I didn’t invent this framework, but when I came accross it, it instantly clicked. At last, I found an approach to answer the question that obsessed me: how to make a game that resonates with enough players to make a living out of it? Nobody can precisely pinpoint the reasons why certain media achieve the level of success they do, there are so many factors to consider. But I refuse to fall into the nihilism trap that success is all about luck. So I cling to this three-step method for designing Drop Duchy, it worked and so here’s how to apply it to your games as well.

Step 1: Making a Compelling Promise

I firmly believe that before you write a single line of code, you need to know precisely how you’re going to catch gamers’ attention. It boils down to two essential components: the genre anchor and the twist. A good promise is simple, concrete, and instantly legible. Balatro did it with “a roguelite where you score poker hands”. Most players know what roguelite and poker are, but how do they combine? That’s the twist that grabs their interest. Drop Duchy has a similar approach: it’s a “tactical kingdom-building puzzle roguelite.” Stray is “a cat adventure game in a futuristic city”. You get it.

Why do lots of indies and AAA studios fumble here? In most cases, I think it comes down to “creative ego”. They either think of a “twist” that only they can appreciate (”it’s set in the super cool universe we’ve created and you play that cool character we wrote that you don’t know yet, but trust me, he’s super cool”). Or, on the contrary, they butcher their anchor, throwing ideas that contradict the experience that this genre’s audience expects.

Finding a compelling promise to build your game upon requires good market research. Players have played other titles before yours, and they’re only going to spend money on you if they sense that your game will make them feel the good emotions they already had before. Market awareness isn’t copying, it’s understanding the emotional terrain players inhabit. If you’re making a colony sim, you’d better understand why players love RimWorld’s emergent chaos and why they criticize its micromanagement. Most of your audience has played RimWorld, and those who don’t have little chance of suddenly becoming interested in the genre because of your game.

Speaking of which, not all promises are equal. Some are stronger than others, some resonate with larger audiences, but the most important thing is to be clear about yours because you’ll need it to know what actually to build. I knew from the start that Drop Duchy was going to sell as “tactical depth in a tiny grid,” so my mission was to ensure that everything from the look to the loop reinforces that promise. Without such clarity, development quickly becomes a scattershot of cool ideas with no unifying purpose.

Step 2: Delivering on It

Once you figure out your promise, your next job is to make players feel it immediately. Players should visualize the core fantasy in a single screenshot, in the opening shot of the trailer, and in the first 10 minutes of gameplay. The GIF of Drop Duchy that I was sending to publishers in my pitch email showed sliding a Lumber Camp tile to gather wood from forest, which tells everything there is to know about the game.

That’s why you have to know precisely the promise you’re making: if players come for strategic depth, don’t bury them under long lore cutscenes, give them what they want ASAP. Look at Factorio: no exposition, just automation from the first moment. Or Black Flag: an epic naval battle to kick things off (before marrooning you to reset the progression). If you’re promising incredible boss fights, do not gate that behind 20 hours of creep grinding. That doesn’t mean you can’t add depth to your systems (see step 3), but holding onto your promise for longer than needed is the surest way to disappoint!

Step 2 is also the moment where feasibility matters: you can only promise something you can actually deliver. Too many developers sabotage themselves by making promises that require massive amounts of content or complex systems they don’t have the resources to develop. That’s the reason why you spent time analysing the market’s expectations. For instance, in 2025, competitive multiplayer demands robust matchmaking, balanced gameplay, and anti-cheat measures. If you can’t do it, you’d better rethink your approach (back to step 1!).

Knowing your promise helps you focus your efforts towards what matters. The guy behind Vampire Survivors knew that his game promised to throw tons of juicy VFX on-screen all at once; he wasted no effort polishing individual sprites. With this mindset, cutting features becomes effortless: anything that doesn’t reinforce the promise can be eliminated. We had many temptations, many suggestions to make Drop Duchy bigger by adding a story or a PVP mode, but we kept its scope focused precisely because we knew what its central appeal was (placing pieces strategically on a rectangular grid). If there’s room to go beyond your promise, that’s good, though, because you’ll need it for step 3.

Step 3: Exceeding Expectations

Once the game fulfills its promise, you earn the right to surprise players. Nothing kills momentum faster than a game whose first two hours are the best ones, then the rest is the same thing rehashed with a different coat of paint. Said otherwise, and I’m paraphrasing Wartales’ game director here, you regularly need to add more code, not just data. Introduce new mechanics, new systems, add rules and modifiers to the existing ones, don’t just repeat the same experience in a different environment.

So, how exactly do you “exceed expectations” when you’re supposed to put most of your efforts into “delivering the promise”? There are two main approaches: surprise and deepen.

Surprising the players is hard to nail down, but it provides the best results if done correctly. The best example that comes to mind here is Portal adding a memorable story to a puzzle game, and doing so at low cost (their story relies entirely on solid voice acting, with no character animation or cinematics).

For most games, Portal 2 included, exceeding expectations comes from deepening the core gameplay loop. To sustain long-term interest, you have to think of elements that enhance the original promise rather than contradict it. For instance, if your game is built around strategic depth like Drop Duchy, the enhancement can be new synergies, new starting conditions, or fresh endgame conditions that require different thinking. What you don’t want is a genre pivot halfway through: a city-builder that suddenly becomes a story-driven RPG, where progress is driven by narrative quests rather than problem-solving.

This one is rarely the issue in modern games because creators tend to have more ideas than they can afford anyway. To find out where you need to invest to deepen your experience, you can either play your own game or read reviews for your closest competitors: you’ll easily grasp why their players didn’t stick out longer.

Wrap-Up

Summarizing “good game design” in a single article is impossible, but I hope this framework will help you step back from the projects you’re making and find clarity. Thinking in terms of “promises” helps me tremendously, as you can tell. I think I’m even somewhat addicted to it now; everything I see is a promise. Each game is a promise, but then each part of the game is also its own promise.

Whatever you’re doing, you’re setting expectations, and your responsibility as a creator is to fulfill them. The worst is that you’ll be held accountable even if you don’t realize you’re making a promise! How did I learn that? Well, I’ve worked (twice) on Skull and Bones, a game that uses pirate imagery to sell itself but doesn’t offer boarding or sword-fighting mechanics.

Before writing this piece, I attempted to find the blog post but was unsuccessful. It may have been removed from the internet now. It’s a bit sad because it has such a profound impact on my game designer’s life, and I’m often referencing it when I share feedback on friends’ games. I suppose the best I can do is to write my own version of it, relying heavily on my experience designing Drop Duchy.

Why Scarcity Matters

Scarcity creates a unique “double tension” that fuels motivation in two gameplay areas: collecting and spending.

First, when collecting resources, you pay closer attention because every unit of resource matters. In Drop Duchy, for instance, you collect resources by completing lines and triggering buildings’ effects. Knowing that resources are scarce pushes you to try to maximize both “sources” at once, which isn’t achievable if you don’t observe and plan.

Second, when spending resources, you must make careful choices precisely because resources are precious. You’re not an autopilot; you’re doing a mental list of every option and weighing which one is worth more. These two tensions feed each other in a continuous loop, keeping players in a state of flow where every decision has weight.

To increase scarcity, the immediate designer’s reflex is to change the balancing. And sure, balancing is necessary, but it’s not sufficient. For instance, ensuring that players have more cards to upgrade than resources at their disposal only matters if each card upgrade feels meaningful. If all roads seem to lead to the same place, players will just pick at random. The same applies if one path seems clearly superior; that option gets picked, and there’s no tension.

Likewise, simply raising prices doesn’t create compelling decisions—it makes the game harder but not more interesting. Your goal as the designer of a strategy game isn’t to use scarcity to starve players, but to present them with dilemmas that make them think.

Scarcity and Opportunity Costs

To feel like true meaningful choices, spending resources must involve real opportunity costs. If a resource is dedicated to a single purpose, the choice is trivial: you either have it or you don’t. In that case, scarcity only helps to incentivize the grind, but you don’t feel strong emotions when spending it. In Drop Duchy, there is a single-use resource for the faction Republic, and it acts more as a power-gating mechanism than an additional strategic layer.

Complexity emerges when resources have multiple potential uses. Players must weigh trade-offs, deciding which goal to prioritize and which to sacrifice. For example, in Drop Duchy, gold can be used to upgrade cards or refill defense. However, you feel bad when you have to refill your defense because you’re explicitly seeing what you lose (an upgraded card). The threat of losing the run is distant and abstract, but this one is concrete, and you promise yourself that the next time, you’ll collect more gold, or avoid losing points of defense, or both. It motivates you to do better so that you won’t feel the pain of abandoning a compelling option again.

Early on, Drop Duchy was correctly balanced, but it lacked the mechanics that introduced such opportunity costs. For instance, buildings were unlimited: each round, you’d see all of them. Adding a slot system introduced scarcity, forcing players to make meaningful trade-offs: do you take more production or military? Short-term power or long-term growth? Pay to unlock one more slot or pay to upgrade current buildings? Now that you’re not guaranteed to see your buildings all the time, picking which to upgrade depends a lot on which you intend to select in the future rounds. Spending resources became less about what you gain and more about what you won’t get: that tension is a powerful motivator.

Everything is a Resource

Once you adopt this mindset, you realize that everything is a resource (not just economic currencies), and therefore, each is an opportunity to create scarcity and deepen the strategy. What do you think is the most crucial resource in Drop Duchy? The grid space. You have a limited number of spots, and what you put in each shapes every decision. At first, it’s wide open, but as the round progresses, that resource is more and more scarce.

Additionally, each piece brings its own rules, lending “meaning” to each of these spots. This one is adjacent to a forest, but also aligned with an Alarm Tower and also connected to a river. As players accumulate multiple buildings and technology cards, maximizing every piece becomes impossible. Sometimes you find a perfect placement (which feels great!), and other times you must choose the least bad option.

In most cases, you’re sitting in between, trying your best to arrange the grid in a way that yields you more than if you placed it semi-randomly. Obviously, as a designer, you want the game to reward players for optimizing: it’s the reason why most card effects we designed aren’t “yes/no” but “for each”. Pieces that always trigger or are too flexible remove depth; carefully designed rules create meaningful tension. You want to tickle that feeling of “I could have done better” after each round.

Quick note before ending this article: irreversibility is a good (albeit risky) tool to heighten the stakes. In Drop Duchy, once a piece is placed, it cannot be moved. Randomness adds an element of risk, requiring players to plan for uncertainty and adapt to unexpected outcomes. Do you opt for the low-return safe military building placement, or gamble on a risky spot hoping the right tile comes later? Making choices when you don’t have all the information can be frustrating, but consequences matter when it comes to creating an engaging decision.

Conclusion

If your strategy game feels flat, the problem is likely not a lack of choices, but a lack of meaningful ones. Balancing better or adding more options rarely helps. You have to sit down and evaluate the whole chain of “cause & consequences” of the decisions you present to the player: what makes “collecting” more compelling is often a stronger “spending”. And vice versa.

I hope you liked this piece. I’m glad I took the time to write it after such a long drought on the Arcade Artificer. If you’d like to read more about game design, based on my learnings on Drop Duchy or not, let me know in the comments (and make sure to subscribe too). See you!

Changing Plans (Summer 2024)

The life of the indie developer (or any entrepreneur) is all about crafting perfect plans that give you a clear vision of where you need to go and how to get there… then having external circumstances crush it into a thousand pieces before you have time to execute.

When we announced Drop Duchy in July 2024, it was nearing completion. We had delivered everything we’d agreed on with the publisher, and we were putting the finishing touches (most notably, debugging and balancing). The closing phase of a project is crucial, and in my opinion, the most challenging to navigate. Often, a dev team rushes to finish as much as possible, either because they’re late or because they want to improve the experience until the last minute. It wasn’t our case; we were pretty happy with what we had delivered, and our minds were focused on the big question: what do we do next?

Once you’ve started your indie studio, finding money becomes an obsession, not by greed, but by necessity. Having found a publisher for a first project is very cool, but publishers aren’t charities; they need to recoup their investment, and they won't pay you for work outside of the agreed-upon project. You can’t count on the first game being a hit either (or at least it’s a risky approach), so you need another viable project ready to roll as soon as the first one is done.

Suffice to say that having to split my mind between the conception of a new project and the polish of Drop Duchy was draining, it felt like those last kilometers in the marathon where you’re pushing yourself harder because you know that the deliverance of the finish line is close. But then something unexpected happened. Our publisher offered to delay Drop Duchy, and I’ll be honest: that made me feel terrible.

I’m not blaming them; they had solid reasons to suggest this because the game’s potential was there, but still unfulfilled in some regards. Still, the delay was ruining our plans and made me feel like all those efforts to juggle between both projects were in vain. On the bright side, their commitment to help us ship the best possible game was also a fantastic opportunity, so of course, we agreed, froze the other project, and came up with a new roadmap: Drop Duchy was now due to release in early 2025.

How the Extra Time Made the Game 10x Better (Fall 2024)

After the summer vacations, I hired (again) a game designer to help me with the content design, and a programmer to handle the more difficult technical tasks (such as connecting to Steam Online Services).

As for what we changed during that time: do you know the saying “a game is never finished, only abandoned”? While the phrasing is a bit extreme, I live by this philosophy. It teaches us that you need to accept that you’ll never get to the bottom of the to-do list, and it’s okay. If an opportunity arises such as this one, you’ll have plenty of options to choose from to make the game better.

We had three main areas of improvement to focus on for this “new production” phase:

1. The military aspect of the game was taking up too much space compared to resources, which impacted the variety of available strategies

2. Deckbuilding synergies were insufficient; it was hard to find satisfying combos

3. The game suffered from a lack of variety inside a run and offered few incentives to play again after winning

Interestingly, those three areas require three different iteration approaches

The first problem was the easiest to fix because we could test our new ideas (skip combat, heal button, mercenaries) quickly. I’m generally not a fan of this approach, but we’ve decided to implement the changes ASAP rather than wasting time on evaluating their impact. In the worst-case scenario, it doesn’t work, and we revert the change. But the new build was actually cool, so we were happy, and voilà. Sometimes (often?) the simplest options are the best.

The second issue required a bigger rework of the systems. The main culprit, in my opinion, was the terrain system because having up to five different terrain types in a round meant you were seeing only a few pieces of each. Spreading the jam thin was negatively impacting your ability to exploit your cards. Our solution was to introduce the “terrain choice,” where at the end of the act, you’d eliminate one terrain. It was a scary change with a risk of ruining the entire economy, which ultimately led us to add numerous cards to the game to compensate. However, we knew it was for the best, as it became much easier to specialize in a particular terrain type, and therefore also improved the variety problem.

Speaking of adding more content, we’ve used the extra time to create new buildings that provide more varied combat challenges during the run, such as making you lose resources or transforming terrains. We also tried to create a “modifier” system, but to be honest, adding extra rules to non-boss encounters added clutter more than depth. We kept the idea of modifiers to introduce an alternative game mode: the Trials. Those challenging runs each have their own specificity, making them ideal for players who get bored with the Normal mode. Oh, and by the way, we’ve added difficulty modes as well. No, we didn’t have those before the delay. What a colossal mistake it would have been not to let the player customize their experience.

This period was the best time we had on the project; we were becoming increasingly confident about our choices and were very productive, thanks to the experience we had accumulated throughout the production. Within this timeframe, we’ve also found solutions to longstanding problems, such as combat and the unlock system, but that’s a story too long to tell. What you need to keep in mind: the game was getting better and better every day, it was now ready to enter its final phase (for real this time).

Demo, Balancing, Demo, Polish (Winter 2025)

As our game was shaping up and our feature list was complete, the new focus became balancing the game. To put it simply, it is a daunting task for a game like Drop Duchy, with its numerous interconnected systems and extensive content. I was lucky to have Lucas, the game designer, for that task. He spent most of his working hours testing, filling spreadsheets, testing again, tweaking the numbers, and testing some more.

At this time, we were also preparing the demo, which is the most critical marketing asset you can have for an indie game. Players want to test a game to assess if it’s worth their money. When the first demo was launched in mid-January, the feedback was great, but not perfect. We were hovering at 80% positive reviews on Steam, and most of the negative reviews came from frustrated people who wanted to enjoy the concept but didn’t. We went back to work right after to improve that aspect. I’ve detailed all the changes in a devlog if you’re curious.

In short, when we set aside some of the usual suspects (poor sign & feedback, unclear tutorials, bugs), the biggest issue was that we were crushing the player’s ego. Having played hundreds of hours and being surrounded by people who were genuinely trying to understand the rules and figure out the strategies (publisher, playtesters), I had never realized how harsh the game could feel to a new player.

When you play Hades and you die, you can always blame yourself for getting hit by attacks you should have dodged, but you will rarely realize your character build was ineffective. As you gain experience in the game, you realize how much grabbing the right upgrades matters, and you get better at making those strategic decisions.

Drop Duchy doesn’t have that mechanic skill, so the only thing that new players blame for performing poorly is the RNG or the poor balancing. The game doesn’t explicitly tell you how unoptimized you’ve played, and there’s no direct loss in case of poor performance: it’s harder to pinpoint what you did wrong. For the second demo presented at the February Steam Next Fest, we improved all of these aspects and received much better reviews.

We had over 100,000 players try the demo, and the playtime metrics were excellent too: some players spent dozens of hours playing such a limited demo. Always a good sign for a game!

It’s release time! (Spring 2025)

Since the end of the (second) production phase, part of the team was working on the new project that we had frozen for a few months. I can’t tell you much about it yet, but we’ve continued to refine the idea we had back in the summer, incorporating various inspirations and figuring out the best approach to convince publishers. We had ambitions and knew that we needed to make a solid prototype to prove our vision if we wanted to secure funding. There was enough money (= time) to create a prototype and pitch it to publishers, but not a lot.

When May 5th arrived, I was frantically refreshing the page and watching the player counts grow. The positive reviews came very fast, after all players had already enjoyed the demo. At that moment, I must admit, I wasn’t sure if the numbers I was seeing were good or not; I was just happy to see the game not being a failure. Releasing a game gives such a rush of dopamine; it feels incredibly good.

Anyway, it was only after the publisher congratulated me and provided points of comparison that I realized the game was also on track to be a financial success. It’s not Balatro, nobody thought it’d be, but it found its audience, and the players are happy about it. Mission accomplished.

And Now?

What’s next, you may ask? Well, well, it looks like our plans need to change again. We’ve received some bug reports and suggestions (which we’ve addressed in patches), and then we went back to the drawing board to prepare updates for the coming months. Putting the game in the player’s hands has re-ignited my passion for the project, and unexpectedly, I’m more motivated to continue working on it than to shift to something else. I still have tons of ideas left on my to-do list. If you want to follow our current efforts on the game, you can subscribe to the studio’s newsletter here.

And what about the other project? It’s not frozen, but progress has slowed down (again). I need to revise the entire roadmap and business plan to assess our current position, determine the number of people we can hire in the studio, and gauge how far we can develop this new project before requiring external funding. I told you: money is an obsession when you’re in the video game business.

Finally, I can talk to you about the game I've been working on for the last year and a half: Drop Duchy, a hybrid roguelite mixing various recognizable influences into an original formula (trailer here). It has just been announced, and I couldn't talk about it earlier because I had agreed with my publisher to keep it secret until then, but damn, it was hard!

