Hey folks,

Today’s piece focuses on player creativity in esports and how game designers can encourage (or discourage) it.

What is Creativity?

To kick things off, let’s define our terminology:

creativity – the use of the imagination or original ideas, especially in the production of an artistic work. synonyms: inventiveness, imagination, innovation, innovativeness, originality, individuality (Google)

Creativity is originality – the expression of novel ideas or the creation of unique works. It is not merely repeating what’s been done before. The relationship with individuality is also interesting, and reasonable when you think about it. If uniqueness is an integral part to creativity, then it’s likely a contributing factor to it will be the the indivisible uniqueness of the individual.

Player creativity in a game, then, is an individual player’s novel ways of thinking or doing while playing. This means that, at a minimum, a game needs to allow novelty in the first place. We shouldn’t assume that this is an essential component of game design. Plenty of brilliant games neuter or outright prevent individual playstyles. I’m currently in the process of an in-depth commentary on Banjo-Kazooie, an absolutely incredible title that limits creative player action in almost every way. Playthroughs of the game by different players will be almost identical, modulo differences in ability that prevent the collection of harder late-game items.

When I discuss creativity in this piece, I’ll be focusing specifically on creative action or creative works – in other words, excluding merely thinking creatively. For instance, a player in StarCraft can use their imagination to see a ladder game as a literal acting-out of the game’s plot, and perhaps the game encourages or discourages that line of thinking. This piece won’t worry about that – here, we’re focused on creative works.

Creative works are both novel (unique, individual, etc) and useful. I pull this definition from the Creativity Achievement Questionnaire, which itself pulls it from a 1955 study on advertising. The idea is that it’s not enough for something to be novel, it must also provide some value… it “must be both original and functional or adapted in some pragmatic way to reality”. Otherwise, any random action in the infiniteness of the human condition could be described as creative.

Utility is a useful concept in esports because esports emphasize competition, improvement, and winning. While they are certainly intrinsically enjoyable to play, they also encourage players to play better. It’s not merely enough to express your individuality to win at an esport. Your expression must provide you with value, whatever that may be.

For example, a creative player might leverage their high individual ability with banelings, burrow, and multi-tasking to effectively use Baneling bombs to win games. I’d describe this as a creative work because it’s both novel and useful.

Another player might creatively design a new build order – ebay first, three CC into first supply depot. This is novel – but not useful.

Why Facilitate Creativity?

I think there are a lot of good reasons for game designers to facilitate (i.e. encourage and make possible) player creativity in games. For one, I think injecting one’s individual playstyle into the structure of a game and seeing how it plays out is inherently fun. I’d say that’s a big appeal of the Souls series. For another, it can enable replayability by allowing players to experience the same game multiple times, but in different ways depending on how they choose to play the game.

When it comes to esports specifically, I think there are specific, compelling motivations for enabling individual creativity separate from traditional games. Here are two I consider to be important:

Content-driven vs. Player-driven

In game design, the designer typically views their role as enabling players to have a specific experience. Here’s Jesse Schell describing the concept in The Art of Game Design: A Book of Lenses:

… And this is the paradox of experiences. On one level, they are shadowy and nebulous, and on another, they are all we know. But as tricky as experiences can be, creating them is all a game designer really cares about.

There isn’t necessarily one “perfect” way of accomplishing this. I’d argue esports share a common way of going about it that’s distinct from traditional games – they’re player-driven rather than content-driven.

In a traditional game, a designer creates a number of gameplay systems and mechanics that work in concert to convey a specific player experience. The designer drives the player to engage with these mechanics using content – obstacles, challenges, worlds to explore, dialogue to discover, a plot to unravel, and so on. The two pieces work together such that if a designer has done their job, by the end of the game the player has gotten whatever experience the designer intended to convey.

Mass Effect (the original) is a great example of a traditional, content-driven game. The designers created a whole suite of mechanics, from the large and interesting universe to compelling characters to enjoyable gunplay. They then build a bunch of content, from the initial descent to Eden Prime all the way to the hard choices of Virmire, that exercise those mechanics in a way that’s intended to convey an experience.

