Gamasutra: The Art & Business of Making Gamesspacer
Skill Ceiling
Printer-Friendly VersionPrinter-Friendly Version
View All     RSS
April 18, 2014
arrowPress Releases
April 18, 2014
PR Newswire
View All
View All     Submit Event

If you enjoy reading this site, you might also want to check out these UBM TechWeb sites:

Skill Ceiling
by Tynan Sylvester on 12/06/09 10:53:00 pm   Expert Blogs   Featured Blogs

The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.


People talk about the depth of games a lot, but it’s tricky to figure out exactly what that means. I’ve been thinking about a new way to measure the depth of games. It’s the Skill Ceiling.

The skill ceiling is basically the answer to the question:

How much can a player theoretically improve their skill at this game before there is no way to get better?


How good can a hypothetical perfect player be at this game? How much better is this from average performance?

It’s easy to measure games on this scale.

Modern Warfare 2 has an extremely high skill ceiling - so high that no human being will ever come close to it. It runs at 60 FPS, game verbs generally have little or no inherent latency, and the game-state can change very quickly with skilled play. It is possible, with excellent tactics and aim, to eliminate entire enemy teams in moments. A theoretical perfect player in that game could singlehandedly defeat an entire team of very skilled human players.

In the middle of the spectrum, Assassin’s Creed 2 has a medium skill ceiling. A perfect player could do significantly better than an normal player, but would not be so astronomically beyond him as in MW2. Many game actions initiate momentary losses of player control - while Ezio swings a sword or grabs another handhold - during which the normal player can mentally catch up with the perfect player. The gamestate cannot change nearly as fast as in MW2. Even with perfect swordplay, for example, it would take at least 50 seconds or so to kill 10 enemies, since the animations to kill enemies take about 5 seconds. A normal player could probably get quite close to this level of performance with practice, since the optimal strategy doesn’t is not that complex and does not require inhuman reflexes. Ezio’s performance is limited by the game instead of being limited only by player skill.

On the low end of the spectrum are games that quickly break down into degenerate strategies, like Tic Tac Toe. Game designers who want their games to play out exactly like the envision them often end up with games like this, since there are so few strategies. This is, for example, the problem with using quicktime events as skill challenges. Since there is a binary outcome which can be secured by an average player, there is no way to improve on this, so there is no difference between an average player and a theoretically perfect player.

Toylikes like The Sims can’t be measured by skill ceiling per se since there is no traditional goal or competition, so this concept breaks down when applied to them.

You’ll note a sharp downward trend in terms of replayability as the skill ceiling falls. The reason for this is that replayability ends as soon as a player consumes all content and hits the skill ceiling. MW2 multiplayer is endlessly replayable since there are always ways to get better, even for players who are inhumanly skilled.

There are two ways to build skill ceiling that I’ve thought of:

The first is the SHMUP method, which is to simply demand such accurate moment-to-moment input that no player could ever do it perfectly. The game simply runs so fast that no human mind can keep up. Racing games, SHMUPS, and fighting games depend on this method to a high degree. Pretty much any game sped up to a high speed will start to do this. Imagine playing Tic-Tac-Toe, but you get a third of a second to make your move. As a mental exercise, you can strip out this element of any game by slowing it down. Imagine MW2 multiplayer played at 10% normal speed.

The second is the Go method. This has nothing to do with twitch skills and all to do with managing complex strategic and tactical information. The game simply presents so many options and variables that nobody can easily see where it is going, even if they can stare at is as long as they want. This type of complexity is harder to do meaningfully than simple high frequency input.

The best games use both these methods. Modern Warfare 2 multiplayer, for example, uses both. Starcraft does as well. Both games are very different, but both have vibrant online communities.

Even in Modern Warfare 2, note the replayability difference between online multiplayer and Spec Ops, which really only uses the SHMUP method of building skill ceiling since the AIs are so simple and predictable.

The above theory is descriptive, but not predictive. I’ll be thinking and writing more about how to design games for high skill ceiling in the future.


Crossposted from

Related Jobs

Penny Publications, LLC
Penny Publications, LLC — Norwalk, Connecticut, United States

Game Designer
Hasbro — Pawtucket, Rhode Island, United States

Sr. Designer/Producer, Integrated Play
Nexon America, Inc.
Nexon America, Inc. — El Segundo , California, United States

Web Designer - Temporary - 3 month
Darkside Game Studios
Darkside Game Studios — Sunrise, Florida, United States

Mid-Senior Graphics Programmer


Luis Guimaraes
profile image
It inspirated to put some more thought in it, even I'm not gonna work in nothing with this any time sooner.