More importantly, let me introduce you to Sleepy Mill Studio, the company I founded and the spotlight of today's article. In the following story, you'll, of course, learn about Drop Duchy's production since it's the studio's first (and so far, only) game, but consider this piece as the “making-of.”

Rest assured, however, that the “developer commentary” will come as well: I'll regularly share more details about the game and the design choices I made in the studio newsletter (subscribe there if you'd like to read about that). So, without further ado, let's dig into Sleepy Mill Studio's ambitious objectives, creeping doubts, and surprising victories of its debut.

End of 2022: The Spark

The story begins in September 2022 when I'm about to leave my Game Director position on the Little Big Adventure Remake, and I wonder what I should do next. Taking another Lead Design position was a viable option, but another idea floated in my mind: what if I tried to launch my own indie studio, too? I'd undoubtedly face new challenges going down this route, yet the opportunity to design my ideal job and challenge myself seemed appealing.

I wanted to take my time to make a decision before diving headfirst into the entrepreneur's life, so my first decision was to study the waters carefully—first things first: the money. In France, we have access to unemployment benefits, and the government is more than happy if you use the monthly allowance to survive while launching a business. Of course, this represents a fraction of my previous salary, but combined with my partner's support and my personal savings, it would only reduce my lifestyle a little (buying second-hand is excellent for the planet anyway).

During the first months, I met people, completed training in video game marketing, and read about the industry's business side. Even if I knew a lot about designing and delivering a game, I would need more than my AAA experience in a massive company to understand how to emerge and survive as a new indie in the 2020s. My biggest learning during this period was how important it is to pick the right genre and game concept, depending on your situation.

In parallel, I also did a lot of introspection, asking myself about the type of company I wanted to build, which games I'd like to create, and how. It's easy to declare what you don't want to be (that's what we did at [2.21]), but if you don't intentionally design your work environment to yield the desired results, these big statements remain wishful thinking. My vision for Sleepy Mill Studio was the following: a game studio with a small & expert team that releases games of different genres every two years maximum, thanks to a smart design & production approach. The closest comparison in the industry would be SuperGiant or Shiro Games.

Setting this goal narrowed down the “how” & “what” significantly. I decided the best option was to bootstrap the company (i.e., to start making the product alone rather than raising funds like startups do). Even on a diverse market such as Steam, the genres/gameplay that were both doable for a micro-studio and had good potential were limited. At this point, I planned to release a first game alone within 3-6 months to 1) prove myself I can do it and 2) gain credibility in the eyes of publishers for a future bigger project.

The temptation to jump on my first idea was strong, and I'm glad I resisted the urge until I had five distinct game concepts of various genres to collect my friend's opinions. As I mentioned, picking the right project is the single biggest decision of a new game studio: the danger of having a single idea is how you'll convince yourself it's the best thing ever (and your friends will politely agree). When you have several options, people will gladly tell you which one sucks more and warn you about the pitfalls you didn't anticipate.

The feedback made me realize all five ideas were still too big for a single dev (who had no art skills beyond a few tutorials), so I took one and distilled it into an even smaller concept: Drop Duchy, a hybrid roguelite-puzzle-medieval-citybuilder. Mixing familiar mechanics (which are relatively easy to produce) in a never-seen-before cocktail was ideal for my studio's debut title. Or so I thought.

Beginning of 2023: From Idea to Concept

I started working on the prototype on January 2nd, 2023, and since you're reading this in 2024 and the game wasn't released a year ago, you're wondering what happened. There are three reasons for Drop Duchy's changing course.

First, while it's easy to recreate simple ingredients, finding the right recipe to cook them together in a delicious cake was much more complicated than anticipated. Turning a one-liner idea into a fully-fledged concept took time and iterations: the deckbuilding components, for instance, weren't there from the start. The design process was both challenging and fascinating to me, and as mentioned, I will write detailed dev blogs about it in the studio newsletter.

Second, I was only a designer, not a programmer, and certainly not an artist: the eagerness to learn can't compensate for my lack of experience. After a few weeks, I had to admit the “rush” strategy wouldn't work: even if I could develop things fast, I had a high scrap rate, and the game wasn't progressing fast. During this stage, the most ‘costly’ mistake was picking a 2D-pixel art style and even commissioning an artwork (with my own personal money) … before realizing it looked like crap and switching to 3D.

Third, I realized that I didn't just want to rush-release games with unfulfilled potential: execution quality matters to me about as much as production efficiency & unique concept. I also admire indies who release quick projects and don't care that only a fraction will be good enough to be successful (such as Edmund McMillen or the Sokpop Collective), but it wasn't my vision. Even if I could ship the game and, by a miracle, it got noticed as an indie gem, it wouldn't help me gain the credibility I needed to seek funds from publishers for the next ones.

The strategic pivot came after I received this fantastic advice from a friend who also founded his company: “if you've got a good gameplay prototype, go ahead and pitch publishers. Best case they fund your project, worst case you've got a foot in the door for your next try”. That sounds so obvious in hindsight, but you've got to consider that at the time, I didn't have an artist on the team and zero money even to purchase a couple of concept art. I wrongly believed it would be an automatic no for publishers. Now that I think about it, the fun prototype is obviously better for my project's typology than a good-looking pitch deck.

Summer 2023: Demo & Publisher Pitching

With this new focus, I laid down a plan for a solid demo and carefully examined the best options to increase the game's scope (what does a 15€ game have compared to a 5€ one?) while keeping things manageable for a small team. I was still committed to iterating on the game's core concept, striking the right balance between its components. What is the essence of each pillar, and how do I showcase it in the pitch demo?

The build was improving rapidly, the game was fun and replayable (everything was already procedural), and my low-poly art was enough to make the gameplay readable. Friends and family's feedback helped tremendously; each extra pair of eyes brought to my attention new improvements … and often showed me that I was, at times, focusing on issues that weren't nearly as critical as I thought. For instance, I had become weirdly obsessed with the narrative logic of the game, which is not something players care about in an abstract game.

I have nothing special to tell you about the pitching process itself. I followed all the advice I could find on the internet, and after all those years as a AAA game designer, writing a convincing pitch deck wasn't too complicated: it's close to writing a document for the director's validation. Studying the market beforehand made things effortless; I was confident in my main message: “Here are some successful games in the sub-genre; they worked because X & Y, and I'm doing the same.”

Once again, I'm lucky to have a network of friends and old colleagues in various game companies who 1) gave me feedback and 2) offered to put me in touch with potential publishers. The first publishers’ responses I got sounded interested, but you can never tell if that's a genuine curiosity to learn more or if they're being polite “just in case.”

However, one publisher was unequivocally hooked and believed in the project's potential: The Arcade Crew (the indie label of retro-focused publisher Dotemu). The CEO, Cyrille Imbert, had played the demo extensively, and I could tell when we discussed that he didn't just “politely” review the build but actually enjoyed it and was invested in the strategies. Given my design background and my personal ambitions, being able to find partners who understand the project's intentions and not just see the market potential was a huge relief. We quickly agreed on the publishing terms. Now, Sleepy Mill Studio was entering a new area: the solo venture became a company with resources (and a contract to deliver a game).

End of 2023/Spring 2024: Assembling the Team & Full Production

The first milestone we agreed on was turning the pitch prototype into a proper vertical slice -representative of the final product- to reduce production risk and have a solid version to convince business & marketing partners (not everyone can see potential through placeholders). It was the perfect goal to hire a team; all I had to tell candidates about the job was: “See this fun prototype? I want you to contribute to making it as great & beautiful as possible”.

I already knew from my previous positions that finding the right people to hire was a time-consuming process, and now I had the added pressure of the budget: if the person was not suitable for the job, I could sink the whole company (yeah, I love being dramatic). Once again, quality takes time. I started by calling a game designer I already had worked with at Ubisoft and proposed to have him work on the content. While I kept focusing on the systems/flow/UX, he designed the type of cards and synergies that would make for varied strategies in the game.

Shortly after, I looked for an art director to establish the vision, then found a 3D artist who specialized in the environment. I also found people who handled sound design/music during the same period. Each time, I had the same approach: find varied profiles, identify how skilled & autonomous they were, and ask myself, “What would the game look like if I let them do their best on the project?”. For example, here's the work of the three artists I hired for a test to bring their creative touch.

Again, the game was improving quickly, and I woke up every day with the energizing feeling of seeing my silly prototype turning into a fantastic game. At the end of January 2024, thirteen months after the start of the project, Drop Duchy had a great vertical slice build. We were proud of it, and the publisher's reception was also excellent: they offered to extend the deadline/budget to give us more time to polish and improve the game.

The deadline extension was essential because, despite the achievements, I also noticed how my responsibilities were piling up and creating a stressful work environment for me. I wasn't working during the weekends or at night; we had delivered everything we had promised, but I needed to figure out how to handle the workload in the long term. There was one crucial responsibility I was neglecting: preparing for the future. How can I want a company that makes several games and be burning myself on the first one? When I was solo, I had time and nothing to lose. Now I have a team to pay, and money burns fast, so I had better bet on the right horse.

Unfortunately, I failed to find the right person to delegate part of my work and hired someone who wasn't the right fit. When I realized it was already too late to start the hiring process again, I had to compensate with even more work. Still, the production went smoothly as we transitioned from the vertical slice (focus on quality) to the alpha (focus on quantity). Everybody knew what to do, the team was working efficiently, and we delivered what we had to on time & on budget. With the massive increase in game scope came new issues, as often in game production. For the balancing, for instance, I hired a freelance designer to play the game extensively & give actionable feedback, something I definitely couldn't fit into my packed schedule.

At the time of writing this, we're about to complete the production, leaving ample time for a polish phase. In this phase, we'll take a step back, listen to players’ feedback (especially from the demo), and continue tweaking the game. So far, the playtest and publisher feedback are excellent; it's only a matter of nailing the landing. The hardest part of the journey is behind us now.

Summer 2024: What's Next?

Now that Drop Duchy has been announced, you can expect us to communicate regularly on social media about it and, as mentioned earlier, to read content about the design & features of the game in the studio newsletter & the Steam dev logs. I look forward to discussing the game with the community when the demo comes out in a few weeks. Hopefully, people enjoy it as much as we do. Drop Duchy will continue to be a significant focus in the coming months, and I plan to learn a ton during the pre & post-release marketing campaign. If the game succeeds, we have big plans for console ports, expansions, and maybe even a sequel.

As I progressively shift towards kickstarting the next project, I'll have opportunities to write case studies and share my thoughts about making games here on The Arcade Artificer. The newsletter isn't dead, even if I'm not writing as often as I used to. Thanks for reading me, and if you've got questions or suggestions for future posts, let me know in the comments.

[EDIT 2025] The sequel article is here!

As I wrote in my last article about Skull & Bones (which was secretly more about the evolution of AAA games), the consumer’s expectations have shifted throughout the years. Graphical fidelity set aside, what was considered hit-worthy back in 2004 would undoubtedly be deemed “too short” or “too shallow” in 2024. Since players have discovered the sweet taste of depth & freedom, they are less interested in linear missions and games that end as soon as the story ends.

If we set aside online experience, the genres that grew most in popularity are those that offer systemic experiences. What does this mean? How does a “systemic” approach enrich experiences? And how do we design it properly? That’s what we’ll see today, and I’ve found the ideal game to illustrate its principles.

System vs Systemic

First things first, let’s take a moment to clarify what “systemic” means. If you’ve read my Intro to Rational Game Design, you may notice some similarities: that’s because RGD is often a sub-discipline of systemic design.

A system is “a group of elements interacting with one another according to principles”. For instance, a democratic political system is a group of elements such as citizens, elections and laws and principles such as “voting” (where citizens elect representatives to more efficiently establish the regulations that will govern all the citizens).

Most aspects of a game can be “systems”: movement, combat, progression, or AI are some of the most common. Save is a system, user interface is a system. Video games are entirely made of systems because, by definition, they’re software created by people who coded principles to establish how elements interact.

However, “systemic” doesn’t just mean “featuring systems”; that would be too easy. In common game development language, a system is “systemic” only if it features significant interactions between its key elements, which have a major impact on the player’s experience.

For instance, “pressing a button to pick up a healing pack” is an interaction (button press) that involves elements (healing pack, and technically the avatar, too) and principles (refilling health gauge). Even if healing is crucial to succeed (major impact) in this example game, this mechanic isn’t “systemic” because “A → B” is a basic interaction, not a significant one. If the player had varied options (through a crafting interface, for instance), the pick-up might be considered systemic.

The systemic design usually correlates with two essential things:

• An increased sense of simulation: that’s when the game makes you believe more in the logic of the virtual space. Traffic cars can run into one another, causing chaos and police to intervene.

• Emergence happens when players can find interactions that the designers may not have directly intended. A problem has several solutions.

From the creator’s perspective, “systemic” is the opposite of “authored”: they delegate the player the responsibility of finding the best way to play the game. They make a solid systemic toolbox and trust players’ creativity and intellect to figure it out.

What Makes a Good System?

So, how do you create engaging systems? Lucky you, I’ve recently played the ideal game to illustrate the rest of this article: Nobody Saves the World from Drinkbox Studios (the makers of Guacamelee).

Nobody is a standard top-down RPG (think old-school Diablo or Zelda) with a unique hook: you can unlock 15 different characters (called “forms”) and switch from one another whenever you want to tackle the game’s challenges. There’s a tortoise that can go on water, a slug which leaves a poisonous trail, a magician and even a dragon! Each of these forms has unique gameplay rules; part of the fun is to unlock and discover new ones as you progress.

Switching forms isn’t a new concept in an adventure (Lego series) or platforming games (Super Mario), but associating it with the RPG genre is quite original. Moreover, it’s ideal for this article as the “form system” exemplifies the three pillars of a good system: simplicity, modularity, and interdependencies.

Simplicity comes first. The elements and rules of the system should be easy to understand, not because you want to trivialize the experience, but on the contrary, because you want the depth to emerge from players’ interactions. An unintuitive system isn’t fun and doesn’t spark creativity, so you’d better keep things as readable & consistent as possible. In Nobody, the controller’s buttons are the same for all characters, so switching is about discovering new abilities rather than getting used to new control schemes. The game also relies heavily on pre-existing player knowledge rather than inventing new things (animals, elements, names, etc.)

Then there’s the modularity. When elements and mechanics can be broken down into small, reusable components, you can combine them in infinite ways without exploding the production cost (that’s one of the goals of object-oriented programming). Nobody has over 100 character abilities, but they aren’t entirely unique if you look at their components.

“Hitting” is a component, and “dealing damages” is a component; that’s the Knight’s basic attack. The Monk has an entirely different ability: an area of effect mechanic that heals. But if you re-combine these separate elements, you get a hit that heals (the Mechanic’s repair ability) and an area that damages enemies (the Ghost’s basic attack). Thinking modular is the perfect way to maintain simplicity while offering tons of possibilities.

Finally comes the interdependencies. Modules are great when used to make varied abilities, but they become even greater when the player can create their own combination! In Nobody, passives typically add cause & effects with words such as “if” or “when”. If you use both the” when you hit, also applies the stunned status effect” and “enemies with a status effect applied take +30% damages”, you’ve created a new interaction between systems. As a player, designing their own “machines” (in this case, combat machines) isn’t just about picking the strongest abilities but also picking ones that interact with one another. It requires solid programming and careful attention to balancing from the creators, but the systemic experience feels fantastic when it clicks.

Toying with the System

If you recall the first part (it wasn’t that long ago), a game can feature well-designed systems without technically being a “systemic” experience. For instance, “car traffic” may be a simple, modular & interdependent system yet only serve as realistic background decor. The key to enabling systemic play is to design the experience to encourage players to interact with the systems. The worst thing that can happen to a designer is creating a great system full of possibilities only to see players use the same few tools throughout the game (thinking of you, Watch Dogs: Legion).

In Nobody, the team put a lot of effort into the variety of enemies and combat modifiers. Some forms have natural edges in some situations, and the damage-type system is a not-so-subtle way to force you to use specific attacks against certain enemies. Regarding movement, there are a few restrictions: you can go anywhere with anyone except for the few places requiring swimming, flying & using tiny tunnels.

The true genius of the game lies in its progression system: there’s the classic XP system to increase stats and unlock gated areas, but there are also Form quests that provide a progression for each form.

The first quests of a form are quick & easy, acting like a tutorial to teach the basic gameplay. Completing them unlocks another ability for the form, then quickly another again and maybe another form too. The progression pacing is quick enough that you stay engaged even if the gameplay of this particular character isn’t as interesting. This design approach works wonderfully to take the player out of their comfort zone (a character they’re used to and already have levelled up).

But anyway, chances are you’ll find yourself strangely addicted to what could be described as “a giant skinner box”, and you’ll want to level up all the forms to maximize the game potential. Another genius idea here is the form customization, which allows you to equip up to four active & four passive abilities. Two of them are locked, but you’re free to equip any other form’s ability for the others. Not only does it encourage theory crafting the most efficient build and identifying devastating combo, but it also encourages levelling up characters you don’t play, so their upgraded abilities help another build.

This design philosophy certainly required a ton of balancing work, but it works perfectly in the end. It is also supported by the “Custom quests,” which are mandatory to progress and require using (or sometimes figuring out) interesting synergies between abilities from different forms. Again, this contributes to the constant switching between characters: levelling up this one to unlock that ability, completing a quest with another one and then unlocking another ability, etc. It is a fantastic web of intertwined motivations.

Wrap Up

I usually like to end these articles with some bonus information, so I tried to contact the Drinkbox team to exchange with their designers, but unfortunately, I never got a reply. Lucky me, there’s an interview on Game Developer where Lead Designer Ian Campbell shares nuggets of design wisdom.

My favourite part is when he explains how the original vision was closer to an RPG. It allowed players to change form and abilities only at save points, so players must commit to a specific build before entering a dungeon (and if they fail, change and try again). However, they gradually realized it was frustrating and did not play to the game’s strengths, so they allowed players to change anything anywhere.

While it sounds obvious in hindsight, anyone who’s shipped a game can relate to the pain of not knowing whether to address feedback by doubling down on your original vision or abandoning it. And believe me, the correct answer isn’t always “stop being stubborn and accept the player’s suggestions”, so props to them for figuring it out!

The first time was during my time at the Ubisoft editorial department. Then I moved on and joined another AAA production -Ghost Recon: Breakpoint- and after its release, I returned to the project, this time to lead the game design in the Paris studio co-development team. My second time on Skull & Bones ended, and I resigned from Ubisoft to join a newly founded startup. I stayed less than a year there, but that was still enough time to kickstart the pre-production of an ambitious adventure game. When I left the startup, I decided to create my own studio, worked hard to make a prototype, raised funds and hired a team to work on the full production of our first game.

And now, after all that, Skull & Bones is finally out.

Needless to say, my contribution to the project feels like an eternity ago. I've had a lot of time to think about the project, its design, and its production, and it significantly shaped me as a game designer.

Since I still need to play it, I won't discuss the game's faults in this article. Instead, I'd like to share my opinion on how it feels to work on such AAAA projects, which face increasing hostility from gamers, and why I believe Skull & Bones is much more than another: it's a symbol of a changing industry.