(Note that you could also view this in reverse and argue that mechanics are sometimes built to exercise content rather than the other way around. You could also note the interdependency between mechanics and content, and ask where the line is drawn. Both very interesting and, for the purposes of this discussion, not important. Another time.)

For instance, choice is a big part of Mass Effect, and also one of its most criticized features. Many folks, including the creative director of the latest Mass Effect: Andomeda, noted that players tended to pick ways of playing (either Paragon or Renegade) rather than genuinely engaging with the dialogue.

I’d argue that these criticisms miss the point, and help explain why Andromeda’s new arbitrary tonal dialogue system has been poorly received by players. The mechanic BioWare adopted in the original Mass Effect was choice, but the experience was a lot more than that. It was about the ability to interactively live as the hero – or a renegade of chaos, depending on your preferences. The core story arc of the first game has more in common with Harry Potter, the Lord of the Rings, or C.S. Lewis than it does with anything in science fiction – except this time, you get to be the paragon of moral virtue rather than just reading about it.

The point of this long tangent is two-fold, and brings us back to our original point. Like many traditional games, Mass Effects attempts to 1) convey a specific experience by 2) driving that experience with content.

I’d argue esports work differently. Relative to traditional games, they feature less content. Despite this, esports also expect players to invest more hours than traditional, content-heavy games.

Take Counter-Strike: Global Offensive. There are only eight or so active competitive maps at any given time, and each is small enough to walk across in about a minute. But play any amount of matchmaking and observe how players measure playtime by the hundreds of hours. It’s rare for a traditional game to engross players with so little content for so long.

Esports do this by enabling players to generate value in their own experiences rather than building content to do it for them. A symptom of this approach is that esports tend to feature simpler graphical styles than other games.

Take CS:GO (an esport) and Battlefield 1 (a traditional multiplayer shooter). Counter-Strike’s worlds are unusually bright and clutter-free relative to other large multiplayer shooters, a comparison I examined in-depth in my video on player-driven design. By contrast, here’s a quote from a Senior Concept Artist at DICE explaining how the company thinks about Battlefield, a game that I would classify as featuring traditional game design:

“…Battlefield is also about the epicness of it, the vastness, and so it’s key we get that into each level. We also cater to a good variation with everything – the idea is that no battle is ever the same. In level design a big factor is our dynamic events, and the destruction this time is more intuitive, it’s more natural. We can reshape the environment around you when you play, but also things like the dynamic weather comes in to give a sort of fresh unpredictability to levels. You can have clear rain suddenly turn to thick fog or heavy rain, and in the case of Sinai Desert there’s a sandstorm – making sure that it’s never the same experience twice – you always have a new one.” – Robert Sammelin, Senior Concept Artist at DICE

The content within the game itself – in this case, the weather and the multitude of graphical effects that accompany it – drives the player’s experience.

Counter-Strike is different, intentionally. The game’s designers want to get out of the player’s way. The game is an environment, an arena, a facilitator – it provides enough structure and content for players to drive and construct their own compelling experiences.

How a player in Counter-Strike approaches a situation depends more on them than it does on anything specified in the game – their sense of where the game is, their belief in their ability to use the weapon they have, their effectiveness in collaborating with their teammates, what they personally enjoy doing, and so on.

A player in competitive Counter-Strike might defend a particular bombsite or rush down a particular hallway thousands upon thousands of times. The value in doing so isn’t content within the game – it doesn’t suddenly start raining in Dust2, the visibility in Mirage’s palace doesn’t randomly change from round to round. The game doesn’t proactively try to make itself more interesting. Instead, the player drives the experience based on their individual style, skill, and what they find enjoyable.

And this brings us back full circle. Esports, because they are relatively less content-driven than traditional games, benefit more from design that facilitates player creativity and individual playstyles. The more a designer can enable a player to play a game “their own way”, the more likely it is that a player will enjoy their experience.