I have an old post that might be useful:

And one practical exemple of what I was talking about in that post:

Smarter than a Movement Macro Record I have thought of...

To be considered deep, a game should be deeper (but not necessarily over-complex) than the average game around. The MMO sphere is one I consider games to be very deep, but sub-considered to the sake of levels, most players just don't get that depth.

I'll use my brother Julio here as exemple, he plays many web-based and mmos and I consider him the best player of all these genres (sometimes he even inspires me to play these games, showing me how interesting they can be if you know how to do it). He's currently playing an mmo called Perfect World (well, everybody in Brazil is). While my other brother and all their friends spend all the time doing quests and "mobbing" as they call, Julio is not.

While many of them are in levels 80-90, he has 3 or 4 chars in level 50, some of them which easily beat any lvl 70 char in the game, but only if it's him playing with them. That's what he does all day, socialize, beat players with 20+ levels above him, and makes tons of money with smart business and social skills (to spend in any useless jokes against his friends, just because he can make all that money again whenever he wants), and quoting Sun Tzu.

Glenn Storm
profile image
That is a great perspective on charting the difficulty progression or flow progression. In a way, this is precisely what the player is gauging for themselves, as an important measure of the game's play value. I like your lines of thinking in offering ways to develop that ceiling. That's cool. Thanks, Tynan.

Bart Stewart
profile image
Tynan, I think you're onto something interesting with your suggestion that the depth of a game can be equated with how soon the typical player of a game reaches the skill ceiling designed for that game.

I'd like to suggest it might also be useful to think about what you mean by the word "skill." What kind of skill does/should a game measure?

Your discussion focuses on a couple of skill types. There's what you called SHMUP-type skill, which I see as being about measuring a player's action-oriented capabilities. It rewards memorization (of static rules and geography) and fast-twitch action.

And there's skill related to complexity, which seems to measure intellectual capacity. Depth in this kind of game appears to come in two forms: complexity of inputs (ability to grasp many dynamic rules), and broadness of choice/consequence patterns (lookahead, or the ability to perceive and assess many potential future consequences).

But what about emotional skill? What would be the upper limit on how well a player can understand an emotional environment of interrelationships and somehow "win" that game? How would you define depth (or "skill ceiling") in such a game?

Deepest of all, it seems, would be the games that offer a high skill ceiling across all three of these modes of play simultaneously -- there would be more for the hands, and the head, and the heart, than could be consumed in just one playing.

That would be an interesting game. :)

Luis Guimaraes
profile image

Strange for an FPS to get that place, but when you talked, my mind went to Counter Strike. I've played it with clans and entered professional leagues. First of all, the basics of the game are about action skill. When you play everything the game offers it has a lot of intellectual skill involved, and when you play it competitivelly, what really wins is how you handle the pressure and concentration.

Bart Stewart
profile image
Luis, doesn't that -- along with much of Tynan's original post -- suggest that the deepest games are those played competitively and cooperatively (at the same time) against and with other human players?

That implies a follow-up question: what would be required to design a single-player game to have an equally high skill ceiling (i.e., to be equally "deep")?

Glenn Storm
profile image
@Bart: Perhaps a strong narrative character progression (story) tied closely to the primary player abilities? Just story probably wouldn't do it. Just ramping abilities and difficulties may not. But both tied together seems like it might.

Luis Guimaraes
profile image

Oh, That does. (And also explains part of CS success that remains until nowadays), I didn't notice that when I replied you.

Good question, and maybe part of the answer lies more on our gamming culture (at least for single player games) than the games itself, but it's too lazy for a designer to blame other factors. By the way I must thank you for the challenge, I would miss a good opportunity.

Single players are a complex case, they often don't have enough replayability as multiplayer and cooperative ones, specially now in the 40-hours-lenght generation. It doesn't help too much for action skills to grow high, the games that benefit from it are often the core-genre type (I go for any FPS in the hardest possible), but relative new gameplays don't have replay enough. I say replay because of the frustation factor, games are managed to be quite possible in the first playthrough, and the player is about to stop playing, or learn as much as needed to progress (nothing wrong with that).