Why Skull And Bones is More Complex Than You Think

Since the day Skull & Bones was first announced, there has been an apparent mismatch between the players’ expectations and what the company could -or wanted to- do. By “company”, I mean everyone, from the developers to the creative leadership and the higher executives. We're not here to blame anyone in particular. These conflicting viewpoints on what the game should be continue to this day. Look at the reviews: there is no consensus. On the one hand, some people enjoy the game for what it is, and on the other hand, some people hate it for everything it's not.

The most common complaint is the comparison to Assassin's Creed: Black Flag. They say, “How come a 2024 game can achieve less than its decade-old inspiration?” And I'd answer, “It doesn't”. Skull & Bones focuses on different areas than Black Flag, which are incredibly complex to develop. The three main culprits? Online, systemic, open-ended.

First, making an online game where players can log in & out at will tends to create many unexpected problems where the usual shortcuts you'd use in a solo game are impossible to implement. For instance, here's one issue we had to solve with the Paris team: what happens if your ship gets sunk while you're on a beach? Do you lose it all and respawn far away? That would be unfair… Maybe your boat is resistant when docked then, but that introduces an exploit in PVP scenarios. If we let the other sink and loot you, we could respawn you with your inventory back, so nobody gets frustrated. Oh, but we've accidentally created a duplicate glitch exploit by doing so.

Everything is more complex online, and it doesn't get any better when the game is systemic too, as in “players have the freedom to do whatever, go wherever, unlock whatever progression”. In systemic games, even something as mundane as an NPC dialogue line to celebrate your last trip needs extensive rules & hundreds of recorded lines. You can't just comment on the heavily scripted previous story mission. If you're wondering why AAA productions need dozens of game designers, you have the beginning of an answer: to think of all the gazillion problems that may happen and prepare monstrous plans to cover all the edge cases.

To give a more concrete example, I also experienced while working on the project: making a cool chase mission. On the one hand, and I don't want to diminish Black Flag's team fantastic work, you can spawn the enemy AI ship discreetly during a cinematic, reload to a checkpoint if the player messes up and tweak the enemy speed according to the speed of the only ship available to the player. On the other hand, you can't do any of this, and other players (friendly or hostile) may also interfere in many ways with the mission. It takes a lot more time & effort to design systems that maintain the activity's fun, pacing and fairness.

And finally, the trickiest of all: you need the game to be open-ended, meaning the activities should be replayable indefinitely and in (almost) any order. Designing for variety & depth has nothing to do with finding the immediate fun for a 10-minute story mission.

The Strategic Mistake of AAAA Games

Game development isn't a contest of “who had it harder”. I'm not trying to find excuses for Skull & Bones in particular: if you give the same constraints to another studio/publisher, they may have done better; who knows?

The more important underlying question is, “Why”? Why do companies persist with AAAA-online-GaaS-live-service-behemoths that cause more harm than good? After the failures of Anthem, Marvel Avengers, Suicide Squad, Redfall and many others, why continue trying out?

The immediate answer could be “greedy publishers who want to maximize shareholder profits, “but it doesn't paint the whole picture. It's true that big companies don't just want to make games profitable but try to hit the home run with money-printing machines, which tend to be live-services games. However, the current gaming market also has an underlying truth that encourages them this way: mainstream gamers want it all (then complain about it).

Let's be honest for a minute: would people really accept a Black Flag ++ these days? In the 2020s, the quality standards required to convince players that a story-based game is worth 70$ are so much higher than they used to be in the past. Only a few studios worldwide, most of which work on Playstation exclusives, have the technical, artistic & narrative excellence to produce such games. Being open world was previously enough of an argument to counter-balance a weaker foundation (read Black Flag's reviews for a reminder of what people thought of the story, the characters, the repetitive activities, etc.)

In the current gaming market, if you want to appeal to a broad audience and justify your price tag, your best bet is to make the game online, systemic & open-ended. Don't get me wrong, that's a great thing: even if I'm more of a solo player myself, I also dreamed of playing Assassin's Creed coop with my brothers and having endless content to play. As the industry progressed and newer games achieved more, it has become harder to accept simpler experiences. Exceptions continue to exist, sure, but the norm for AAA games has changed.

And unfortunately, it dooms them.

Can The AAA Industry Change?

If you're familiar with game development (and I assume you are since you're reading a game developer newsletter), you've heard this a million times this advice: start small, don't over scope, find your audience, and expand over time. Part of the reason why live services fail is because they want to skip the “start small” part. Unfortunately, trying to build a skyscraper on shaky foundations is the recipe for a sure collapse.

Here's an interesting thought experiment: what if, instead of launching a first lackluster Assassin's Creed and then making a sequel that improves on all fronts, they delayed the first by two years to make it « perfect » from the start? I’m 99% certain the commercial performance of this super-AC1 wouldn’t have been that much higher, and certainly nowhere near the two episodes cumulated. So, this hypothetical super-game would have been considered less profitable and refrained Ubisoft from greenlighting a sequel. This is what big companies seem to have forgotten these days: the creative process isn't a linear function of time & money: no delay can replace the power of shipping a product, taking a step back and doing it again.

There are several causes for this. First, teams get exhausted and burned out working on the same product for an extended time. The more time passes, the higher the risk of turnover on release and, therefore, losing precious insights on how to improve for the sequel. Second, big budgets massively increase the financial pressure and, consequently, the odds of designing a game that appeals to everybody but pleases nobody. Third, chasing infinitely replayable systems typically forces one to spread the jam thin on the bread. Every gamer has figured by now that systems such as loot or procedural missions make games longer but not more fun.

Of course, the graal is to figure out a way to have both the quantity and the quality. Some franchises successfully evolved in that direction (such as Assassin's Creed or GTA), but it took several instalments to get there, it's impossible to go for everything right away. The team progressively learned what makes their IP the most entertaining; every opus gave them key learnings.

What's the solution, then? How can we get high-quality new AAA again? I suggest trying out building new franchises through smaller & faster projects, letting the team learn and grow until they can eventually hit the jackpot. Companies that operate like this today are the winners of tomorrow. Elden Ring would never have been a massive hit without all of From Software's previous games. The same goes for Supergiant and their game Hades (not a AAA budget, I know, but definitely AAA revenues) or Baldur's Gate 3 or Helldivers 2.

Now, it's a bold assumption to assume that AAA companies can become more flexible overnight and accept to make (slightly) unprofitable games. Hence, the solution is probably to hope for indies to scale up rather than AAA to scale down. After all, the big companies all started as what we'd consider indies nowadays.

Conclusion

So, what's the real tragedy of Skull and Bones? A subpar game isn't anything new; we have some every other week. But this one is different: it feels like the end of an era. Due to the development length and all the expectations that it should be “just” a Black Flag sequel, Skull and Bones represents all the games which will never exist because the window of opportunity has passed.

There used to be a time when games could be hits with “just” a cool fantasy, great gameplay mechanics and a nice story to follow, set within a free-roaming open world you could explore between missions. And that would arguably be perfect for our childish view of pirates’ adventures (we've seen the potential thanks to Black Flag). There used to be a handful of studios capable of delivering such experiences accross the globe, and now they have either shifted their focus or now work only on their in-house properties.

And don't get me wrong, I'm not whining about video game companies being “anti-consumer” or “misunderstanding expectations”. This is just the reality of an evolving market, player changing tastes and spending their money differently. The 2010s open-world era was always set to be replaced by something more modern, just like the 2D platformers who used to dominate the 90s eventually faded away from the mainstream market. Until, of course, indies picked the genre back up.

Suika Game came out on Nintendo Switch in Japan in 2021, and in the summer of 2023, it took off and ranked as the best-selling on eShop in September of that year. Watermelon Game (the literal translation) is the port of a video projector game whose gameplay mixes Tetris and 2048 (with physics). The mechanic is simple: you drop fruits in a closed rectangular area, and whenever two of the same type touch, they merge into a bigger fruit. Your goal is to get to the biggest fruit, the watermelon, a more challenging feat than it seems at first glance.

In this issue of The Arcade Artificer, we won't scratch the surface to unearth a “magic” underlying system. There's no secret that makes it entertaining, but there is still a lot to learn from it. I wanted to write about the watermelon game because it exemplifies several fundamental design principles that I notice & try to apply when it comes to entertaining games. Here are five crucial lessons that came to my mind while I was compulsively merging those damn fruits.

#1 - Clear Visual Design

Let's start with the most obvious yet often overlooked: the screen is always readable. Thanks to the use of simple shapes & colours, you can instantly distinguish fruits from one another and understand how they fit in the play area (which is one of the keys to great physics, see point #4). Only a few UI elements are on the sides, while the central area remains laser-focused.

You may object that it's easier to achieve clarity when you've got such simple rules, but I'd disagree for two reasons. First, it's part of the game designer's job to propose mechanics that can be exposed visually rather than introducing rules and leaving it to the UI/tutorial to figure out how they'll be represented.

Second, even in this game, there are lots of details which could have gone wrong. The “dropping line”, for instance, could have been a preview of how the fruit will bounce like in pool games, or maybe it could have been eliminated, but the team found the proper middle ground, and it's for the better: imagine how different the experience would feel like in those scenarios.

#2 - Correct Use of Theme

Finding a suitable theme for an abstract experience can take time and effort. From an accessibility standpoint, it's not better than numbers we can immediately compare & classify (like in 2048). They compensated for this by including the fruit chart in the bottom right to help you remember how fruits chain.

However, the fruit-based theme is an excellent idea because it adds flavour (literally). Everybody knows fruit, they're colourful & you've got an intuition for their relative side, so there's a sense of logic & order (more about that later), which adds to the satisfying game feel. Crafting an entertaining experience also involves finding visuals that delight the player, and when they can help to inform the mechanics, that's even better. Did you know that board games frequently start without a theme, and a publisher may pick a design prototype and then figure out the art that fits?

#3 - An Obvious Core Conflict

Here's something where many other games fail: an immediate tension emerges from Suika's core mechanics. What's a “core tension”, you may ask. I'd define it as such: to be compelling, any choice-based game must present a strong duality, with two sides at odds, and put the player in charge of pushing the system towards the “winning” side. If there's no clear right & wrong, players may do stuff, but they won't feel strong emotions. It doesn't mean that every action has to give immediate feedback, there can be long-term goals & strategies, but there needs to be at least a good correlation between the player's actions and the desired outcome.

The typical duality in many game genres is the conflict between life & death, but here, the watermelon game has another: chaos vs order. To win, you must position fruits in the correct sequence: if you intertwine another fruit, the ones that should merge will never touch (thus feeling the space for nothing). Despite physics, it's a mostly stable system: nothing will move unless you do something. After each fruit you add, you can tell whether you contributed to bringing order (i.e. you merged it or you're about to merge it) or chaos. Playing this game feels rewarding because you're constantly contributing to a raging battle, and every step of the way, you ask yourself: “Am I winning or losing?”

#4 - Meeting & Breaking Expectations

Mentioning “chaos” naturally leads us to the next point: the use of physics in the game. The principles are simple: it's only 2D discs reacting to one another, and it's ideal to fulfil its role in the experience: being predictive, yet sometimes a little bit surprising. Achieving this balance is more important than it seems at a glance, and when you play one of the clones, you can intuitively feel how the differences in physics tuning impact the experience.

Humans are reasonably good at estimating how two balls bounce against one another without needing a Physics PhD. Since our baby years, our brain has conditioned us to enjoy when we can correctly anticipate something and see it unfold. When you finally get that big chain of merging fruits, you're not surprised but still enjoy it: we're still inside the kids laughing at their parents doing peek-a-boo.

However, making the game too easy to predict comes at the risk of making it dull. What separates the watermelon game from 2048 is the little dose of variation the physic allows: a minor imprecision can lead to unexpected outcomes. It frequently happens when fruits merge and push one another: whether you get a good or bad surprise, you feel an extra dose of emotions.

#5 - Pacing & Randomness

On breaking expectations, there's another interesting difference with 2048: the fruit you get to drop isn't always the smallest but any of the first five. In 2048 terms, that's equivalent to getting tiles from 2 to 32, and it's a smart design choice to reduce “spamming” and shorten the duration of a game. The goal of Suika is to organize the fruits; what's the point if they are already? In a game that strives for predictive output, having a little bit of input randomness is ideal for renewing the challenge & keeping you focused.

But wait, from what I experienced, it seems that the sequence of fruits isn't completely random: when you get a smaller one, you frequently see another of the same type or the one right above. By manipulating the odds, the game still maintains a good pacing: each fruit is a challenge, but some are easier than others. Indeed, input randomness is a plus for strategy, but if it's too chaotic, it gets frustrating & counter-productive.

Final Point

Does clever design pay off? Well, the company reported having sold three million copies on Switch alone, and since they keep topping the charts, they're on their way to gaining a lot more. Did copycats on mobile & PC cannibalize the sales? It may be counter-intuitive, but those probably boost the popularity of the concept & help to spread the watermelon craze. In a way, the clones act like a free demo, and they aren't going to make millions with no monetization but a bunch of ads anyway.

If you liked this piece, you may also enjoy this one:

I'm a big advocate of studying how great titles work in order to learn their tricks and sharpen our craft, but a bad game can teach you just as much. What went wrong? Why is it missing the mark?

I recently grabbed Marvel Avengers, a game under fire since its announcement for attempting to turn a beloved franchise into a “Game as a Service”, which in their case meant bizarre RPG mechanics, tons of DLC cosmetics and endless repetitive missions. Gamers are suspicious of this type of game, but if the core gameplay & the fantasy are solid, they'll gladly accept to sink hundreds of hours into them. So what happened?

In this issue of The Arcade Artificer, I'm taking things differently. First, you will read the detailed account of what went through my mind during the first hour of the campaign. Then, I'll recap the key issues and quickly touch on what comes after the intro, which helps us see how the problems of Marvel Avengers extend far beyond corporate greed.

The First Hour in Details

False Start (5 minutes in)

Like any decent AAA, the game opens with its settings menu and tons of accessibility features, which is fantastic for those who need them. Too bad the main menu, on the other hand, isn't all that accessible. I'm unsure what “Operations” means, so I click the “Play Now” button.

After a 5-minutes loading screen, I get a cinematic montage which drops tons of story exposition on me. What is that? A band-aid to help non-Marvel fans catch up on the universe? It drags on & on, seemingly recounting major story beats, and that makes me doubt my choice.

When the cinematics ends, this screen warns me that I'm in post-game mode and that it spoils the campaign. Huh, thanks, but too late? Why didn't you show that in the main menu?

Proper Start (11 minutes in)

Ok, no big deal, we're in a classic case of “the UX gets messy after three years of post-launch updates.”: back to the main menu, I found the main campaign by scrolling through the “Operations”.

In the (actual) intro cinematic, we meet Kamala & her dad: she's a young fan of the Avengers who got selected to participate in a fan fiction contest and was invited to meet her idols in the festival they hold on their base of operation. She's afraid of being ridiculed, so her dad gives her a motivational speech to comfort her. It feels somewhat artificial, to the point I'm wondering if the dad is about to die to give her superhero backstory (if you don't know, Kamala will become Ms Marvel).

Gameplay! (16 minutes in)

I now get to control Kamala, and my first impression of the gameplay is good. It's not Naughty-Dog-level good, but the camera & movement feel polished.

Not even twenty seconds after being taught how to run, I get a tutorial about… photo mode. That's an uncommon learning order, but it's consistent with the context of this intro sequence, I guess.

The lady at the entrance gave me my first proper objective: collect five comic books through the festival to gain access to the VIP balcony. Hmmmm, ok, it feels forced again, but it's not the dumbest tutorial quest I've ever seen in a game.

The small festival area offers some mini-games where I can earn comics as a reward: they serve as shooting & QTE tutorials. A popup also explains how the comics give stat boost: again, I wonder if this is the right moment to teach me about that; I'm trying to get immersed into a universe and don't need a first exposure to the RPG layer.

Exposition dump (25 minutes in)

Then, the game remembers you've paid for an Avengers experience and decides to introduce the heroes you love. Kamala first meets Thor, who acts like an arrogant jerk (because she was checking an Iron Man toy).

Then Captain America witnesses Kamala standing her ground against some bullies and comes to praise her courage. Black Widow shows up and asks him, "Who's the kid?” to which he doesn't reply (to my great disappointment…) “What a dumb question, she's obviously one of the invited kids to our frigging festival, and her name is written on her badge”.

At this point, I realize that the writers are rushing to showcase as much of Kamala's personality as possible and give her scenes to interact with the superheroes, but it feels more like hospital kids meeting Make-A-Wish cosplayers.

And it gets even worse when Kamala sneaks backstage to collect the last comic book and -of course- overhears a totally secret conversation between Bruce Banner/Hulk and two scientists who fear their new great technology isn't ready just yet. Just to double up, the designers throw at you a second scene seconds later where a journalist asks Nick Fury about the safety of this new secret technology. Oh, I wonder what'll happen next.

It's happening? (36 minutes in)

I finally get to the VIP Area, a small & lame balcony where we can't even sit (but at least the view is good). Blah blah blah press conference about the super technology, blah blah blah Iron Man shows up and interrupts like a jerk too, and…. finally, some action! An explosion happens on the Golden Gate Bridge in the background, which might finally free me from this long & flat opening sequence.

First hero impressions (42 minutes in)

Quick cinematic, and I then get to control Thor: the combat gameplay is solid, with satisfying animations and great VFX. Grunts rush towards me, and I dispatch them with ease. Don't feel any tension or sense of danger so far, but it doesn't seem to be the point of the encounter anyway.

However, two minutes later, I'm surprised when the game allows me to use my “Heroic” ability already: too soon? It's not exactly a tutorial since the game doesn't tell you how that works (cooldown?), but it's more of a showcase of a cool effect.

Iron Man time (47 minutes in)

Before I have time to get used to Thor gameplay, the game switches to Iron Man, flying and shooting bad guys above the bridge. The heroes keep talking while fighting, which normally annoys me, but I guess it's consistent with the Avengers franchise.

The game continues to move relentlessly, adding new enemy archetypes and mechanics constantly… which I forget as soon as the next one arrives. It's the opposite of Batman Arkham, which would let you get used to dealing with certain enemies before introducing new counters & abilities to renew the challenge.

Moving to Hulk (50 minutes in)

The Iron Man segment also ends with a tutorial on his Heroic power, which again feels undeserved.

Then comes Hulk with a short sequence where you run accross the destroyed bridge. The whole scene is chaotic, and the cinematics casually insert shots of endangered civilians between two funny dialogue lines, but the game still doesn't want to take its time. Run, punch, heroic power, move on.

Captain America in Danger (53 minutes in)

Switching to Cap provides a nice change of pace as he's still inside the vessel, and enemies attack him in a smaller arena, making the combat feel different and more strategic: I have to play smart & avoid the attacks from the newly introduced archetype.

The downside of including an actually engaging gameplay sequence is that I barely listened to what the characters were saying over the radio, but then comes a cinematic with a clear message: a bomb-loaded truck is rushing to the city, and we need to stop it! I'm unsure how they can get anywhere with such a destroyed bridge, but it's Black Widow time anyway.

Chasing Taskmaster (the bad guy driving the truck) is another well-executed sequence, mixing cinematics with explosive gameplay … but the thrill comes suddenly to a halt when the actual boss fight begins. I have to avoid his flying attacks & shoot back, but the window of opportunity is so small that it takes a lot of back & forth until I'm allowed to move to the second phase.