Watchability

In today’s edition of “what a time to be alive”, how a game comes across on an online video platform like YouTube is a reasonable consideration to take into account when building it. This is true even for very traditional, very content-driven games: here’s an example from the very wonderful Overcooked.

Esports are often even more concerned with this than traditional games. Many esports are also spectator sports, with massive tournaments, professional players, even live television broadcasts. The professional scene builds engagement with the title and helps drive revenue – it’s win-win. As a result, it’s part of an esports game designer’s job to care about how their game comes across on the big screen.

I’d argue that facilitating creativity and individual playstyles goes a long way in improving the watchability of an esport. One way that it does this is by enabling professional players to stand out from their peers and showcase unique styles. This allows these players to build fan bases and attract people to events by name alone. Tournament organizers can build on this by promoting specific storylines. For instance, I really enjoy watching TRUE play StarCraft II, because his ability to single-mindedly focus on ling/bane and still play at a Code S level is incredibly fun to watch.

Individuality can also counterbalance weaknesses in the meta-game. If the moment-to-moment gameplay in professional competition becomes too predictable, it will also become boring. This will drive spectators away from the sport. The ability for players to make plays and showcase creativity can offset this and keep things interesting.

Game Systems

Next, I’d like to talk about a couple game systems that I think are interesting to think about through the lens of facilitating player creativity. By thinking about how these systems work and what their advantages and disadvantages are, we might be able to tease out some good, general ideas for facilitating creativity in esports.

Game System #1: Weapon Viability in Counter-Strike: Global Offensive

Counter-Strike: Global Offensive is different from other modern shooters in that its competitive mode emphasizes the use of a very small number of weapons. Most weapons are difficult or outright inadvisable to use in professional matches – at the very highest level, weapons like the AK-47 are simply better than almost everything else a player can purchase while on the T-side.

I’m not sure how intentional this design was. Counter-Strike was originally a mod for Half-Life, whose online Deathmatch emphasized the use of a small fraction of its total offering of weapons. The developers of Counter-Strike at least made some effort to balance out the guns (remember the original G5?). Regardless of what the intent was, the result by 1.6 was an even stronger version of what’s described above. Hardly anything was viable in competitive play, and entire matches would go by where the only weapons that were purchased were the M4, AK-47, AWP, and Deagle.

I’d argue that this has a positive impact on player creativity and individual play style, in a few respects. Before I do that, I’ll first differentiate between “generally viable” and “situationally viable”.

A weapon that’s generally viable in competitive play is one that can be purchased in most, if not all, situations. The game’s Assault Rifles fall into this category.

It’s not really possible to make multiple weapons of the same class “generally viable”. If two weapons can be used in most or all situations, then one of two things is true: either they’re identical or they’re not. If they’re identical, then we haven’t accomplished our goal of making different weapons viable. If they’re not identical, then one will be generally better than the other, and players will always choose that one.

By contrast, a weapon that’s situationally viable is the ideal weapon for a player in a limited number of situations. For instance, sub-machine guns in Global Offensive are situationally viable. They’re only the optimal choice some of the time; however, in those situations they’re arguably better than rifles because they net the player a larger monetary reward for getting a kill. This can make it easier to fund an AWP purchase a round or two later.

Thus, it is possible to make more weapons situationally viable, but not generally viable. General viability is a zero-sum game. Trying to increase the number of situationally viable weapons creates a bunch of different problems, and I’ll discuss the ones related to facilitating creativity.

On a practical level, I think a smaller number of viable weapons reduces designer interference and allows the meta-game to develop organically. To state this in reverse, if a designer makes more weapons situationally viable, then a player’s actions in those situations will by definition be less creative. The designer has injected themselves into the situation and proclaimed that a particular set of behaviors – in this case, using a particular weapon – should be prioritized over what the player individually thinks is best.