The intellectual side is more of the single player strenght, in the point where this skill isn't affected by the 3rd type (emotional skill) in most strategic-driven single-players. Desperados and Heroes 3 are some of my favourite of all time.

The emotional side (which can be a problem if not well handled) in single player works better in scary games, but this time they are most likely to affect the quick reaction than the smart thinking. Even Resident Evil 4, not exactly a scary game, made me have real nervous time when all the ammunition, that I saved with so much effort of action and smart skills, was running near the end, and the bosses were still in combat after many minutes of running and shooting.

(Here I was making a short list of single player games that do great in one of the 3 skills types we're considering)

While thinking of which games I'd not put in the study list (for exemple Splinter Cell) and the reasons why, it gave me a thought: skill measure is about possibility measure. I'll try to exemplificate:

Let's say our possibility space goes from 0% to 100%, and the following analisis is taken:

the average-player skill is 20%-80% (not fixed number, but a constant oscilation),

the poor skill is 1%-19%

the awesome skill is 80%-99%

the perfect skill is 100%

For the game to be really deep (in that one given skill type), the performance result should match the skill levels. In other words, if a performance of 60% is the "perfect" performance, then the game has 40% less depth it has potential for. The average player can reach the perfect result in 1 out of 4 tries. If a performance of 20% is enough to progress in the game, the average would (theorically) never fail (and get bored).

After allowing real high ceiling for skill, the game must find a way to reward and incentive the search for more. In other words, there's a clear difference of goal in most singleplayer and multiplayer games.

In Bioshock, which I played entirely in stealth the first time, I had a constant goal of "having MORE ammunition" "MORE health and adam packs" and "MORE pics of the enemies". That, aside from the game goal, brought a rich experience, exploration, hunting down enemies and "mobing" around long time before I proceed. I loved each place and spent a lot of time in each one, taking the best I could from it.

In multiplayer games, the goal (aside from the casual part most players have) is MORE, while in single player it's ENOUGH.

Luis Guimaraes
profile image
@Glenn Storm

One step ahead! =)

Story is the best way to give a goal to the player (the worst is probably and Trophies and Achievements).

But in multiplayer, getting better (and in which direction) is a choice. In single player, succeding through some obstacle at any cost, and succeding with full health and not using potions or bombs are different choices, that tie to the skill-seeking which leads to experiencing the depth of the game.

Using the story with smart non-cliche goals and the right dose of choice (and consequences to it, as using your rocket laucher around will leave you in trouble against the next boss), can make a real difference towards the goal, which is, in the end, teaching your player how to enjoy the best of the game (not too different than a waiter suggesting the best drink to follow your dishe).

Perceived depth is as important as real depth.

Joshua McDonald
profile image
"what would be required to design a single-player game to have an equally high skill ceiling"

There are a lot of approaches to move in this direction, but I think first and foremost is to make a game that is less scripted than modern single-player games typically are. Many games seem to be aimed at the "Simon Says" style gameplay where you use specific abilities or items in specific places to progress through the game. One that seeks a higher skill cap needs to simply give the player a set of tools, and environment to use them in, and ideally, maximum rewards for insanely difficult tasks.

While there are no perfect examples of this that I can think of, there are a couple games that at least make some progress in this direction:

Battleforge (the single player part): My personal favorite RTS: Going into the exact mechanics would be too long, but their expert difficulty levels can be ridiculously hard, not because they demand an exact pattern to follow or perfect twitch/micro skill but because the player needs to come up with an effective way to manipulate the rules to overcome the challenges. Watching different players is interesting because they devise totally different ways of handling these challenges. Often what seems utterly impossible simply requires a bit more creativity in your tactics.

Tactics Ogre: Old rare PS1 game: This one is a highly difficult strategy-RPG that, while it provides NPC's and a story, also allows you to ditch the characters you are given and completely craft your own army. Once you dig into the mechanics of this game, you'll find that you can craft highly interesting armies that are far cooler and more effective than simply playing with the provided NPC's, and the game becomes ultimately more satisfying as you're prospering based on your own ingenuity, instead of your ability to follow the game's instructions.

Single player games that choose to focus on story and tightly controlled experiences will probably always fall short in this area. For many people, that's a worthwhile trade-off, but if you get your enjoyment from mastering deep and interesting skills, you'll need a game where designers focus more on giving you building blocks for an experience instead of a pre-built one.