The second phase is classic combat, showing again that when the game slows down, the mechanics can be deep & challenging. Chaining attacks, blocks, rolls & other abilities makes for a satisfying flow when it ticks. I'm still unsure why the game insists on throwing even more tutorials at me; it even pauses the combat to show a popup at some point: just like the story, I have the feeling they want to dump all the exposition and be done with it.

End of the introduction (1 hour & 5 minutes in)

The first hour ends with a cinematic where we learn that the bomb truck was only a diversion; the real danger was the weird force field emanating from the bay that caught the vessel. It makes the dangerous new secret technology explode, and the shockwave impacts the whole city, causing random people to develop superpower abilities.

During the credits, we learn that politicians turn against the heroes for “parading their untested technology” (I mean, sure, but are we going to ignore the whole terror attack?). With Captain America likely dead, the Avengers disband, and the backstage scientist seen earlier has kinda replaced them with his company AIM by promising to “bring science to the world & fix the mess caused by the heroes”. Why isn't he accused of reckless behaviour, too? How is that more efficient/less dangerous than superheroes who have done enough good deeds in the past to the point people are fans of them?

So, What Went Wrong?

With this detailed breakdown, I hope to have shown that the problems of the introduction of Marvel Avengers have nothing to do with its ‘game as a service’ orientation: those “annoyances” show up later in the campaign. Besides the minor hipcups, there are two main reasons to explain why the introduction is so underwhelming: the convoluted story concept and the unimaginative gameplay sequences.

First, picking Kamala as the protagonist could work on paper, but only if the studio figured out a better context than this festival to introduce her. Kamala's biggest personality trait is her unconditional admiration for the Avengers, and seeing her lack of self-confidence about the fan fiction contest isn't enough to help us connect more with her. It feels too distant & artificial, something we're told rather than something we see in action. As soon as the explosion happens, we forget about her anyway, and it feels like our time was wasted (despite the fact she's the main character of the whole campaign!).

The “Avengers plot” is also complicated, and, as a result, there is a lot of exposition dump throughout the introduction. The game has to explain why the Avengers fell out of grace, how the scientists of AIM rose to prominence and why new superheroes emerge everywhere. This is a compressed first act rather than a prologue, and it feels indigestible. I'm not saying it's impossible to present all of this, but doing so (while introducing Kamala) meant, in this case, that scenes had to be stitched together unnaturally. It doesn't make for an entertaining experience.

This brings us to the second big issue: how this story is conveyed through pretext objectives. There's nothing wrong about cinematics delivering unexpected twists, but as a rule of thumb, to make players feel something, it's better to make them play the thing!

Kamala is obsessed with the Avengers? She could illegally infiltrate the Chimera rather than being a perfect girl who collects comics in an Avengers festival. And while we're at it, something weird happens when she gets close to the technology, hinting at the fact that it might give people superpowers. “Show, don't tell.”

The same goes for the heroes: you could easily imagine a sequence where you have to protect the technology from hordes of enemies but eventually fail, then Captain America heroically sacrifices himself to give a chance to the rest of the crew. That'd be an excellent way to make you play those key events rather than take all your attention to the bridge: ironically, this distracts the heroes and the player.

In a nutshell, this introduction is subpar because of design problems, not because of the game concept & Destiny inspiration. The team is clearly talented, can craft outstanding sequences, and has solid combat gameplay. It doesn't pay off in the first hour of the game for a reason (that might be outside of their control).

Aftermath

After writing this, I continued playing the campaign, which isn't perfect but still enjoyable. We follow Kamala/Ms Marvel, who tries to reunite the Avengers again & prove their innocence. The story eventually solves the events of the intro, but nothing that makes you go “Oh I understand now, they couldn't do this sequence any other way”.

The gameplay sequences are well-crafted, and the combat continues to shine. The RPG looting mechanics are indeed annoying & out-of-place, so we feel the identity crisis of a game that couldn't choose between Destiny & Uncharted. I didn't try the coop missions & DLC.

One last thing I noticed during the campaign is how they repeat some tutorials again, but this time allowing the player to digest the mechanics. Why put everything in the introduction, then? Did they initially intend to use the campaign as a lengthy introduction to the coop missions, then feared that the players who skip the story wouldn't understand the game?

One thing is sure: this game had a massive potential, and it unfortunately doesn't fulfil it.

If you liked this post, you may also enjoy:

Taking time to set myself on a course and regularly evaluating my progress, method & chances of success is an excellent way to clarify my priorities and stay focused. However, my brain also tends to hate unfinished stuff, which is often helpful and sometimes annoying when I should drop an objective before wasting too many resources.

There are three situations where you should drop an objective. The first is easy on paper and complex in practice. The second is the easiest to spot, yet often poorly handled. And the third is probably the most challenging for the human ego (at least for mine). Let’s dig into it.

Cursed Objectives

“Cursed objectives” are fundamentally flawed goals which can never be attained because they’re built on conflicting premises. For instance, I want to “be a sumotori who runs marathons” has two individual components that contradict on a deep level. Motivation & training can’t always overcome human limits; you have to choose one or the other at some point. (Note: I discovered after writing that a sumotori did try this challenge but walked for the majority of the marathon).

It’s never as evident in the game industry, but we see many of those, probably because there is a galaxy of nebulous sub-objectives behind every objective. The role of the game designer is to establish pillars & intentions for each part of the game using specific vocabulary, listing success criteria and anything suitable. If you skip this part, you risk spending forever iterating towards a utopia. There isn’t always a “third way”, a “sweet spot”, or other expressions managers love to say to make you work harder.

So, you must learn early in your career to spot them. The biggest indicator that you’re facing a cursed problem is if you feel that pursuing a given part of the objective takes you further from the rest. The second trap I’ve often encountered is what I’d call “the eluded basics”: when we start a project, we tend to write down the specific guidelines & intentions about the game but not the mandatory basics (such as “accessible”, “pleasing the fans of the brand”, “memorable story”, etc.) I’d always advocate writing everything, even if it sounds dumb: I can guarantee that many indies would be better off if they clearly stated that one of their goals is to make the game financially viable for the people working on it.

When to abandon a cursed goal? As soon as you spot one! You’ll typically drop a part of the objective and keep the coherent one, giving you a sudden boost of energy, as if you dropped a weight (literally).

Impossible Objectives

Sometimes, there’s nothing wrong with your goal, but you still won’t attain it because you lack the necessary resources: it’s what I call an “impossible objective”. It’s typically another side of the same not-identifying-properly-your-constraints coin than the above. AAA are often guilty of it: after all, if some other studio pulled it off, why not go for it? I see two main reasons: expertise & motivation.

People often believe that time is the most crucial resource to manage. They can hire to increase man-months or throw some overtime to catch up. List tasks, estimate, and trim around the edges. Too few plan their objectives around expertise, despite being the resource everyone lacks: in the game industry, we’re “discovering” a lot more than we’re “optimizing”. Therefore, we all lack expertise when we constantly face uncharted problems. Your game project has a lot of issues that you must learn to solve because the people who already acquired this expertise aren’t around (and they’re busy with something else anyway).

But … aren’t people motivated to take on new challenges? Don’t creative people explicitly dislike staying in a comfort zone? Sure, but motivation is also a resource. You only have so much before it runs out. Motivation isn’t a “mindset”; you can’t add “more” simply by thinking about it. You have to manage it, or it’ll drain faster than you imagine. Making any project rely on motivation is pure madness: only healthy work systems & habits can take you far. See it as such: motivation is the fuel, methodology is the rails. Both are needed for the locomotive to move forward, but burning fuel without rails won’t get it anywhere.

So, what happens when you realize you don’t have the necessary resources to reach your goal? First, don’t hope to compensate it through “better efficiency” or “luck” or anything. These aren’t the resources you lack; it’s just wishful thinking the problems will magically solve themselves. If you don’t have resources to allocate to a goal, abandon it and celebrate everything you’ve accomplished. Even if it wasn’t the end you imagined, you still have learned something valuable that’ll help you down the line.

Unworthy Objectives?

We all have limited resources in life, yet an infinite supply of possible goals, which brings us to the third case: what happens if you have the resources to reach an attainable goal, yet you don’t want to allocate them here?

I’ve always enjoyed discovering new topics and engaging in a variety of activities throughout the year. I set achievable goals for each, such as participating in a half marathon, building a scale model or learning to play a particular piano piece, which gives me a direction and a finish line. The importance of each varies throughout the year, and I can regularly change focus while still enjoying the satisfying sense of completeness.

However, in the past months, I became unfaithful to this principle as one side activity took over a significant portion of my schedule: writing this newsletter. Sticking to a set rhythm was a fantastic system to support my intrinsic motivation, and I certainly enjoy the variety of topics I can explore. Still, committing to drafting, writing, revising, illustrating & publishing a new issue every two weeks occupies a large chunk of my free time. I do not aim to make my writing lucrative; I already have a lot of my work schedule between kickstarting a new game studio and running a game production.

When it’s neither the goal that’s impossible nor the resources that aren’t available, I think it’s important to ask yourself: what did you really want to achieve with this objective? What is the opportunity cost of pursuing it, aka all the other things you could do if you dropped it? Abandoning such goals may be painful because, deep down, you still believe it’s valuable: maybe there’s a way to reframe it in a less demanding way. It happens a lot in game design. You must learn to “kill your darlings” and trust your gut.

What’s Next

Is it the end of this newsletter? No! But it’s evolving.

The biggest change is going to be the schedule. I won’t publish every two weeks like I did so far and switch to a “when it’s ready” policy instead. I still have many ideas and unfinished drafts in the drawer. In the coming months (and years, hopefully), I’ll share more about my own game company/project, sharing advice & helpful behind-the-scenes.

I’d also like to introduce a new type of issue (tentative name “react”), where I’ll share some interesting articles or videos and share my point of view, why I agree or not. It’s akin to writing a long comment on social media without going through the whole writing process.

To avoid missing out, the best way to be warned of new content is to subscribe. I’m not selling your email, but I promise I’ll send you interesting content about game design, the industry & how to make better games!

Subscribe now

In just six words, we’ve got a whole development strategy designed to address the biggest fear of any developer: spending a lot of time to create a game that just isn’t good. A simple mantra for a powerful mindset: when you’ve got an idea, prototype it immediately and playtest it.

If it’s crap, throw it away, you learned a good lesson at minimal cost. If not, congrats: list the areas of fun and work towards expanding those. You either win or don’t lose; how awesome is that? Surely everyone in the industry must agree with the mantra (they usually do…) and apply it in their job routine (… they usually don’t).

Indeed, as you may expect, the reality is more nuanced than a catchphrase can convey in six words. When a recipe looks too easy, it’s suspicious.

Even worse, some people divert it and somehow make it say the opposite, leading to disaster. I’ve had trouble working with this approach on personal projects and had questionable experiences in my professional life, too: here’s what I learned about iteration, when and why it works (or doesn’t) and how to leverage its fantastic potential.

The Limits of Human Creativity

Making games, or innovative consumer products in general, is a backloaded process: you only truly know if the consumer enjoys it after you’ve built it. You can use all the tricks in the book to predict success (and you should), but still, you never get a specification from your clients that tell you everything that should be included or not. Players have wildly different expectations and like to be surprised. It’s your job to design something you confidently think can find an audience. That is a considerable risk since the development of a game takes a lot of time, energy & resources.

So, to seek early reinsurance and encounter the problems while they’re still solvable, we create intermediate versions that resemble the desired final result -with caveats and limited scope- and play it in similar conditions to eventual consumers (internally first, then through external playtests). These sessions are helpful in discovering that features do not fit as well as you imagined; maybe there are some problems you didn’t anticipate, and sometimes it unexpectedly turns out better than you imagined. No problem, now you know it and can work on concrete improvements.

However, confronting our work with negative feedback is challenging for our self-serving brains. We get attached to ideas while they live in our minds because, until contested, they always feel better than they are. The attachment is a trap. It drives us to dismiss criticism and to keep finding reasons why the tangible prototype isn’t comparable to the intangible concepts we have in our mind (”the idea is great, the prototype just needs this, this and that to be better”). The longer we stay attached, the harder it gets to face reality.

The ‘Follow the Fun’ philosophy directly contradicts our natural human bias as it encourages us to stay in the passenger seat for once. We’re still creators, it’s not about letting feedback decide for us, but we try to observe the direction the game drives us, not where we want to go. Doing prototypes faster reduces the emotional attachment and the sunk cost fallacy: it’s still tricky to get the necessary distance, but with a fully open mind, there is a good chance we notice dead ends and unexpected opportunities.

Why It Can Fail

Presented as such, the strategy looks solid & rational, however, since it still involves several human decisions, there are at least three ways it can fail.

#1 - Anything can be fun

As a lead designer, I like to use the analogy “playing football with a can” to explain to colleagues that anything can be fun. If you have never tried, playing football with a can is super enjoyable because 1/aiming & shooting things are intuitively satisfying for humans (for some reason), 2/erratic bouncing creates unpredictable situations, and 3/you’re laughing with friends. The whole activity is fun, but the can on itself isn’t. If you were to develop an entire sport around it, chances are it won’t find an audience.

The same can happen for a video game. When testing a multiplayer or a silly gameplay prototype, you may feel entertained and yet your design might not have anything to do with the fun. Basically, people aren’t always great at understanding the origin of fun; thus, sometimes, “following the fun” might lead you to gimmicky dead ends.

#2 - You can’t always fail fast

Some games & even whole genres aren’t good enough as early prototypes to properly assess if they work.

For instance, highly systemic games (management, 4X, cards, etc.) require a lot of content, inter-connected systems, some balancing and polished enough UX before you can adequately assess their fun. There was a Reddit post (which inspired me to write this article) explaining how a team had a vision for an RTS, but playtesters didn’t like it much, so they steered towards tower defense.

In this case, ‘following the fun’ meant focusing on the more immediate fun because it’s what stands out in an early prototype. Enjoying strategy & assessing depth in an early version requires a lot of projection, which is almost the opposite of the method. Don’t get me wrong; it doesn’t mean you should not seek to confront your ideas, just that if you can’t “fail fast” in a harmless way, you should approach “follow the fun” differently.

#3 - It’s a guide, not a religion

The worst misuse of the strategy is when it serves as an excuse for taking the wrong direction with a product: “Trust us, we’re not there yet, but in the end, it’ll be fun”. It seems to happen more frequently than it should in AAA; some even have a name for it, such as the “Bioware Magic”, as reported by Jason Schreier in his in-depth report on the failure of Anthem.

In a way, it’s recognizing that games aren’t fun at the beginning, but that devs have the power to make anything fun eventually (they don’t). Turnarounds happen frequently, and if you’ve shipped games before, you know how things improve rapidly towards the end. However, using “follow the fun” as an excuse not to challenge concepts early is as risky as it sounds. Some fundamental issues are cursed and can never be fixed.

What You Should Instead

If you feel like your game may fall into one of these traps, here are three ideas to approach « Follow the fun » differently (sorted from most to least controversial).

Follow the Fun … on a Design Document

I know this is the opposite of the enjoyable rapid prototyping strategy we all love, but hear me out.

If you can’t write a short document to convey the intentions of the mechanic/system and get people excited about it (yourself included), how will it be fun in the game? Just write what you have on your mind, leave it to rest for at least a week and come back. What is the more interesting aspect of it? Can you emphasize it? What’s the least? Can you scrap it?

It takes more projection efforts, but since it costs fewer resources, you should also feel liberated to eliminate anything that doesn’t bring enough joy. Don’t write okay-ish designs and hope to turn them around through prototype iteration. Fail faster.

Follow a Variety of Fun

As explained in the first part, we tend to get too attached to our ideas, to the point we corner ourselves to conceiving a single possibility. While confidence is great, it can be an efficient technique (although less fast) to also brainstorm and prototype alternatives.

Artists work this way all the time, making several sketches before picking the one they prefer and enriching it with the good parts they noticed from the discarded ones. Even if you have a preference from the start, exploring different directions can help reinforce the selected concept and reinsure you. Knowing you have fallback alternatives can be liberating to dig into the exotic options: if it goes poorly, you can always return to the safer plan.

Follow the Fun of the Fun

Finding the fun is easier when you already have an idea of what it looks like. Using the method to discover new fun mechanics out of the blue is a tricky endeavour for inexperienced devs. For instance, if Ubisoft Montreal didn’t release Assassin’s Creed with its flaws, I’m not sure they could have figured out all the improvements needed to elevate the franchise and make a fantastic sequel.

I’ve heard once in a conference that a game evolves so rapidly that even the team making it has an entirely new understanding every three months or so. To create additional content and expand on solid foundations, it might be better to wait until you’ve released a decent initial version and let the dust settle in. Trying to get all the answers too early is as risky as hoping they’ll all come magically at the end.

Wrap Up

You know the problem with doctrines: when they are too simple, they turn into religious scriptures that need to be analyzed to become useful (like I just did…). If you’re at the beginning of your career and didn’t form a strong opinion yet, I’d always advocate maintaining distance with whatever you read or learn about game dev.

Use several sources, and analyze your experiences. There’s no one-size-fits-all here: don’t prototype for the sake of it, and don’t spend forever waiting for the perfect idea. There’s a right in-between for each project, and it’s part of the job to figure it out. Good luck, and have fun!

I wasn’t a fan of the Tour when I was younger, I thought it was the most boring way to waste summer afternoons. Now that I discovered you don’t have to -actually- watch the whole thing, I’m really into it: watching the key moments is sufficient to understand everything that’s going on in this fantastic three-week race.

Like any other sport (and e-sport), there are fascinating storylines regarding the riders, with underdogs, favourites, underwhelming results & surprising victories. However, this format of cycling race (le Tour, Giro & Vuelta) is also a unique game worth studying: which rules are in place to keep the competition fresh & exciting for riders & spectators? Which strategies emerge, and how do designers respond?

In this issue, I’m giving you the keys to understanding the game of Tour de France and how its design evolved over time.

Cycling Basics & Break Formation

The Tour de France is, first & foremost, an endurance race, with about 3500 km split into 21 stages over the course of three weeks. The pacing is relentless; riders have an average speed above 40 km/h (despite climbing many mountains). About 20% of riders abandon the Tour before the end, either after an injury or because they get eliminated for not making it to a stage finish within the maximum allowed delay (usually around 40 minutes after the winner of that stage).

On the road, riders have a supporting staff to give them water & food, as well as team cars following them to assist in case of mechanical or medical issues. It ensures they’re in the best condition, ready to spend their whole energy fighting against two main forces of nature: air friction & gravity.

Drag (or fluid resistance) is the physical reason why cycling is a team sport: when you ride close to one another, the cyclists in the back are protected from air and spend up to 40% less energy. All stage long, you’ll see riders relay at the front of the peloton to pull for a few minutes, then rest while someone else takes the spot.

Sometimes, a group accelerates to form a “breakaway” and get ahead. If you don’t know about the sport, it may sound like they’ve got a much better chance of winning, but the reality is more nuanced: the break riders need to make more efforts to maintain the gap, especially if they’re only a few in the group with limited possibility to relay. A cooperating peloton will always be able to control the gap & close it eventually, like in stage 7 this year when Pierre Latour went in the break 80 km before the finish, only to get caught 5 km from the end.

