The Curious Case of Complexity


One thing that's been weighing on my mind in recent months has been the role of complexity in game design. When I speak about complexity, I'm referring to the number of mechanical tasks or important decisions that a player is asked to make in the course of gameplay. Some games might have a relatively low level of complexity to their design, like a puzzle game that simply asks the player to perform a simple sequence of tasks faster and faster as time progresses, while on the other end of the spectrum, there are strategy games, MMOs, and the like that ask the player to juggle a dizzying array of subsystems in order to succeed. I wanted to type up something on this subject because the concept of complexity in game design is one of the most misunderstood topics that I see floating around on the Internet. The typical poster on Reddit has a poor understanding of what will lead to good gameplay in this regard, and as a result, I often see all sorts of nonsense thrown out there. Even worse, sometimes this kind of misunderstanding gets thrown into actual professional game designs, as developers feel the need to "listen to the fans", resulting in all sorts of bad decisions. Let's try to address this topic and clear up some misconceptions.

First and foremost, there's a very common misconception that adding more complexity to a game makes the game better. This is posted everywhere on the Internet, and particularly in strategy game forums. In practice, game design is not nearly that simple. Increasing the mechanical complexity of a game can sometimes result in stronger gameplay, but it can just as easily have the opposite effect. Piling on more and more subsystems can also bog down an otherwise good design in torturous busywork, sucking the life out of whatever goal the developer was trying to achieve.

The ideal that everyone should be looking for is not complexity, but DEPTH. These are completely different concepts, unfortunately often mistaken for being the same thing. A complicated game mechanic is something like the economic system in a Paradox strategy game (Europa Universalis and the like). The player has to manage income, stability, research, casus belli, vassals, alliances, cultural and religious affiliations, their "bad boy" rating, and so on. It's a vastly complex system to understand, and don't even get me started on the nightmare that is the economic engine in Victoria, but that's precisely why these strategy games have their fans. The appeal of the game design is managing a very complicated system of interlocking mechanics.

I'm a strategy gamer, and this is too much even for me.

A complex gameplay setup like the one used in the Paradox games may or may not have depth to it. Depth in the gameplay refers to the QUALITY of the overall mechanical and decision-making process, not their QUANTITY. A well-designed game will have depth to it, offering the player tough choices and real alternatives throughout. I'll use the Civilization series as an example of games that typically have offered a great deal of depth in their gameplay mechanics. The Civilization games usually strike a good balance between competing priorities: do you use your production on building wonders or military units? Make friends with your neighbor and trade resouces, or go to war and take their land? Do I win by building the spaceship or by conquering the world? The point is that the game shouldn't just have complicated mechanics; there also needs to be gameplay depth lurking there behind the actual task of managing cities and moving workers. When gameplay depth and mechanical complexity come together, as in a well-designed game like Civilization 4, the results are impressive and can stand the test of time.

However, Civilization is in many ways a poor example, since it's a game that is quite obviously complicated in its design. Newcomers are easily intimidated by the vast array of buildings, wonders, technologies, units, leaders, civilizations, tile improvements, and so on. These are not games that someone will simply pick up and play on a whim. Too often, the general public thinks that this is how game design is supposed to work. Civilization is a big and complicated game, which means that adding more bigness and more complexity will automatically make things better. More stuff = more good. The result is user-designed mods full of excess and bloat, poorly balanced and frequently poorly documented, the exact opposite of good design. Nor are professional game developers immune from this either; eager to fill out the bullet list for the back of the game box, professional developers have crammed all sorts of poor ideas into Civilization expansions: vassal states, corporations, espionage, coastal blockades, and so on. All of these things are classic examples of feature creep, watering down the core gameplay and introducing needless complexity for complexity's sake. Even the professionals can't stop themselves from making this mistake over and over again in their designs!

What readers should realize is that complexity and depth have nothing to do with one another in game design whatsoever. Nothing at all. Adding complexity to a game's mechanics might result in gains in gameplay depth, creating a richer and more satisfying experience. Or it might not. Depends on the situation. Complexity is neither good nor bad... simply complexity. I've mentioned Civilization and the Paradox strategy games as examples of complexity and depth going hand and hand together. However, I can just as easily mention examples of the opposite. Master of Orion 3 is a game infamous for having vast amounts of mechanical complexity without any kind of depth at all.

You would think all of these numbers mean something. They do not.