There’s also another practical issue with respect to game balance. Esports are complicated not just by design but in practice – they’re played at a professional level, which means that there’s a constant race to identify new nuggets of information and find new depth within the gameplay. If there’s a balance or design flaw within the game, a professional player will eventually find it. Its use in a professional game will then spread awareness to the rest of the world all but instantly.

Each additional gameplay feature, like another weapon that’s intended to be situationally viable at a professional level, makes the job of balancing the game exponentially more difficult. All the different situations the designer needs to keep track of suddenly have a new dimension worth thinking about. It’s inevitable that this will create pathological problems; it’s inevitable that these pathological problems will become widely known; and in the end, every single one will limit creative thinking. Players are encouraged to do whatever the meta prescribes in this “broken” situation rather than doing what they individually think is best.

This demands a response from the designer, who issues a balance patch. But the problem was created by designer interference in the first place, meaning that another patch risks creating more problems. This can turn into something of an infinite loop.

That’s compelling, but we’re still not done making our argument. Twice now I’ve made this statement – “doing what the player individually thinks is best” – and cited it as a benefit of facilitating creativity. Surely there are readers who have experienced the wonder that is queueing up for matchmaking and witnessing a teammate buy pump shotguns every round. Perhaps we don’t want players to individually determine what’s “best”?

It’s definitely true that there lots of “wrong” responses to particular situations. But the solution to this is not for the designer to prescribe a specific approach to every situation; it’s not like that’s going to convince the guy who doesn’t see the merits of buying Kevlar.

Rather, what I appreciate about the lack of weapon viability in Counter-Strike is the way it enables a spectrum of correct choices. What it lacks in viable weapons it makes up for it in depth. As I discussed previously:

Spray patterns, interactions between accuracy and movement, optimal movement across a map, playing your opponent’s positions, coordinating effectively with teammates and so on are deep, complex, and fine-grained skills that can never really be mastered, only improved upon. Because the skill ceiling is unreachable, a player who works hard to develop skill in a particular playstyle will beat a player who works less hard on a theoretically more optimal way of playing the game. The high skill ceiling removes the concept of diminishing returns and the harder-working player wins out.

This is pure facilitation of creativity – whatever it is that needs to be done, the game isn’t giving you any more tools. This is great, because it means the player doesn’t have to think about what the designer wanted them to do. They can just buy an AK and practice. As in the pump shotgun example above, this is not perfect. There will always be lots of wrong ways of playing the game, where wrong is defined as a decision that all but certainly reduces your probability of winning a game. But this does not mean that there should only be one right decision – there can be lots, and it’s very player-dependent and situational. Not only is this fun to play, it’s extremely fun to watch as well.

A counter-intuitive takeaway from this is that there is such a thing as too much choice. The limitation in the number of viable weapons actually increases the number of real player choices because the depth in the handful of weapons that the player does use enables the player to choose many different correct responses to a situation. They can play aggressively or passively, close or far, and so on. The outcome depends on their skill level rather than what the designer has prescribed, creating genuine choices.

I think there’s a larger principle at work here relating to structure and its impact on creativity. I don’t think we’ve really “nailed it”, but I do think we’re onto something. More on this in the future.

Game System #2: All-in Rushes in StarCraft II

StarCraft II features a large number of early game cheeses and all-in rushes. Coming from an Age of Empires background, this was initially quite frustrating. After the wide prevalance of feudal age rushing in Age of Empires 2, Ensemble substantially buffed early game defenses and reigned in the sprawling nature of the economy, dealing a death blow to most rush strategies. As a result, the competitive meta typically featured longer games that ended in the mid-to-late game.

Over time, I came to appreciate this difference in design philosophy and now consider StarCraft’s approach to be superior. I’d argue that cheesing facilitates creativity and individual playstyles in the larger meta-game.

It does this by preventing players from defaulting to the greediest possible openers. Aside from the first thirty seconds or so of gameplay, openers begin to diverge depending on how safe or greedy a player wants to play. The right decision depends on a lot of factors, including the map, the match-up, what they think their opponent is going to do, the scouting information they have, and, importantly, how they individually like to play.