However, you have to keep in mind the peloton isn’t a single hive mind: there are actually teams in direct competitions here, and they face dilemmas that fall in the realm of game theory. If you spend all your energy at the front of the peloton to catch up with the break, then the riders who were patiently waiting behind will beat you on the final sprint. The same thing happens in the breakaway, and you will often see them slowing down drastically near the finish line, sometimes to the point they get caught, just because they don’t want to be the person who made all those efforts only to help a competitor win.

Cycling involves a lot of strategies, bluffing, opportunistic alliances & betrayals.

The General Classification and How to Win the Tour

Breakaways may also happen in mountain stages, but it’s completely different. Since air friction isn’t as important as gravity in steep slopes, the peloton doesn’t have an edge: what matters more is the individual effort. Mountain stages are the most important of the Tour since this is where the best riders gain time on their opponent in the general classification, whose leader wears the famous Maillot Jaune.

This is why the Tour de France doesn’t follow along the coasts & frontiers like the first editions but instead always goes through the Alps & Pyrenees, where the highest climbs are located. Mountains are the key stages, so the course designer ensures these happen during weekends (when most spectators are watching). Thierry Gouvenou, the race director, works all year long to design an ideal route that’ll give the best spectacle, juggling with many constraints & figuring a way to make it as diverse as possible, like a level designer would.

How do you win the Tour then? Well, it’s a team effort. Several teams revolve entirely around a leader who has a fair shot at the general classification, and seven mates riders will do everything to preserve their “leader” for 95% of the race for him to have the most energy in the challenging climbs. In this year’s edition, we had a clear duel between reigning champion Jonas Vingegaard & previous runner-up - and winner in 2020 & 2021- Tadej Pogacar, but a bunch of other teams were also going for general classification & aiming at the podium.

In the flat or hilly stages, teammates pay attention to the riders who get in the break and will chase immediately if a direct competitor tries to go for it (which is known as “defending” or “controlling”). Riders who are allowed to get in the breakaway usually lag far behind in the classification.

In such an environment, crashing is almost the bigger threat; hence you’ll often see the GC leaders ride at the front to limit the risk, even if they don’t really want to sprint for that stage. There are bonus seconds to gain at specific checkpoints to spice things up, and it could almost have decided the Tour this year since the two favourites were less than 10 seconds apart at the end of the second week.

In the mountain, teammates are still essential to set the pace, motivate you & help ride downhill, but the individual effort is the only thing that matters in the end. We’ve seen it in the particular time trial of stage 16, where the best two dominated the others completely. The gap between Vingegaard & Pogacar was already almost impossible to catch up, then it got even worse in the following stage, where he lost five more minutes. In an endurance race, you’re one bad day away from losing. Pogacar finishes second in the Tour again, claiming the White jersey (awarded to the best rider under 25 years old) for the fourth time.

The Games Within the Game

There are 176 riders in the peloton at the start, but only one Maillot Jaune for the best overall time, and very few teams/riders capable of contending, so what do the others do?

First, winning a stage on the Tour de France is also very prestigious, and it often comes down to having the right team tactics or getting in the breakaways (as explained in the first part). On the flat finishes, when the peloton is likely to arrive as a group, the best candidates for stage wins are the “sprinters”, a category of riders capable of short bursts of massive efforts.

As you can imagine, your chances increase drastically if you are in the front of the peloton in the final kilometre and if you can have a teammate riding in front of you to save your energy until the last second. The best sprinter this Tour was Jasper Philipsen, who won four stages & came second twice. He’s a skilled sprinter and had the best teammate in the category, Mathieu Van Der Poel.

Sprinters are also ranked on the “points classification”, a separate ranking system whose leader wears the Green jersey. The first 15 riders to cross the line each day score points, and there are intermediate sprints throughout the race to incentivize early efforts & competition (rather than saving it all only for the finish). It’s interesting to note that the balancing has evolved: the points for intermediate sprints have increased a lot since 2011, and the flat finishes give more than mountains to make it almost impossible for non-sprinters to “accidentally” win the Green jersey.

Something similar happened to the other particular ranking, the mountain classification, whose winner wears the Polka dot jersey. In the 1990s, riders would complain that the scoring system encouraged riders who go for early breakaways to grab points in the first climbs of the stage, thus ranking high there without actually being the best climbers at the Tour (which, as explained previously, are normally the same people who win the General Classification). This led the organization to tweak the points awarded: less for smaller climbs & double the points for climbs near the end of the race.

Funnily enough, this caused the opposite problem: the winners of the general kinda “automatically” won the Mountain classification. For instance, in 2021, Tadej Pogacar won with 107 points, 80 of which came from winning two stages in the Pyrenees. Sure, he was the best climber, but for 2022 the organization decided to cancel the “double points” rule. This incentivized other good climbers to animate the race, and the new balancing paid off: in 2023, Giulio Ciccone made it his goal, collected points on many occasions, and eventually won it (despite finishing 32nd on general classification). The Tour may be at its 110th edition, but it continues to receive updates to stay fresh & exciting.

Conclusion

There are many more topics worth digging into, but you get the point. It’s fascinating to me how, even in such a large-scale game, the organizers play a significant role in designing & adjusting the rules to keep the challenge interesting for people to watch & for riders to compete.

One last fun fact: did you know that some riders were purposedly going slower to be ranked at the last place since it was an excellent way to stand out for the public and could lead to invitations to more races afterwards? The organizers didn’t like it much, so in 1939, 1948 & 1980 had a special rule to eliminate the last riders at the end of some stages. No matter the game, players will find a sneaky way to optimize the fun out of it!

If you liked the topic of this article, I’m sure you’d also enjoy this one:

I’ve never participated in a game jam. I’ve been making games for over ten years now, so it wasn’t the lack of opportunities or groups; I just never liked the concept of working all day & all night on prototypes that often end up being too bugged & too basic to be interesting.

Today I’m finally trying out. I don’t want to die stupid. I’ve learned a lot about development in the past months, including coding best practices & making realistic estimates, which gives me the confidence I can produce something worthwhile (without working more than twelve hours a day).

I picked the GMTK Jam because its timing fits my schedule perfectly. Indeed, I had decided to entirely dedicate the first half of 2023 to the pre-production of my first commercial indie game, and now that this is behind me, the jam will be a good moment of break before starting publisher pitches (if you want to connect, hit me up!)

This is also the ideal opportunity to write about the behind-the-scenes, as some readers suggested: finally, you’ll get to glimpse how I approach design, what I focus on, the choices I make, the corners I cut. I’m writing this devlog along the way, with editing afterwards for clarity & conciseness, so you can experience the story as it happened.

The theme drops in half an hour, but I’m not in a hurry anyway; I’m going to the restaurant first, that’ll give some time to brainstorm. It’s definitely not going to be a hardcore all-nighter jam, I hate that.

How I picked a concept (Day 1 - 22h52)

I thought about this year’s theme, “Roles Reversed”, during the dinner. Although I entered the jam solo, having a sparring partner to exchange thoughts about the theme was definitely helpful. So, how do you find a good concept?

First, I quickly eliminated meta-ideas (such as “players switching roles with the devs”) and playing uncommon characters (such as “you play a shopkeeper”) because I always fear they turn into gimmicks. Not saying you can’t find neat execution & make fun prototypes, just that they wouldn’t be as enjoyable for me to create for me. And I guess that’s the first lesson: don’t rush on a concept you’re not 100% comfortable with.

Then, my self-imposed challenge became to “reverse roles” during gameplay itself. “Reverse” isn’t the same as “switch”; it implies some adversity, so I wanted to explore a concept where you play … against yourself? The tricky design question with such a setup is: “How do you win?” If you’re playing two avatars, you can lose on purpose with one and win with the other. So, I started to lean towards “coopetition”, where you play two characters with a shared goal but where actions tend to cause problems to the other, thus feeling like you play against yourself.

After these reflections (and the dinner), I started to scratch down ideas of the form it could take. Shooter? Platformer? Given the short deadline and my immoderate love of board games, I settled for a card game: I immediately pictured how the presentation could convey the conflict between the two identities, with cards that have text upside down & the camera switching side every turn.

As for the theme, I had to be extra-smart about it because I can’t make anything but basic art. I first thought of the life/death duality or something along those lines, but it’s a bit overdone. Again, discard the first ideas if you’re not satisfied, better invest more time now than regret later.

While we were discussing initially, there was this idea of alternating between hunter & hunted, and when I thought about it more, it clicked: why not a ‘Jack the Ripper’ theme? I like the Victorian Era, the atmosphere would stand out compared to the usual colourful card games (and other jam entries), and I could use archive pictures for the art.

Since I had this idea of coopetition towards a shared goal, I meshed this theme with Dr Jekyll & Mister Hyde to create a more original & coherent setup: what if Jack the Ripper worked for Scotland Yard during the day and couldn’t help but kill during the night, causing massive problems for his alter ego? This is the definition of a good theme for me: when it reinforces the mechanical foundations and has a great potential of enriching it further (reading the Wikipedia page helped me a lot to find ideas!).

I scribbled a bunch of notes and brainstormed some core mechanics & systems I could implement. The design “document” (more like a page of notes) look messy, and frankly, I barely referred to it afterwards, but it helped me to settle on ideas & organize my thoughts. Not going to explain it all to you now; you’ll get to see it before your eyes in action. I’m going to sleep now, but for you, the story continues in the next paragraph.

Always start with a plan (Day 2 - 8h25)

After a solid breakfast, I’m sitting in front of my computer, ready to code!

Huh, well, actually, I’m going to start with objectives: there’s a clear deadline, and I can easily make a retro-planning of where I should ideally be at what moment.

Being done with the “main tasks” on second day is ambitious, but I prefer this to under-estimating the landing time, aka all the gazillion secondary tasks that accumulate on top of those. It’s a jam, it doesn’t need to be super-polished, but it’s important to make have an MVP (minimum viable product) as early as possible; it makes you realize earlier what the game needs and not just what you thought you could/should implement back during the paper design phase.

I then took some time to break down the tasks into a big hierarchized list, thinking about what I needed for each topic & how to do it. Again, it might be a loss of time for some, but trust me, rational organization helps to be laser-focused and remember everything in the heat of the moment. By the way, ticking boxes is also a proven psychological trick to 1/keep the motivation, 2/notice when things are slowing down, and 3/take shortcuts because you want to tick boxes faster. It just works.

Slow foundations & basic game feel (Day 2 - 13h32)

For the remainder of the morning, I started a blank Unreal Engine 5 project and worked on the basics: camera & controls. This isn’t the ideal engine for a jam, but it’s the one I know best; again, you’ve got to be comfortable to be productive; this isn’t the time to learn new techniques.

Establishing the foundations took more time than I anticipated. It’s, by definition, the kind of task you perform once per project and somewhat forget about until the next. Here’s what I had going into lunch: a board with two sides, cards that can spawn in hand & on the table, and the ability to move them around (within boundaries).

At this stage, the most important is the game feel: with limited time, I focused on the minimum amount of details I could implement quickly for the biggest impact. Firstly, changing of cursor/colour on hover is a must-have. It immediately makes the game more intuitive too. Secondly, smooth movements go a long way: for that, I used “springs”, some functions that give this sort of elasticity. Cards follow the cursor without getting stuck to it, and they can naturally leave/join the hand.

Cheap UI art & tricky interactions (Day 2 - 20h29)

During the afternoon, I deviated slightly from my plan as I figured out that making the card art after the UX functionality would be more work than doing a decent v1 right away.

So, what’s the technique for cheap UI when you’re not an artist?

• My first idea was to use paint brushes & splashes to make everything looks “organic”; plain colours & geometric shapes would ruin the atmosphere I was going for. I found a free asset pack on opengameart.com, selected the best-looking ones and imported them into the engine.

• Secondly, a good font goes a long way: again, there are lots of free ones on the internet, it just takes a bit of searching to get them. The first I picked fitted the theme, but its “shaky handwritten” look was unreadable, so I also grabbed an “old newspaper style” one.

• Finally, if you’re not an artist, use an existing colour palette. I picked one from the website Lopsec with a good-looking yellow/blue contrast. Using fewer & more coherent colours is a massive shortcut to make your game look better early on.

After that, I continued working on the core mechanics, mainly the card drawing at the beginning of the round. It doesn’t look like it, but spreading them in hand & repositioning them on-demand took was not as straightforward as I thought. Maybe I’ll regret it and could’ve taken shortcuts to implement this in one less hour; we’ll see.

What I know for sure is that I’m late on my schedule: I won’t be able to make much art & sound design, but I can still finish the main features of alpha & turn this interactive prototype into a game with rules.

Day 2 - 23h16 - Taking design shortcuts

Back from dinner, I had another three hours to spend on the prototype before my energy runs out, so even if there was no playable content, I wanted at least to have all the systems ready for it.

While eating, I had a clutch moment and realized I could re-use my hard work on the hand to make a design that’s both simpler & more interesting for the central “encounter” cards. My original idea was still loose. It involved stacking cards together & real-time interactions, just like Stacklands, which I already covered in a previous issue. All of this meant implementing many new mechanics, and I couldn’t tell if that work was worthy.

So, instead, I figured the central cards should also be in a line (re-using the same neat auto-arrange code). Now, the “action cards” you’ve got in hand represent the possibilities, and you can spend them on an encounter card to affect the line. The strategic challenge appears more clearly, and the content starts to design by itself magically: you’d move encounter cards around, flip & discard them to trigger various effects. It’s another reason you pick a genre you’re comfortable with: the inspirations flood in naturally (from board games, in my case).

After implementing that & before going to bed, my last task was to prepare all the economic loops & HUD that goes with it. Again, it’s relatively effortless since those are simple & similar mechanics: you gain & spend points doing stuff, which triggers effects. Action cards cost “time points”, and you only have eight each turn. Suspicion augments or diminishes when performing certain actions, and it triggers a loss if you reach the max. Same (but inverted) for the Duty, which represents how the inspector gets fired if it reaches zero.

Here’s the result at the end of Day 2. You may notice how using the downloaded splash textures & the fonts go a long way to make it feel like a decent UI, even though it’s just text, I would never have time to make cool-looking filling bars/circles anyway.

Timeboxing art (Day 3 - 10h00)

This morning, I decided to kick off with a one-hour burst entirely dedicated to the visuals (it turned into 1h30), where I do the maximum I can and then move on. This powerful technique called “timeboxing” is an efficient way to force yourself to focus on the most important aspects and those with the best efforts/outcome ratio.

The scene I have in mind is simple: a wet paved street with changing lighting between night & day. Since I’m not an expert, it took slightly more time than it should have to nail down, but here’s how I did it:

• Grabbed a texture and slapped it on a plane, without even the “sidewalks” model I thought about at first. Looks a bit weird, but who’s going to notice?

• Downloaded a free lamppost 3D model, added a dynamic spotlight & instanced it a bunch of times on the side

• Followed a tutorial to create a mist-like particle system

• Followed another tutorial to have a puddle/wet look on the floor material using vertex painting

• Coded the function to smoothly change the parameters of the sunlight between low-intensity blue & high-intensity orange-ish. I would have preferred a rainy atmosphere during the day, but no time for that, I gotta get back to gameplay!

Designing & scripting content (Day 3 - 13h05 )

So after the art, I opened a big spreadsheet & started to design a list of cards, their effects & how they interact with one another. I went for the smallest amount of cards that would still feel like a complete game, providing a good variety of interactions & delivering on the thematic aspect (I couldn’t make a Jack the Ripper game without the Whitechapel letters, for instance!).

The first draft was quick to make, less than 45 minutes, since the figures are purely based on gut feeling. Normally, I’d do a bunch of math formulas to spot holes in the economic balancing, but with this little content, I should be able to tweak it fast through playtesting. My goal with this table is, once again, to be organized so I can implement the cards quickly without forgetting anything.

Here, I took another risky shortcut. I’ve written down card descriptions manually rather than coding functions to generate them automatically based on the scripted action. Manual approach is faster initially when the content is slim but requires more maintenance while making iterations. Given the deadline, I wasn’t going to make many changes anyway.

What’s coming next for this last afternoon? Here’s my plan:

1. Test & debug the cards’ interactions

2. Add a winning menu & a main menu

3. Add sound

4. Add art in the background of encounter cards

When this is done, hopefully around 15h30, I’ll package this V1 & upload the build to itch.io, work on the page & submit it to the jam. Even if things go wrong, I’ll have a backup “shipped although not great” build, at the very least. With the remaining time, I’ll continue testing & iterating on whatever feels more important, then export a v2 close to the deadline. Hope this works out; fingers crossed!

We’re done! (Day 3 - 19h05)

My plan did work, except it was a bit ambitious: the page went live at 16h30, however, I skipped the card art part since finding & integrating sound effects took longer than anticipated.

So, after the page, I scrambled to find vintage illustrations quickly, ran them through Photoshop to harmonize the colours a bit & put them as card backgrounds. I’m glad I did because it looks much better than plain backgrounds.

And finally, I used the remaining two hours to test, update the balance & add some more feedback. I figured the suspicion meter was too easy to keep low, so I added another rule to increase it by +1 every turn, which somewhat fixed it. Sometimes, the simplest solutions are the best, I could have revised all the other cards, but a new rule was the correct choice here.

Here’s the walkthrough of the demo where I -almost- win at the end:

However, this rule (and the others) made the game harder to understand; I wish I had more time for pop-ups, tutorials & better feedback. For instance, the loss screen appears as soon as you lose, with no delay to show you which counter is at fault: normally, good feedback requires proper preparation & flow, as I explained here. To compensate, I took to write detailed instructions on the game’s page (which we’re allowed to update after the deadline). I’m unsure if players read those, but it’s better than nothing.

And now, I need are two things: stepping outside & getting some rest. Although the jam was exciting, and I’m happy with the result, it was intense for the eyes, the body & the mind. I feel washed even if I took proper time to eat & sleep. I’ve never been a fan of working this intensely, I’ll never be, but at last, I can say I’ve participated in a game jam.

Epilogue (Day 4 & beyond)

What happened the next day? I checked the jam page to realize I had zero comments & reviews, which is somewhat normal for a jam this big. So I updated the page again, made a better miniature to catch people’s attention and gifs to showcase the gameplay. Yes, you also need to do basic marketing in a jam.

I also tested many games to improve my “karma”, but I’m not sure it worked: with this much competition, if your game doesn’t take off early with strong hooks & visuals, you get buried by the algorithm. I never expected to rank too high anyway with a tactical card game made in solo, with no humour or catchy character.

However, I intend to enrich & polish the game once the rating period is over; there’s a lot I have to explore with this concept, so please have a look & share your honest feedback; it’ll help tremendously. Thanks!

https://sentrydown.itch.io/jack-ripper-at-night

Thanks for reading all the way to the end! This is issue isn’t like the others, it’s more personal, more concrete, what did you think of it? Your comments always help to improve!

I can't remember the first time I heard this “fact”, probably on a forum or during my game studies. All I know is that I just accepted it. It felt a bit sad because, like everyone, I grew up playing so many demos for games I never intended to buy. In a way, it felt like exploiting the system, feeding my insatiable hunger for fresh content while saving money for the bigger fishes. It's not worse than piracy, for sure, but it somehow didn't feel morally right. And that's probably why I easily accepted the idea that “demos hurt sales”.

But it was a lie.