David Wipperfurth
profile image
I like the general scope and target of this concept. I also like the way you chose to break-down skill, into number of processed outcomes over time (GO * SHMUP). This has the benefit of a simple empirical analysis, basically comparing a human's skill to a computers FLOPS. This has some down-sides, but I think there is a lot to be said for the general nature of it.

While I think using the idea of 'Skill Ceiling' vaguely might be helpful for conveying the gist of some arguments, I don't think re-evaluating a games difficulty from a learning curve to just the curves global maxima is more than slightly useful for analyzing/describing games. Most games have no calculable maxima, and the rate at which humans stop progressing along the curve has little to do with the maxima of that curve.

More over, I don't feel 'Skill Ceiling' equates to depth if you measure it as GO * SHMUP. This may be an overly technical analysis of your concept, but GO in that equation represents all possible states, both depth and breath. SHMUP represents the rate at which the player must progress through those states. depth is just depth. 'Skill Ceiling' is (depth*breath)/time.

If on the other hand you mean depth to be how engrossed a player can get in a game, than analyzing a game as (GO * SHMUP) doesn't work out so well. (GO * SHMUP) leaves out a crucial element, which is what makes (GO * SHMUP) a powerful concept, but also why it doesn't work for this type of 'depth'. That element in heuristics. Humans don't actually process all possible game states(obviously), the best human player is not the one that can process the highest FLOPS, but the one managing the best system of heuristics. How engrossed a player can get in a game is based on how long they can continue to develop and refine their system of heuristics while still maintaining an adequate amount of satisfaction. This is linked to GO in that the possible complexity of a heuristic system is limited by the domain of the game. It is also linked to SHMUP in that a player has a harder time developing their heuristic system when limited by time. The quantity of the game state domain alone, however, says nothing about what patterns can be found in the game state data, or how the player gets rewarded for continuing to search for those patterns.

I would still enjoy using the term 'Skill Ceiling' to vaguely describe the phenomena of the point on the learning curve where players fall off.

Thanks for the provocative read Tynan.

Tynan Sylvester
profile image
Bart Stewart: "That implies a follow-up question: what would be required to design a single-player game to have an equally high skill ceiling (i.e., to be equally "deep")?"

Without strong AI, I'm not sure it's even possible. The complexity of a gamespace that involves the actions of other real humans, each with a personality, mood, skillset, and so on, is instantly and automatically vastly more complex than a pure machine-driven gamespace. Most advanced MW2 tactics (near the top of the skill curve) are really about deception, manipulation, and prediction of other players. They're human interaction more than anything else.

J. McDonald: "There are a lot of approaches to move in this direction, but I think first and foremost is to make a game that is less scripted than modern single-player games typically are."

This is really tricky for a lot of reasons. I think it is possible to hit a decent compromise. Batman: AA, for example, is scripted up the ass in a broad sense, but within the individual gameplay challenges (fights and stealth attack sections), the player has endless options. They even sliced their core gameplay out into challenges which you can play over and over to get better. I tried the challenges for a while, but got bored or trying to improve my scores because the game didn't hint at better Go-level strategies. The twitch skill component could always be improved, but when trying to improve the actual arrangement of actions I was taking, it was hard to think of new strategies because I had no examples to work from. I think this is part of the reason multiplayer works well this way: you see superior players using strategies and can learn from them. This is very different from trying to think of a better strategy in a Batman Predator Challenge, which requires you to figure it out in a vacuum. But ah, I've digressed.

D. Wipperfurth: "basically comparing a human's skill to a computers FLOPS".

Interesting that you bring up this idea. I think this comparison is actually flawed, for all the reasons that computers with insanely high FLOPS can still be challenged by skilled chess players. Human skill is a complex phenomenon, but I understand that it isn't about your FLOPS (which never changes), but about how much knowledge you have about the game and how well organized it is in your head. This is how skill builds - building knowledge. So you could restate the concept of skill ceiling by asking, "How much does a person need to learn before they can play the game essentially perfectly?"

Per the difficulty curve: You're right, this model doesn't take into account information along the player's learning curve, just its end. In the real playing/learning process, the entire curve needs to be gratifying. It's a simplified model; I may sit down and try to think of a more complete one.

"'Skill Ceiling' is (depth*breath)/time.": Cool, but see my above comment about knowledge. In abstract, you're right, but I'm not sure this is how human learning and skill works.

Heuristics: I can see you already anticipated my earlier point about knowledge, and probably expressed it better.

Thanks for the replies everyone.