This leads into a huge variety of mid-game compositions. This is important because, going back to our Counter-Strike example, the outcome in any given situation often depends on what a player has practiced and what they’re comfortable doing rather than a prescription from the designer. Enabling variety in the mid-game means that a lot of different players are going to get good at a lot of different things. This helps contribute to strategic diversity in the competitive scene.

For instance, watch professional level games and ask yourself how often a player could have opened in the most greedy way possible. The answer is “most of the time” – so why don’t they? Because they know that they could get cheesed, so they play some variation of a safe opener and scout to see how greedy they can be. This makes things a lot more interesting and creates a huge variety in games instead of the same thing over and over.

You can see additional evidence of this by reading a retrospective on some of the top players in StarCraft II. The number of different referenced playstyles is stunning. The best players were often famous for the way they played the game particularly well, for their individual styles and what they liked to focus on. I’d argue that cheesing was part of why this was possible – by creating genuine strategic diversity in the mid-game.

All that said, what’s interesting about cheesing is the way its balance sits on a knife’s edge: there are instances where it hampers creativity, too. Three-rax reaper in TvZ and TvT prior to the Cyclone buff and grenade nerf, particularly on certain reaper-friendly maps like Orbital Shipyard and Galactic Process, created a huge number of one-dimensional games. TvT especially became more and more one-dimensional until the Cyclone buff in Patch 3.8, with many top players adopting the standard reaper-helion opening in virtually every game.

Why didn’t three-rax reaper facilitate creativity? I’d argue the problem was its lack of depth. With most cheeses, there’s a spectrum of all-in-ness and a spectrum of safety. There’s a ton of compelling decision making where players need to decide how much to commit and what to do next. Scouting and using game sense to make decisions with incomplete information come into play. A lot of a player’s success hinges on their individual ability to control their units and macro effectively under pressure.

Three-rax reaper diverged from this in two important ways. One is how early it hit – in TvT, if a player did a standard fast expand opener against three-rax reaper, they would take game-ending damage before their factory was complete. This negated even the “standard safe” openers, effectively forcing players to open reaper-helion.

That might not sound so bad, but there’s a big difference between openers diverging at the thirty second and three minute marks – the latter has already stripped away a huge amount of strategic variety from the mid-game.

On top of this, reapers’ grenades and health regeneration create a snowball effect in which the strategy actually gets stronger in the early-mid game instead of weaker, in contrast with most other cheeses. This meant the defending player needed to assume the cheese was continuing longer than they would for other cheeses.

These two assumptions culled the spectrum of creative responses – it limited the number of viable openers and limited the number of mid-game “forks in the road” for players to choose from. This created a more one-dimensional playing experience.

The takeaway, for me, is that cheeses need to find the right balance between preventing overly greedy play while also not forcing overly safe play. They need to facilitate a complex and deep mid-game rather than forcing it in one particular direction. When this is done well – as I’d argue it generally is in StarCraft II – it enables lots of creativity from players and makes the game a whole lot more fun to watch to boot.

Before I go further, I do want to address a common counter-argument to the existence of cheeses. Many folks argue (correctly) that they are a pain point for players. Many players would rather play macro games without worrying about the huge variety in early game rushes.

I completely understand this perspective; I don’t think cheeses are a free win, from a design perspective, by any means. But I also think that addressing this pain point is more complicated than simply preventing rushes. Without the threat of a cheese, a lot of depth gets taken away from the mid-game because players all open the same way. This ironically takes away a lot of the fun of the macro games we were trying to encourage in the first place.

Final Thoughts

Creativity in games is a huge topic, deserving way more coverage than what’s written here. I hope I’ve at least managed to do it some proper justice with this piece. I’d argue esports benefit greatly from facilitating player creativity, and we can learn a lot from existing games as to how to do that effectively.

Thanks for reading! If you enjoyed this article, please consider following me on Twitter or Facebook and checking out my game-design-focused YouTube and Twitch channels. All the best and see you next time.