Or, at least, it was a claim made without strong supporting evidence. For this weird reason, what was a staple of video game marketing faded out for years. To this day, some developers keep believing that demos negatively impact games, even when current-day data says otherwise.

In this issue of The Arcade Artificer, we're debunking an old myth, examining its origin and trying to understand why it stuck. In a future article, we'll see why demos are crucial again and how to design them, so feel free to subscribe!

An Inspirational Speech Full of Approximations

I was first exposed to this story by reading this great article from game developer & marketer Tavrox titled “The Game Conference That Cursed Demos for 6 years”. It immediately made me want to dig deeper, and I wasn't disappointed by what I found.

Before going further, just a quick note: in this article, whenever I use the term “demo”, I only mean “consumer-facing demos”, not internal prototypes or vertical slices used to pitch to other industry professionals (publishers, journalists, etc.)

With this clarification out of the way, we can introduce the protagonist of our story: Jesse Schell. You may be familiar with the name of this American game designer & teacher thanks to his best-selling book The Art of Game Design, one of the genre's earliest (and arguably still among the best). Jesse worked at Disney & founded his company, where he serves as a CEO, and he used to give a lot of conferences.

The specific talk we're interested in today was part of the DICE Summit 2013. Mysteriously titled “The Secret Mechanism”, the 20 minutes keynote covers a wide variety of topics sharing the common theme of “things aren't always the way they seem”. Schell makes funny remarks about how he wrongly predicted that iPad was useless and would flop. He also talks about how humans are biased regarding pricing & discounts. There's a lot of vaguely inspirational stuff in there, but the guy certainly has a talent for captivating audiences (I learned while researching for this article that he's also a circus artist & comedian, which explains his talent on stage).

After making a point about how seeing someone wear cool cosmetics in MMO ignites your desire to own the same piece of customization, he jumps to a graph with several Xbox 360 game sales curves. Time slows down as Schell dramatically reveals how games with only a trailer sell 2.5x better than a game with both a trailer and a demo. Boom. Everyone's shocked.

He then elaborates with the following argument: when you see a trailer, you make a plan to play the game (like with MMO cosmetics); therefore, not having a demo available means you have to buy it.

We won't have much more information since Schell switches to yet another topic. Still, it was enough to generate many articles in the press, which were promptly discussed on gaming forums afterwards.

Journalists should have dug further into Schell's source because they would have realized the story was five years old already.

Where does the data come from?

So, below the graph Jesse Schell shows, there's the mention “Data courtesy EEDAR via Geoff Atkin”. EEDAR is a research firm co-founded by Mr Atkin which specializes in collecting factual data on video games: camera type, protagonist genre, presence of multiplayer mode or not, etc.

By collecting only verifiable facts, they could claim that their database & statistical analysis was purely objective. However, EEDAR still crossed their data findings with other industry analyst companies, such as the NPD Group, which specializes in estimating game sales for North America. Estimating physical sales & revenues is a notoriously tricky task (and even more so on a global scale), leading to potentially wide margins of error.

Like Jesse Schell, Geoff Atkin gave lots of conferences, including one in GDC 2013 titled “Awesome Video Game Data”, which provides several insights on the state of the game industry, yet this crucial information about the negative impact of demos on game sales is missing. Why?

With more Google Search, I discovered that the graph shown by Jesse Schell at DICE 2013 comes from a 2008 talk given by EEDAR at MI6, another professional event specializing in game marketing. Unfortunately, we don't have any video recording on the internet, but by searching “EEDAR MI6 2008”, we find another bunch of press articles relaying the news that “demos can hurt sales”. It's cool because now we can retrieve the same graph but for PS3 game sales, and it already tells a different story: the differences between the groups aren't as impressive.

But more importantly, now that we know that the data was presented in early 2008, and therefore regroups games released up to the end of 2007.

Firstly, in 2007, Youtube was starting out: the most popular game trailers only had a few hundred thousand views, which is negligible compared to today. People watched trailers in other places (such as gaming websites), but it was a lot less prevalent than today. Trailers were reserved for a certain elite who could afford them with their high marketing budget (for TV & store ads, to show to journalists at E3, etc.) And as Geoff Atkin explains himself in his conference, a bigger marketing budget correlates with higher sales.

Secondly, 2007 was the beginning of the generation, and relatively few games were available then. According to Wikipedia's list of all Xbox 360 games, there were more than 200 games released physically in North America (NPD didn't track digital sales at the time). This is a small sample size, even more so when we split games again into four categories: a few outliers can have a massive impact on the data.

One of these outliers is Halo 3. That game was such a hit when it came out that it reportedly sold close to 4 million in its first month on US soil alone. Thanks to its extraordinary sales, Halo 3 can make the whole group of games “with a trailer but no demo” looks like they sell 2.5x better on average. In reality, for all we know, there could be 25 games that sold 100k copies (the same as the “trailer + demo” cluster) and adding Halo to the group would be enough to raise the average to 250k.

And Halo 3 wasn't the only candidate susceptible to influence sales: Call of Duty 4 or Guitar Hero were also big hits of that year which didn't have a demo. So what happens in years without such hits? Or, more interestingly: what could the story be if those hits had a demo? According to the Xbox 360 marketplace, there are 90 demos for games released pre-2008; therefore, if Halo 3 had one, it would be enough to raise the average of this entire group by 45k copies.

You see the problem with using data to tell a story: what you want to say (" games without a demo sell better because people have no choice”) may absolutely not be the story behind the numbers (" a few hits without demos skewed the whole group's average”).

The Difference Between Correlation and Causation

I'd speculate that EEDAR had stopped mentioning these findings by 2013 precisely because the data was five years old, and they couldn't replicate the results again. But there's a bigger problem still: the issue with mixing correlation & causation.

They were partly aware of the issue since an article from 2008 was written to point out the logic's flaws, with a reader mentioning that big games with high anticipation & high marketing don't need demos anyway. EEDAR's co-founders answered "Causation is a hard thing. There are so many variables in the industry. What works now may not work later on causing something. What we want to do is point out really strong correlations".

And the quote continues "If you have a game with a $100 million marketing budget, you probably don't need to release a demo. What our evidence showed us was, if you only have an average game and an average marketing budget, putting out a demo that's bad is going to hurt you even more than [if it] were a triple A title [where] marketing might make up for it.”

It's an entirely different story!

Did the misinformation influence the industry, though? Maybe, but since demos didn't disappear overnight, that should probably have raised questions in people's minds: if demos had a negative impact on sales, why would Activision, Warner, Capcom, EA, Sony & others continue to make some? Not to mention the elephant in the room: what about PC, where players often use demos to check if the game runs on their computers?

We must not forget that creating a solid demo can be arduous since you have little playtime to prove the game's quality, and you can't afford to waste it on tutorials & story exposition. Furthermore, with games getting more open, complex, and online-based, demos became harder to produce in the 2010s, so if it wasn't going to be a decisive marketing factor, that time was better invested in polishing & debugging, for instance. So maybe not making a demo was the right choice, even if the justification is incorrect.

Conclusion

In my game design career, I can't count the number of times some higher-ups had questionable interpretations of poorly contextualized data. It takes time to develop the reflex to take a step back & question if the causation we think we see (because we're human, we make links) is real or if we're missing the actual correlation.

Many people would love to figure out the secret to selling more games, but it's impossible: every game is unique, there are thousands of factors that go into their success/failures and without a time/multiverse machine, we can never measure the impact of each decision. So, always be suspicious of people claiming they've found the secret (especially if they don't use it for themselves).

If you liked the topic of this article, I’m sure you’d also enjoy this one:

Indeed, the idea of a story which hands over the key to its characters’ fate to the reader can be traced back to the 1930’s novel Consider the Consequences! with its 43 possible endings. The Choose Your Own Adventure series massively popularized the genre in the 80s, with more than 250 million books sold. Around the same time, computer games were exploding as well; it was only natural for them to let the player decide the fate of the avatars they play.

However, a problem arises from letting the users shape their experience: the developer has to adjust to the consequences of all possible choices, and if there are many options, it translates into an absurd amount of work. Yet if you only offer a few choices, there isn’t that much interactivity anymore, and it falls flat.

Fortunately, in the past forty years, clever designers have worked on that issue, so we have many examples of tricks & strategies to leverage branching narratives without exploding costs. No need to be a writer to learn something from them: choices are everywhere in games, from quests to progression systems, and we can all use a bit of inspiration.

Humans vs Exponential

A tree of possibilities, where each fork leads to several branches and, again, is exponential growth, something we humans have a hard time intuitively grasping because it’s unlike any law of the observable nature (including for actual trees).

Let me take a quick example: suppose a new virus comes, and an infected person transmits the virus to one another each day. On the first day, there’s one, 2nd there are two, then 4, 8, etc. How many days until everybody in a population of 10 million is infected? Only 24 days. Exponentials grow fast; that’s their very definition.

But here’s where people are usually tricked by their intuition: how many days until just 5% are infected among the population? The answer is 20. While 80% of the time is elapsed, the problem is inexistent for 95% of people. Then suddenly, it becomes a massive issue for everyone.

Why does it matter for game development? The work required to keep up with exponential story possibilities isn’t linear. While it feels utterly manageable initially, developers might be a few choices away from complete overload.

You may say that every option doesn’t have to be unique and that making adaptations might not be a vast deal of work, but it adds up quickly. Writing becomes more complex, debugging & playtesting as well, not to mention the costs of recording & localization.

Even if you can make it work production-wise, offering too many choices can be counterproductive since it will be harder for players to remember what they decided and understand how it influences their narrative. Despite having more choices, they feel less involved. That’s the paradox.

So, to address those issues, you need to be smarter about the quality of your choices rather than focus on the quantity: it takes a great deal of creativity to do so, and you have essentially two options (omg, is this a branching article?): sheep herding & embracing the tree.

Sheep Herding

The first solution to trim the exponential tree is to consider your player as a sheep and yourself (the developer) as the herding dog: the sheep might wander off, but eventually, you ensure they don’t deviate too much and keep going towards the destination you have in mind.

The king of such games (before its complete demise, anyway) was Telltale Games, which reached massive critical & commercial acclaim with their adaptation of The Walking Dead into an episodic choose-your-adventure story. Their recipe has two secret ingredients.

First, they figured they could lie about the importance of choices. One of their key “mechanic” was to display “This person will remember this” in the HUD after you made a decision, even if that wasn’t going to influence anything. Since the interface is the privileged source of info from developers to players, there was no reason to doubt it. This idea was brilliant since it helped highlight crucial moments in the scenario and trigger players' imaginations who’d make connections that didn’t exist (our brain loves to do this).

This leads us to their second “breakthrough”: how they mastered the art of hard course correction. There are countless examples in their games of decisions that feel major yet eventually lead to the same outcome. There’s one example in the graph below: at the end of the fourth episode of The Walking Dead, you can choose to save a character, which seems important. You ignore that the character barely intervenes in the following scene scenes, and he’s dying a bit later anyway.

With this approach, all players will see a maximum number of scenes, optimizing production costs for a lengthy enough playtime—no need to create tons of additional scenes.

However, these corrections were sometimes noticeable (even before replaying). A gripping narrative and surprising twists can help to disguise the smoke and mirrors, but when you start figuring out the trick, it becomes less and less engaging (which explains how quickly sales dropped for Telltale Games despite the massive franchises).

Embracing the Tree

But if you want the player to see all the content, how about doing the opposite and acknowledging the tree in order to encourage the player to explore all possibilities? This design approach made the success of the other big episodic game hit, Life is Strange, and the visual novel Virtue’s Last Reward, among others, games which use time-travelling as a narrative justification to explore the consequences of choices.

Recently I also came across Cartaventura, a series of narrative card games where you explore the life of lesser-known historical figures. It functions like a choose-your-adventure novel, and due to the limited space to write on the card, the designer had to be clever to maximize the number of scenes & dialogues: let’s dig into all the best practices I observed.

Have separate areas

Each Cartaventura makes you visit several places, which fits the exploration theme and is convenient for writing. When they move, the player gets a sense of an evolving story, which easily justifies the frequent “soft reset”. In each new area, the designer introduces new characters and therefore doesn’t have to make them react according to past players’ choices.

Structure the world like a Metroidvania

To complement the point above, the games typically include several ways to reach the same place, just like shortcuts & hidden areas in Metroidvania. It’s a way of sheep herding for sure, but one that feels more rewarding for the player. One way to reinforce this approach is to ensure that the player gets something new or different in the mandatory area, depending on which path they took to get there.

Use items to track choices

As a physical game, Cartaventura can’t hide its logic in the software; the player must know which options are available & which previous choice influences his narrative. To do so, the player gets item cards after making crucial choices, a simple but genius design. Owning an item is more readable for the player, and he can make immediate connections (compared to remembering how a dialogue option influenced a relationship with characters).

Turn the tree into a puzzle game

Items are also heavily used in the series as “keys” to unlock paths, which ignites your desire to replay and discover how you could have obtained that item that opened this door. I wrote extensively about such a “door & key” design in this article.

With just a few items, each Cartaventura can turn its branching tree into a puzzle where your goal is to understand how to discover all the outcomes. The endings are hierarchized: the “bad ending” happens if you make a few mistakes, while the “good ending” can only be unlocked by collecting everything & knowing exactly how to navigate the tree. Once the player knows the narrative, he won’t re-read the cards anyway, so you may as well switch from a pure narrative experience to a strategic puzzle feeling.

If you want to try it out, I recreated the tree of the Oklahoma episode below with all its doors and keys.

Shake the initial conditions

Proposing choices to the player doesn’t mean you must offer all the options immediately. As a designer, you may know that some outcomes are more desirable than others and, therefore, would like to drive players into seeing some branches before others. In Cartaventura, you always unlock new possibilities when finishing the game for the first time, which is a great way to encourage replay with a twist.

Design many keys, not merging rails

It’s a lot of praise for such simple games (and I didn’t even mention the fantastic themes & funny choices), so I’ll balance it with a critique of the Hollywood episode, the worst, in my opinion. It has an interesting idea where you must explore all the branches to find 8 “scenario ideas” in order to unlock the best ending. However, in some instances, like the Chicago arch I recreated below, it tries hard to funnel you towards a single encounter which feels noticeably worse than having several options to collect a given item to move on. There’s a thin line between escaping fate & feeling in control of the puzzle.

Conclusion

So, which is better in the end? Sheep herding or embracing the tree? The designer in me leans towards the second, obviously, I admire when a big constraint is turned into a strength through creative design choices.

However, I also know we must always keep in mind the unexpressed needs & desires of players: some will enjoy choices when they’re a flavour on top of a conventional story but won’t like it if it turns a laidback experience into a head-scratching puzzle.

If you liked the topic of this article, I’m sure you’d also enjoy this one:

As designers, artists & creators in general, it's easier to explain our medium using concrete facts: the game is that long, it uses this and that mechanic, it has this many items. However, by explaining games for what they are, we lose sight of the most important: how they feel.

Rogue-lite is one of the most fascinating genres to me precisely because of its (too-often-eluded) emotional core. The best games in the genre stand on a thin frontier between fun and frustration, luck and strategy, familiarity and variety.

If you only focus on the big factual terms, such as procedural, perma-death or metagame, you will not understand the experience underneath. You’re at a high risk of failing at recreating the feeling of rogue-lite, even if you tick every box in the checklist.

So today, we're having a look, not just at what rogue-lites are, but also at why they work and how they create strong emotions.

The Central Symmetry

Rogue-lites may use varied core gameplay, shooting, platforming, cards or action, yet they all present the same equation to the player: kill or get killed.

I discussed the question of omnipresent health systems in a past issue of The Arcade Artificer, and the main points remain valid here: in such “me or them” situations, there's an obvious tension with two clear outcomes. You may stall out a little bit, but rogue-lites always force you to face danger, even if you don't feel prepared. No time for grinding.

In such a ‘do or die’ configuration, you have two basic needs: means of offence and means of defence. The designer's role is to give the players enough options to cover both: through the permanent mechanics (action to perform/rules of the system) and additional tools to collect while playing. For instance, a rogue-lite with top-down shooting will typically let you avoid damages by moving around but may also include armours or health pick-ups.

All twists and turns are allowed to create a diverse experience, from economic loops to secret combos. The defence could mean blocking damages, healing after the fact or simply preventing the enemy's damages (using a stun mechanic, for instance). There is also an incredible variety of ways to deal damages: combo, over time, one-time, charged attack. RGD principles can help to find all sorts of meaningful variations, but at the end of the day, the player always intuitively translates it all into damages gain/loss.

To be more accurate, the player is typically focused on time: the damages per second/per turn they can inflict and sustain. Indeed, in most cases, you have a permanent option available to evade damages; it's just that the margin of error is thinner as the difficulty ramps up.

We all end up making mistakes or getting unlucky, so when the combat drags out for too long, we start feeling a sense of dread and impending doom. Mistakes, just like good decisions, compound; small fails here and there will ruin your run eventually.

Run Completeness

All of that is great, but it works for all sorts of survival and multiplayer games too. What sets rogue-lites apart is another critical component: the sense of completeness.

Indeed, a rogue-lite is structured in “runs”, a series of combat encounters & choices with an end and, therefore, a purpose. You want to finish it. The final boss is like the graduation exam in the “kill before getting killed” class. If you can't complete this challenge, you need to sharpen your strategy and come back stronger.

This sense of completeness is integral to the satisfaction a rogue-lite provides. They start simple, you're weak, but as you survive increasingly difficult challenges, you grow so that at the end, you're at your peak. Rogue-lites deliver their whole (simple) story in a single session; conveying this feeling is more important than the factual run duration.

A run-based structure gives the genre another perk: general knowledge aside, no need to remember what came before or to care about what comes after. You feel present and focused—the choices you make count now in this run.

Input Randomness

And so, we get to another key ingredient of rogue-lite that is often misunderstood: the procedural generation. The procedural isn't about randomizing stuff so it's less boring when replaying, it's about renewing the strategic feeling. Any fixed game, even a difficult die & retry, will eventually shift its mental challenge from « thinking » to «memorizing». Once you've figured out the strategy, you can just focus on the execution (and it can be a great deal of fun, like in Souls, it's just different).

Rogue-lites want to challenge the brain again and again, therefore they have to present the player with some input randomness (a unique situation) and let them figure out how to deal with it. Which rewards to specialize in, which risks to take, how to explore the dungeon, etc.

You may ask: «isn't random the enemy of strategy? If good or bad luck can destroy your strategy, how is it more fun? »

To make an analogy, random in rogue-lite is akin to poker: you get dealt a hand with a certain power level, influencing your chances, but you can still make the situation work regardless of your initial luck (through bluff or simply by minimizing your loss). Like in poker, the randomness of rogue-lites is diluted with the number of occurrences (which reduces variance): in most cases, you'll have a bit of both good & bad luck, your challenge is to succeed in all contexts.

The procedural generation generates a unique puzzle each time, sure, but again it's not as much about randomness as about the rules. Rules are the appealing part of the rogue-lite variety: you jump in the session quickly and learn by doing. Things such as enemies patterns or more obscures facts such as how secret rooms appear in The Binding of Isaac —this is a knowledge you acquire to help you survive and typically compliment the strategic aspect: you don't “solve” the game, you learn patterns to get better at strategizing.

Having more info about the inner working of the game helps you make more informed choices and use your resources efficiently. Humans feel content when they get better at executing a task.

Fun Again & Again

However, designers had to solve a last problem to truly enable the potential of rogue-lites: alleviating the pain of losing. Losing sucks, it always sucks. Even when it's “part of the game”, and you learned something new, it feels terrible.

The solution has a name: meta-progression. Changing just one letter from rogue-like to rogue-lite, designers leveraged modern design techniques to do two things: materialize the importance of each run and incite to try again. When they die, games let the player keep a currency: even if it's 5% of all collected items, it feels better than zero. Using this currency to then marginally improve the starting character (or unlock another one) may be symbolic, but it matters.

Games also frequently use the meta-progression to unlock new content & mechanics, which both encourages to try another run, and helps to space out the strategic aspect. Indeed, any deep system with many rules will feel chaotic and indigestible, but by introducing notions one after the other, the player can absorb a surprising amount of knowledge.

There's a tricky question to address here regarding the game’s fairness: is it genuinely possible to beat the game on the first try, or is the meta-progression a sneaky way to upgrade your raw power? Some players prefer to feel tangible progression, while others like it when their skill has to grow. The most important, however, is that the player doesn't feel like meta-progression was the key factor: you don't want them to lose the sense of presence and blame impossible starting conditions, or worse, to feel like a win is only achievable long after grinding.

Wrap-Up

Some say that asking good questions is half of the design work. I think it's even more than that. Questioning our feelings when we play is integral to learning how to design. If you only read a review that says “frustrating”, what is it about? At the system too opaque to strategize properly? Or the choices you make don't matter? Unlike technical jargon, these are the sort of things that few people discuss excplicitly, so you only have one option to figure them out: listen to your emotions.

If you liked the topic of this article, I’m sure you’d also enjoy this one:

It’s normal, adults have busy lives, and the idea of grinding a game for several dozen hours for months sounds exhausting. Despite this “popular opinion”, long games still have a massive advantage in our industry.

Hardcore gamers, those who spend the most, typically weigh in their purchases according to the value they perceive, and the ‘cost per hour’ is one of their most important criteria.

It’s nothing new: since the dawn of gaming, developers have had to develop strategies to lengthen the experience they produce. On the arcade, it was the absurd difficulty, then came optional objectives, completionism, multiplayer modes, grinding and many more.

In production, designers often have the hard job of finding ways to increase the game’s length without exploding the scope. With more mechanics, more game loops, the complexity increases exponentially, so you have to be smart about your approach.

How can we reuse mechanics to provide additional enjoyment? Where is the limit between meaningful variety and tedious repetitiveness? Where to invest efforts so a few new elements reinvigorate the previous content?

All these questions (and more) can be approached from a methodical perspective, and it has a name: rational game design.

What is Rational Game Design?

Rational game design has a vast and unclear definition, the industry is still young, and we don’t have a unified vocabulary yet here, so I’ll explain the concept the way it was taught to me ten years ago)

The basic idea behind the rational game design is to break down the rules & parameters of a mechanic to understand how each of these atomic components influences the final experience.

It’s a cold & clinical look at the software, which voluntarily avoids artistic or emotional opinions. Using this method helps to apprehend abstract concepts such as the variety or the difficulty progression. We could say that the rational game design is the flipside of the heuristic decision-making we intuitively resort to (”well, it seems a bit hard, maybe tone it down a little”).

Suppose, for instance, you’re a game designer in charge of proposing a diverse cast of enemies for a shooter game. The AI system is in place, with just placeholders, and now the directors want to see varied opponents. With the RGD method, you may start to identify their relevant characteristics: size, speed, resistance, etc. Of course, which parameters are interesting depends on your specific game mechanics: if the player can select from several ammo types, enemies could have weaknesses & resistances to the different types.

With those parameters in mind, you could propose a tank enemy which is huge & slow (easy target) but with a lot of HP and resistant to fire ammo. And also a rusher who’s weaker but moves fast. Those are two distinctive archetypes, challenging players differently yet using on the same system.

Enemies don’t have to be all balanced, of course, especially towards the end of the game: why not combine all difficult parameters at once? For example, we could propose one that’s very small and very slow but deals incredible damages: easy-to-neutralize taken alone but deadly in chaotic situations.

Of course, AI archetypes are indissociable from the levels you put them in, but you can apply the same method to them too (that’s rational level design): is the area narrow or open, how many covers does it have, how many enemies are combined, etc. With such an approach, you can ensure that the content you create is noticeably different from the previous and that difficulty & complexity rise over time.

Taking time to break down the parameters also allows you to spot unexplored design spaces. By flipping a standard rule (’enemies die when their health reaches zero’), you can often create unique mechanics for minimal cost (’one archetype explodes on death’ or even better ‘, one enemy heals its peers when it dies’). Using a rational design viewpoint helps to make games more cohesive and achieve more variety with fewer mechanics.

Tile Laying Core Gameplay

Now that we have the theory, let’s see the rational design in action in a simple yet highly varied game invented by no less than a (german) doctor in mathematics: MyCity from Reiner Knizia.

MyCity is a city-builder where players position buildings one after the other on their boards until they run out of place. There are only a handful of rules to respect: each building must touch one placed previously, and buildings can’t go across the river or be placed (even partially) on forests or mountains.

The goal is to have the most points at the end. You must fit buildings together in tight spaces to avoid the penalties (-1 per empty green square, or -2 if it has rocks). If you can leave the trees uncovered, they award a +2 bonus. During the game, skipping a building you can’t (or don’t want to) place, also gives a -1 penalty.

Since every player has the same set of pieces, they can plan ahead, but the random order can force them to re-evaluate strategies. The board is constrained and the odd shapes don’t fit all with one another, it’s an easy-to-understand challenge mixing strategy with a luck.

What elevates this not-too-original concept though (Blokus came 20 years before), is the campaign mode.

Indeed, in the box, you’ll find 8 chapters envelopes with new components & rules variants which tell you how each of the 24 episodes (=games) plays out. This type of board game is called “legacy”, a popular subgenre where you play several times and some of your choices have consequence on future games.

Legacy games typically use stickers to permanently change the boards & pieces. In MyCity, for instance, the winner of an episode will get a malus rock sticker while others get a bonus tree, an easy way to re-balance chances for future games and keep it competitive & fun for everyone.

So, as you chain games and open envelopes, you receive new stickers (and sometimes other bits such as cards or pieces), which add new mechanics & scoring rules. Churches, mines, rails: the way MyCity can effortlessly introduce novelty yet maintain its approachable gameplay is admirable.

It doesn’t seem like it, but being able to hook you to play the same simple game 24 times is quite the feat.

5 Strategies to Renew the Fun with RGD

In the final part, I’ll examine some of the RGD principles that Reiner Knizia used to add variety to its core system. I’m not sure he even calls it that, but judging by his background, I’m confident he has a methodical approach to game design.

#1 - Make promises (and keep them)

It sounds simple, yet it’s often forgotten: what pushes players to get invested in the whole campaign is the curiosity. What’s inside the next envelope, what do those symbols on the board means?

Having crazy varied mechanics isn’t going to help if players don’t stick with you, so tease, tease, tease! Don’t make false promises, just find ways to expose quikcly enough that there are lots of possibilities to come.

#2 - Expand & exploit the entire design space

As I explained, RGD is all about identifying relevant parameters of a mechanic to create variety, but at some point, you’ll reach limits regardless. The temptation in that case may be to invent another one but sometimes, introducing just one new parameter is enough.

In the game, the place-a-tile mechanic is enough to have scoring rules for:

• Covering or not a square

• Having a square surrounded on all four sides

• Touching a side of the board

• Touching the last-placed tile

Which is ok at first but needs to be deeper in the long term. So, Reiner Knizia separated the buildings into three colours (with the same shapes in each), and suddenly, this simple parameter opens a much greater variety potential:

• Groups of the same colour

• Surrounding a building with all three colours

• Touching a tree with a specific colour

• Biggest group overlapping a particular element on the board

You don’t need much to drastically increase the variety as long as you have good combinatory.

#3 - Change the context

Just like changing the level design can significantly influence how an AI is perceived, deviating from the usual context can greatly help the sense of variety.

In MyCity, most of the scoring depends on how you deal with your personal board (the norm), it has a bunch of multiplayer rules where the first to do X, gets a bonus over the others.

Another seemingly simple, yet exciting change, was when the game introduces a reward for surrounding a square with four red buildings that is an immediate tree sticker rather than points: technically, it’s similar to a +1 bonus (so not outstanding, given the efforts required), but it feels even better!

#4 - Shake the basic rules

To expand the previous point, another strategy used by Knizia to give a feeling of complete renewal is to make exceptions to the basic universal rules you learned at the very beginning: it immediately stands out by contrast.

For instance, at first, you are forbidden to place on the forest, but once you get the sawmill, it suddenly becomes allowed. You can skip all buildings except churches (which you start to dread seriously). Punctual exotic modifiers can have a fantastic effect.

#5 - Add (and remove) systems over time

And finally, one of the big lessons from MyCity comes from the organization of all those rules. It’s not nearly enough to generate tons of ideas; it’s also crucial to introduce them gradually, associate rules that work together or, on the contrary, eliminate some before they get too stale. RGD is a tool to maximize the sense of variety & progression, not a way to generate artificial complexity.

You can see on the graph below how mechanics are introduced and then disappear to make room for others. Subtractive design is sometimes scary (fear of losing depth), but it’s a powerful tool to make things feel different.

Wrap Up

As always when exploring such a complex topic in a single article, I feel like a drop way too many information at once. So, if you had to keep a single takeaway: rational game design is a powerful lens to maximize the potential of systems and ensure varied & progressive content.

It might be tempting to see it as a cheap way to create artificial variations (and it can be), but conversely, it’s an excellent method to design cohesive experiences which explore a system or a mechanic entirely.

If Nintendo approaches design from this perspective, I’m sure we can learn a thing or two.

If you liked the topic of this article, I’m sure you’d also enjoy this one:

The Arcade Artificer

Quest Design Lessons from Monkey Island

Fetch quests deserve their bad reputation. Doing NPC's deliveries was never the peak of entertainment and boredom awaited at every corner. Yet, in essence, a point & click adventure is just a series of fetch quests that has their players look for items and give it to the correct character…

Read more

3 years ago · 4 likes · JB Oger

Since the Steam submission process became open, more games are released each year. Everyone has a chance, as long as they can compete with all the other new games to come out and also those from past years which still receive updates & regular discounts. Naturally, large & experienced companies have a massive edge in such a hit-driven market, however, in the arsenal of indies, there is one weapon to avoid standing pale in comparison: lower prices.

Pricing a game is a tricky prospect. On digital platforms, there are no costs of goods (fabrication, transport, storage, etc.) which can normally be a good basis for figuring out a price that leaves a decent profit margin. Unlike physical products, it's financially equivalent to selling 1000 digital games at 2€ or 100 at 20€. The problem is that you have no idea of how price & volume correlate.

So, most indies try to position themselves on the market through comparison with similar games. If you're not too confident in the quality of your game, you may as well sell it for a few bucks: players won't be as demanding, and you'll make up for it in volume (the hard discount strategy). It's precisely what worked for today's subject of examination, Vampire Survivors.

Or did it? As I said, in the indie scene, there is a debate. Was the game's 2.69€ price the main reason for its success or, on the contrary, just an accelerator to propel a strong game already?

Many people believe Vampire Survivors is just another low-quality pixel art game with unimpressive (but efficient) gameplay. You know I'm a contrarian: I think Vampire Survivors has intrinsic merits, and its lucky virality couldn't just have happened to any other very-cheap game.

From Bullet Hell to Bullet Heaven

First, we have to understand the core of Vampire Survivors: which genre does it belong to? People seem to settle on the term “action roguelike” (on Steam page tags, at least), which refers to its rogue-lite-inspired progression structure, but I prefer the elegant “bullet heaven”, as it explains the core game loop better.

A “bullet heaven”, by definition, is a reversed “bullet hell” (also sometimes called shoot 'em up). In both you play an auto-shooting turret and have to move around to avoid bullets & kill enemies. The difference between “hell” & “heaven” lies in the balance of power: in a bullet heIl, you’re a fragile hero narrowly evading complex projectile patterns, whereas in a bullet heaven, you are a demi-god spitting out millions on hordes of enemies.

There lies the first “secret “of Vampire Survivors already: the core challenge is crystal clear, and you can understand the stakes from a single screenshot. You don’t need more to feel the tension and figure the objective.

There's a character with a health bar standing out in the centre of the screen from hundreds of enemies damaged by visually impressive attack effects. All the codes used are simple & well-known by gamers (damage numbers, bats as enemies, the timer, etc.), so already just with its core loop Vampire Survivors is able to communicate its promise efficiently (a topic I covered in the very first issue of The Arcade Artificer).

A bullet heaven's goal is to survive as enemies swarm you, but there is also a fundamental difference in the difficulty of evolution. The obstacles & the type of challenges don't change much throughout the experience (enemies move in a straight line towards you), but your avatar will shoot so many more projectiles (that deal higher damages).

Instead of a “David vs Goliath” scenario where you feel proud for outsmarting a tricky boss with a giant arsenal, you are the boss crushing thousands of ants.

And it's another key to understanding the Vampire Survivors’ appeal: it's a casual game at heart, where you don’t need to be mechanically skilled to feel the thrill of power. Actually, skill doesn't matter all that much here, after all, you can only move and you have few parameters to account for (avoid enemies, collect powerups & dropped XP pills, visit map locations). Your capacity to survive resides a lot more in the upgrades you chose, an other area where the game shines.

Power Race In The Casino

Vampire Survivors’ player progression is as simple as selecting the best rewards the game randomly proposes to you (this is the rogue-lite part). The bar on top of your screen is an XP meter, and every time you fill it, you get a new item to increase your stats (the usual RPG ones, damages, speed, rate of fire, area of effect size, etc.).

There are only about 30 items to collect in the game, most of which need to be unlocked (as we'll see in the next part), but since you can only hold six weapons & 6 passives. Each item has enough horizontal differentiation to colour your run: you can somewhat specialize your character with different strengths & weaknesses each time, so it doesn't feel too repetitive.

The game shines a lot brighter on vertical progression, however: rather than unlocking just those 12 items, each starts somewhat weak and has many upgrade tiers (8 for weapons, 5 for passives). The intention behind this design choice is easy to understand: rather than a reward every 2 min 30 (12 items in 30 minutes), you can have one every 23 seconds!

A trickling stat progression helps to maintain the pacing better than a step-like difficulty curve where players get boosts of power infrequently. The being said, smoothing out the progression curve doesn’t exclude the possibility of major bumbs (you can see several in the graph above), as they also feel rewarding for the player. The game’s difficulty inevitably catch up afterwards to maintain the pressure.

There's another slight advantage to such balancing: the game is more frequently paused while you're picking an upgrade, which is essential to keep it from becoming exhausting through the constant attention (and visual strain) it requires.

That being said, the upgrades aren't anything exciting (you don't get to choose how you upgrade your weapon, it's predetermined), and choices aren't significant for the most part (since the options you decline will probably show up again in the next few minutes anyway). The strategic aspect comes from selecting the right combo items to maximize your damages & survivability, upgrading them is more on auto-pilot.

Why does it work, then? Well, I'm not surprising anyone if I say that the loot mechanic has proven its power to engage players. The designer of Vampire Survivors, Lucas Galantes knows it better, too: he explicitly noted in interviews that his work on online casino games had inspired him. The chest opening animation, for instance, received a lot of polish (compared to the rest of the game): it feels like a slot machine, and it uses similar techniques to get you addicted too!

Indeed, chests have a considerable variance: you can loot from 1 to 5 items and anything from 100 to 1000 gold coins. As you can imagine, better outcomes are very rare, but the game uses a nasty trick to hook you anyway: the first six chests you get are hardcoded with respectively 1-1-3-1-5 items inside. This sets your expectations high before drastically reducing the odds.

More Structural Tricks

At this point, you're starting to see how Vampire Survivors is addictive by design, not by chance. However, this ‘casino fever’ coupled with arcadey gameplay could still quickly fall on the boring side if it wasn't for the following aspects.

First, Vampire Survivors has a simple meta-progression loop, the « lite » aspect of rogue-lite; you don't restart from zero-zero each time. During the run, you've collected gold with no particular use on the spot, but unlike the rest, you get to keep it at the end and spend it on PowerUp in the main menu to increase your starting stats (again the speed, damages, rate of fire, etc.)

The increments are tiny individually and again separated into many little bonuses, so it doesn’t feel grindy, but it’s also a sneaky way to upgrade your baseline. All cumulated powerups make a massive difference in raw stats: about 2.5x the damages and at least x2 health. Yet, you do not feel as such and believe your margin of progression has more to do with your skill & strategy.

Such a meta system, of course, has the advantage of eliminating « zero progress » runs: even if you didn't make it as far as the previous, you do not feel as if you've wasted time (or even better, you want to retry immediately with your new perks unlocked).

The other clever aspect of Vampire Survivors is its immense achievements list, giving you constant side objectives while playing the same content repeatedly. Nothing complex once again, just a reason to taste a bit of all the content: survive with each character, level up the weapons, visit maps locations.

The rewards are plenty, with new items, new characters, new locations. Nothing ground-breaking, but it just works to give you the illusion of plenty. In a game that doesn't set expectations high, unlocking alternate versions of the five levels or a bestiary menu feels like a bonus. And that might be the last secret of Vampire Survivors: it always goes the extra mile to put a cherry on the top, and the small attention leaves you contempt.

Conclusion

So, was Vampire Survivors lucky to go viral? I still believe so.

Did its creator design efficient mechanics to create a compelling experience? Certainly too.

Would it be as enjoyable if it costed 20€ and had better art? Of course.

People forget that it wasn't the low price that made people stay and enjoy the experience enough to tell their friends “about this tiny game that is much more fun than it looks”. Vampire Survivors isn't a gimmick, it's just a great (small) game players enjoy. What else does it need to be?

If you liked the topic of this article, I’m sure you’d also enjoy this one:

The Arcade Artificer

The Counter-Strike Clockwork

Why is Counter-Strike still relevant to this day? In today’s market, big publishers try hard to maintain a community by feeding their flagship titles with regular content updates. Valve, on the other hand, barely does it, and yet Counter-Strike is still in the top-played rankings, 20 years after its initial release…

Read more

4 years ago · 7 likes · 5 comments · JB Oger

Unlike other disciplines, the story creation process isn’t well known, probably because there are many notions & methods you may use to write an engaging narrative. The typical game review stays on surface, giving a summary of the plot & throwing adjectives at the characters (what are we supposed to understand when they say someone is “memorable” or “forgettable”?).

As a game designer, I consider myself a scientific person: I have a rational approach to systems, components & rules and always attempt to analyse how & why they work (or don’t) within a given game. The Last of Us became an instant personal favourite since its release, yet I struggled to really nail what made this game so great beyond the “obvious reasons” (outstanding actors, unusual themes, cinematographic realisation, etc.)

I’m far from perfectly understanding the game’s narration and even further from writing one myself, but I’ve now got a clearer picture of the intrinsic qualities of The Last of Us. It all started with a book...

Robert McKee’s Story formula

Throughout this article, I’ll repeatedly refer to the same book to explain critical concepts: Story by Robert McKee. I didn’t pick it randomly: Neil Druckman, creative director & writer of The Last of Us, has explained on several occasions how influential this particular storytelling book was for him & the Naughty Dog studio.

Breaking down a story into individual components can help us understand what makes them engaging. However, remember there’s no magic here: even with a model, it takes a lot of creativity & iterations to reach a high-quality story. It worked for Naughty Dog, so let’s learn from the best!

In Robert’s McKee model (simplified for this article), each important character has desires and values that shape their perception of the world.

Throughout the plot, we see them in scenes (40-60 in an average movie) where the different values are opposed to create conflict.

Several scenes are stitched together to form sequences with higher stakes & climaxes. Ultimately, the story can be seen as a series of events from the characters’ lives that are strategically selected & organised by the author to express a point of view & deliver emotions.

That may not sound particularly helpful or innovative so far, but let’s dig into the concrete example of The Last of Us.

Dimensional Characters: Desires & Values

The Last of Us main protagonist is Joel and he’s introduced as a protective father during the prologue, set on outbreak day, then re-introduced twenty years later as a stone-heart smuggler in Boston’s quarantine zone. This two-part exposition is crucial to present to the audience Joel’s dimensions, which, to McKee, are the character’s contradictions that make them nuanced.

Protagonists are driven both by a conscious & an unconscious desire, which are typically conflicting. Joel’s conscious desire is to survive and protect his tribe at all costs (as seen with Sarah, then with Tess), but his unconscious desire is to experience a parent-daughter relationship again.

At first, he’s sticking to his concrete goal “deliver the girl to get my smuggled weapons back”, but as he fails to accomplish this task, he gets deeper and deeper towards the place he never wanted to return: the fear of losing the loved one.

The character’s motivations & internal struggles are explicit right from the start, we develop empathy without a word. The prologue could never have been a flashback after the fact, (not that Naughty Dog dislikes them, it worked to tell Ellie’s story), but because we need to understand intuitively where the story goes. After seeing Joel capable of loving his daughter, we wonder if/when he’ll return to that point.

On a fundamental level, Robert McKee also explains that the story’s emotions, or as he calls them, the values, are dual (freedom & slavery, life & death, hope & despair). The character’s arch is therefore a slow change from one to the other, as their view of the world gets challenged by the events & other characters.

Joel’s main characteristic is indifference when the story begins (yes, The Last of Us inciting event is Joel meeting Ellie, not the outbreak or the death of Sarah)—the opposite of love.

A classic mistake in storytelling, and particularly in video games, that Naughty Dog avoided there, is to use values that are contrary instead of contradictory. Contrary values such as “love & hate” or “good & evil” accept in-betweens compromises, and changing from one to the other feel cheap & forced (you don’t go easily from ‘hate’ to ‘love’, but you may become ‘acceptant’)

Another mistake would be to have a single value driving the character: each choice becomes so obvious that it’ll be difficult to generate tension & drama. Imagine if Joel was egoistic rather than a protector: he’d abandon Ellie on the first occasion, and the story wouldn’t work. Instead, we have a nuanced character who faces difficult choices and whose values oppose the other characters. His cynism contrasts with Ellie’s innocence and hope.

Each character in the story has a different viewpoint, with no clear right or wrong, an essential setup to generate the “currency” of a good story: conflict.

Layers of Conflicts & Intense Scenes

Indeed, all those rich & subtle characters would go to waste if they didn’t experience situations which reveal their values to the audience and challenge them. We don’t want the writer to simply tell us the struggle through dialogue, we need to see it in action.

Those “forces of antagonism”, as McKee calls them, can pressure the protagonist on several levels: the inner conflict (that we’ve discussed already in Joel’s case), the personal conflicts between characters and extra-personal conflicts (against the threat of nature).

In video games, the latest tends to receive a lot more attention since it provides obstacles for the character, and The Last of Us is no exception with its violent post-apocalypse setting. The game doesn’t stop there, however, as it expertly weaves the layers of conflicts together: Ellie being immune & a potential cure for humanity plays a significant role in the narrative but it’s also a tool to push the tension between characters, and in turn, affect Joel’s inner conflict.

The events (or scenes) have increasing stakes that shake the balance of the protagonist, who struggles to find a new equilibrium. Change is slowly coming; thus, the narrative tension remains high. What keeps us riveted to our seats is the fate of the key characters: will they finally realise what we all guessed? Can they overcome their inner demons?

Joel, as the protagonist, undergoes the most dramatic evolution throughout the journey, and we notice a clever pattern from Naughty Dog: a dangerous (playable) situation leads to a climactic scene (not always playable) such as a QTE or a tense dialogue, then a pause. At this moment, the incertitude is still high, the suspense is unbearable, until finally we have a resolution with a calmer discussion & Joel (secretly) re-evaluating its choices.

Specific game events are always the catalyst for the character’s change: their impact is increased by holding on to the tension just a little longer. Again, I do not deny the quality of the art, the acting & realisation, but it’d be for nothing without the expertise in scene construction.

Structuring the Plot: Expectations & Surprises

We’re slowly getting to the biggest chunks of the story: sequences, or the subtle art of arranging all your nice characters, scenes & conflicts in a coherent order that maximises emotional impact. In The Last of Us, there are 11 sequences (or slightly less, depending on how we count): after the introduction, all of them follow the same structure of a mini story-within-the-story.

We first have an initial situation, which establishes the baseline for the characters: how are they feeling and interacting with one another? They have an objective that sounds simple enough, but quickly, there is a big event that raises the tension and changes the paradigm. After that, several complications may happen until a resolution is found, putting Joel & Ellie back on track.

Each sequence puts its twist to avoid the feeling of an “episodic formula”. For instance, the Pittsburgh sequence is longer to let the game breathe and slowly builds the characters scene after scene. The penultimate sequence in Salt Lake City starts particularly slow as if the game didn’t want to reach its conclusion, mirroring Joel’s feelings.

All sequences have the same goal of providing us with meaningful variations on the themes. While the main story only loosely connects the dots, most of the character’s evolution comes from these side story arches: Tess, Bill & Henry each depict an aspect of this violent world and serve as a cautionary tale for Joel, who doesn’t want to suffer the same fate. The arrival & departure of side characters provide major turning points which shake the status quo frequently and keep the story fascinating in the long term. I could explain how each sequence works, but you get the point.

Another McKee lesson I’d like to touch on is the importance of respecting the audience by giving them what they want, but not the way they think. When we start the game, we know that Joel & Ellie will end up bonding; there is no suspense there: what hooks us is how their relationship unfolds, the surprising low & high moments that stick with us. And Naughty Dog knows that: they purposedly lied in pre-release interviews to keep the secret on playing Ellie or the giraffe moment. They are masters of their craft.

Final Words

If Robert McKee’s work influenced Naughty Dog a lot, I’d also like to thank Neil Druckman & the team for openly discussing their craft. Video game creators are (arguably) often scared of talking about the behind-the-scenes. Yet, for The Last of Us in particular, we have incredible books & podcasts that give great insights on how to conceive a story that works on an emotional level.

Does peeking behind the curtain kill the magic? For me, and probably for you if you read this article, it never will: understanding all the work that goes into creating these experiences inspires me and the talent behind the blockbuster’s money-printing machine.

If you liked the topic of this article, I’m sure you’d also enjoy this one:

The Arcade Artificer

Quest Design Lessons from Monkey Island

Fetch quests deserve their bad reputation. Doing NPC's deliveries was never the peak of entertainment and boredom awaited at every corner. Yet, in essence, a point & click adventure is just a series of fetch quests that has their players look for items and give it to the correct character…

Read more

3 years ago · 4 likes · JB Oger

Most character-based games use an arbitrary points system to classify the varied type of injury & heal. Equip a new armour to increase your HP limit by 10, lose 35 when falling from a cliff, eat a loaf of bread to restore 50. It doesn’t make sense, and yet it’s everywhere; we’re accustomed to it.

While the past decades of gaming have brought many innovations destined to increase immersion & recreate more accurately our world, it feels odd that we don’t question the relevance of this mechanic anymore. Everybody (myself included) accepts this feature, and nothing seems to be able to replace it.

So, why is the health mechanic universal and can nobody invent something else?

All Games Are Immersive

I love to begin my analysis with a (too) broad statement, so here’s one: 95% of games attempt to be immersive.

When we exclude experimental art projects and look at commercial game genres, one could define video games as “pieces of software whose primary goal is to deliver quality entertainment by allowing players to escape reality.” Games, by definition, create a separate reality driven by rules that are necessarily distinct from ours. The game rules give a new cohesive meaning to the visible items, the displayed numbers & the pixel on our screens.

Virtually every human on earth is also a player (at least as a child), with unique tastes regarding the genre & types of experience they like to engage with. In particular, each person may appreciate different levels of abstraction & distance from the reality they know. Still, it’s fair to state that most games help the player get immersed in a new “reality” by relying on existing knowledge & leveraging concepts they can immediately grasp. Even chest.

At a basic level for instance, most video game genres involve controlling a single avatar who can interact with the play area: it makes immersion easier, because that’s also how we, humans, experience life. Our world is full of rules we grasp intuitively, such as the law of gravity or that fire burns things.

Simulating those rules is a way to blend the frontier between game & reality, thus facilitating our acceptance of the unique twists the game has to offer. If everything is new, the step is too high, and we have difficulty approaching the experience.

However, our world is complex, so the difficult task of game designers is to simplify the real-life mechanisms while keeping their essence. Over the years, we stumbled upon pretty good approximations of our reality. For example, a stamina bar is a decent model representing how humans can sprint short distances before they need to jog/walk to rest. Hunger is another example I like: it’s ‘realistic’ that the sentiment of fullness decreases over time until the lack of food feels painful and has a physical impact.

Meanwhile, health feels poorly represented when it’s often just a simple pseudo-economic system where different actions cause your “health bar” to go up or down. Why do you lose the same arbitrary amount from a gunshot as an explosion? How does eating food heal your burns? Looking at the health mechanic through an immersion lens can’t explain the standardized form it usually takes.

Who Needs Health?

If we observe how health works in real life, and I don’t mean after earning a medical degree, we may create various smaller mechanics to represent different bodily functions. This is the approach taken by the popular extraction shooter Escape from Tarkov with seven separate meters whose status can influence movement speed or vision.

This is a more complicated version of the traditional health bar, yet still ends up centralizing a variety of inputs (damages) & outputs (maluses) with its own logic of states & effects. Ultimately, though, the system has the same goal as simpler ones: tying all the “negative” actions to a common defeat condition, because it helps players make sense of the rules.

In a way, directing all the attention to an abstract “health meter” makes as much sense as unifying the varied “positive” actions in a single scoring system. Losing your avatar is the ultimate punishment a game can inflict on its player since you literally can’t experience the entertainment anymore.

No need for heavy consequences, sometimes, restarting a couple of seconds earlier is enough of a threat to push you to engage with the features designed to help you survive (such as a cover system) and generate tension.

Even games who seek maximum immersion have to forfeit a certain degree of freedom to players (or they’d be movies), hence the need to have mechanics to communicate boundaries & re-create intuitive logic. Rather than putting an invisible wall, I’m going to let you jump out of the window, but you’ll lose health because gravity hurts.

Health takes this form in games because its main role is to communicate limits efficiently, and the infinite complexity of -actual- human health isn’t a suitable candidate.

The Injury System That Didn’t Work

So, if we shouldn’t reinvent the health mechanic each time, maybe we could just enrich it? That’s the question some of my former colleagues on Ghost Recon: Breakpoint had to answer when they were tasked by the direction to introduce an injury mechanic.

The creative director’s idea was to reinforce the theme of the military squad behind the enemy lines and introduce injuries to raise the tension. For instance, while limping, your speed was massively reduced, and you could only shoot with a handgun. The sudden change of controls would (in theory) force the player to go to cover to apply a bandage, giving enemies a chance to close on him… but it never really worked as intended. You’d typically either die shortly after the injury came or could comfortably use the bandage.

Moreover, if you didn’t have a bandage in inventory, you’d limp for very long minutes until finding a bivouac spot, which wasn’t a fun experience for any of the playtesters. You had “successfully survived” the encounter, yet it felt more like a punishment, and you’d wish you died & respawned instead.

Of course, this raises many debates in the team about the fun of the experience and how to account for maluses. Deserved or not, applying negative effects to the player is always a difficult design decision, because everyone has a tolerance threshold for frustration.

After many iterations, the eventual solution to this design problem was turning the bandage into a permanent inventory item while adding a consumable “quick-heal” syringe. This still rewarded the cautious players who prepared their loadout and was less than punishing for those who didn’t. However, it also decreased the intensity of the experience too, but it probably wasn’t the right game to introduce such hardcore mechanics anyway.

Conclusion

In the end, even though I didn’t work on the mechanic directly, seeing the process & the iterations helped me realize that the lack of “deep” health systems isn’t a design oversight. When the game takes away some of the avatar capacities, it’s rarely a great experience, so it must be handled cautiously.

I’m not saying it can’t be done, there are certainly great games with inventive health systems. However, when a mechanic is that powerful and omnipresent, there is always a good reason. When the ‘defaults’ serve us well, there’s no need to reinvent the wheel (even if we often love to as designers!)

If you liked this issue, you may also enjoy that one:

The Arcade Artificer

The Inevitable Open World Towers

When I joined Ubisoft in 2015, the ‘open-world towers’ cemented a large portion of the criticism of the flagship titles of the company. The latest releases, Watch Dogs & The Crew, were raising eyebrows with their questionable integration of that mechanic…

Read more

3 years ago · 6 likes · 3 comments · JB Oger

If we think about it, playing a video game is nothing more than interacting with software whose programmed rules decide what to display on the screen. Our brain conveniently provides the necessary abstraction layer to make sense of those pixels. That's why we need signs to discover the actions available to us and feedback to understand their consequences.

In the past few months, while researching for my first game, I've playtested many demos of indie strategy & card games. Unfortunately, the same problem kept coming back: poorly designed signs & feedback. They are there but often too subtle; as a result, the experience feels dry and difficult to grasp. And it's a shame because signs & feedbacks are the foundations that can make or break the experience.

So, in order to get better at them, I took a look a Hearthstone, the king of digital card games, and discovered it all comes down to two design pillars: sequencing and dramatization.

Layers of feedbacks

Hearthstone is a card game, so the temptation is to look mostly at the interfaces, but one should always remember that signs & feedbacks also include animations, sound effects, AI, FX, voice & more. Making a game clear & responsive is everyone's job in the team: most signs & feedbacks have several layers, and the number one mistake is picking only one.

For one, you never know exactly in which context the indication will be perceived: is the screen small or big? Are people watching or controlling? When was the last time they played, and how much do they remember?

Even in a turn-based game with no time limit, a lot of information can fly over a person's head, you don't want to put all your eggs in the same basket. And you'd also rather have huge Easter-painted eggs than small discreet ones.

Look how bold & explicit Hearthstone can be when it comes to mulligan (the selection of starting cards): there are large green glows around cards you can click and a huge red cross & a “REPLACED” label on the one you'd like to change. The top of the screen reminds you of the actions you can take here, the button at the bottom spells out “Confirm”, and to top it all, the rest of the screen is darkened to ensure your focus stays there. You can't make it more straightforward.

That being said, signs are relatively safe in 2D games with no time pressure: you have tools to direct the player's focus & they have time to process it. On the flip side, feedbacks have to be even better to surface the game's internal rules: the player has to understand the chain of consequences of their decisions to make informed choices. Hearthstone handles those dramatically and with particular attention to sequence.

Designing outstanding feedback in three steps

The key objective of feedback is to inform the player of the consequence of an action and how it changes the state of the game. Most feedbacks in Hearthstone seem to follow a similar structure in three phases:

• Preparation: where you build anticipation & let the player know that something is about to happen, giving them time to pay attention

• Emphasis: where the actual action is performed, and the animation reaches its peak (this is where Hearthstone doesn't shy away from going 300%)

• Aftermath: to give enough buffer to let the player process the information and ensure that several actions won't chain abruptly

Each of these steps is critical to maximizing the readability & the impact of the feedback. Let's take a first simple example with the victory animation below:

• after the last hit, there is a brief static moment when damage numbers are still displayed (preparation)

• then comes the explosion, with pieces of the enemy's portrait flying towards the shaking camera, a loud sound and even fireworks! (emphasis)

• finally, you can notice another short pause once the camera stabilizes before the “Victory” pop-up apparition (aftermath)

Here is another example with one long feedback sequence to inform the player that they reached their turn time limit:

• The rope is deployed & the hero says a voice line (preparation)

• Then plays the sparkling fuse animation, in the middle of the screen (difficult to put more emphasis than this)

• And finally, the “end turn” button explodes (aftermath)

I have a hard time imagining better feedback than this one: it's explicit, it's intuitive, and even if one might think that using a ticking bomb analogy is over-dramatic (it is), it's certainly far more memorable than a boring microwave countdown popping in the corner of the screen. The designers here knew precisely how to catch the attention & to convey critical information visually with the proper emphasis.

Playing with the rules

Once you get the basic principle of sequencing, you may start tweaking each phase to give a different feeling. Indeed, feedbacks cover a lot of contexts, some are positive, some are negative, with more or less importance.

Let's take, for instance, the card draw: in Hearthstone, you have a limit of ten cards in hand, and if you get extra, they are discarded. This potentially frustrating mechanic has to be adequately feedbacked to avoid confusion, especially since it's an exception to the standard rule.

• Preparation: the animation of the card coming from the deck (builds anticipation)

• Emphasis: the card is shown in full size (as usual), but then it explodes, and the hero says a voice line (always accompanied by the speech bubble)

• Aftermath: almost absent here, giving a feeling of abruptness compared to the “expected” aftermath (the card joining your hand)

Missing a card draw is far from ideal in a card game, and this animation adds salt to the injury, leaving you with a sour taste in your mouth while being perfectly explicit about the rule applied.

Once your deck is empty, the rule is that you take increasing “fatigue” damage for each draw: again, they chose similar preparation feedback to remind you it's a consequence of the card drawing phase, and then the animation places the bigger fatigue card in the centre and making it goes in black fume.

For the added dramatic effect, the hole that used to keep your deck is now a lava pit, which you can hover with your mouse to learn more about the rule. Indeed, the text on the “Fatigue” card is difficult to read in the time it appears, but sometimes, it's preferable to keep the timing short and rely on the player's curiosity to dig deeper. In most cases, once the player knows the rule, they won't ever need to read that text again anyway.

One Last Thing

When working on such a topic, there's a risk as the developer who sees them all day long finds them too long & the game less responsive.

In the two gifs above, notice how the player regains control (green outlines) while the animation finishes: this is the key to ensuring the player never feels locked in an animation they might have seen a hundred times. Of course, developing such a system that decorrelates animation from interaction takes more time, but it's worth it.

A developer is never the best person to assess the quality & relevance of their own signs & feedback anyway; they know the game too well for that. So remember always to add more.

If you liked this piece about signs & feedbacks, you may also like:

The Arcade Artificer

Stacklands and the UX of Cards

Among our 270 273 274 board games collection, a whopping total of 236 have cards. Whether they’re actions to perform, unpredictable events decks, characters, items, abstracts, or figurative: cards are everywhere. Why? Why is this rectangular cardboard piece the most used component in board games, far ahead of pawns & dice…

Read more

3 years ago · 4 likes · JB Oger