For those fortunate enough never to play MOO3, the game overwhelmed the player with all sorts of complicated-looking stuff. You had DEAs to assign, ships to design (with something like eight different hull sizes), a tech tree with six different fields, a galactic map with scores of stars and hundreds of planets to colonize, diplomacy with 15 different races in which you could negotiate with multiple different "attitudes" (do you reply kindly or aggressively to their offer? some races like one attitude and some like others!), a galactic senate with voting, four different victory types, and so on. This looked like an amazing game on paper: so much going on! So many choices! So much complexity! And... it was all completely meaningless. The game had no depth at all. None of the choices that the player made had any real effect on what happened. It was entirely possible to turn on auto-colonize, click the next turn button a few hundred times, and find that you had won the game. The AI literally played the game itself, with or without your input. Master of Orion 3 is a clinic on exactly what not to do when crafting a strategy game, with mechanical complexity out the wazoo and no gameplay depth at all.

It's easy to find examples in the opposite direction as well, games with simplistic rules and little to no complexity that nonetheless feature deep and engaging gameplay. Many of our oldest games fall into this category; think of something like chess, where learning how the pieces moves takes all of five minutes but people still spend entire lifetimes learning to master the intricacies of the game. You should also see how seriously some people take a game like competitive Scrabble, wherein something that's viewed as a basic children's game by most becomes elevated to a cutthroat engagement. (Complete with its own cheating scandals!) In the realm of video games, I've spent close to twenty years now playing Tetris, a game that I learned how to play in mere minutes, still without ever feeling as though I've mastered the full intricacies of its gameplay. There is always room to practice and improve, finding ever-more perfect ways to optimize the placement of a series of falling blocks.

Illustrated with amazing Paint skillz.

I can go on, but I hope that by now the point is clear. Complicated game mechanics have nothing at all to do with the depth that lies in the gameplay. It's entirely possible to have a game that's complex + deep (Civilization), complex + shallow (MOO3), simple + deep (chess), or simple + shallow (99% of the free download games on the iPhone). The problem comes from the idealization of that which is complicated, a belief that is rife in the online gaming world. (I can pull out my academic credentials here and write that online gaming has fetishized complexity as a design element, but that's probably getting a bit too obnoxious.) I'll explore this further by means of one of the more interesting feuds in the emerging field of esports: the rivalry between upstart League of Legends and its more established competitors, Defense of the Ancients (DOTA) and Starcraft.

Most people probably do not particularly care about this rivalry. I play and enjoy League of Legends, but without any kind of malice towards fans of DOTA or Starcraft. To each their own, life's too short to waste it hating on other people's entertainment habits. I even love watching competitive Starcraft between top players, truly a fantastic spectacle. That said, many of the fans of each of these games can be completely obnoxious, in the way that only teenagers on the Internet can be, and fans of each respective game keep pushing this whole "feud" aspect, particularly between League and DOTA. There's also a real logic behind this competition, as there are only so many sponsorship dollars out there to go around, and if an esports organization like MLG or IGN makes the decision to carry League of Legends, then they're less likely to carry DOTA as well. This makes the rivalry between the fans of these games somewhat understandable, if no more palatable.

The main source of the controversy has been the meteoric rise in popularity of League of Legends, which went from gaming non-entity in the beginning of 2011 to the most played game in the world by the end of 2012. This development caught everyone off guard, from developer Riot Games (which has been stuck using an outdated client for League due to their own shock at the game's success) to esports organizations, which continued to promote Starcraft and feature it much more prominently in their advertising despite League drawing much higher viewership numbers. I remember watching one of the IPL tournaments in 2012, which gave Starcraft multiple streams and commentated every match with casters, while League was relegated to a single stream and with half of the games completely unwatchable... and that one low-quality League stream had five times the viewers of Starcraft. When League of Legends released its Korean client in early 2012, the number of active players quickly dwarfed the number of Koreans playing Starcraft, something which would have been almost unthinkable just a few years earlier. In a very short span of time, League had pushed aside the older and more established esports games to put itself on the top of the competitive gaming pyramid, with the largest viewership numbers and the biggest cash prizes in the world.

The League of Legends Season Two World Championship.

It was inevitable that this would breed resentment and jealousy from the old guard. But there was something even worse that brought out the bile from Starcraft and DOTA fans everywhere: League of Legends was a much, MUCH less mechanically complicated game than either of its rivals. This is not something that is up for debate. Starcraft asks its players to manage a resource economy, build a defensive base (actually multiple bases), control dozens of units, and use active abilities (e.g. Psi Storm) all at the same time. The mechanical skill required to do this at a high level is almost inhuman; the top players use a dizzying system of hotkeys to micromanage an intense battle while at the same time actively controlling their economy and replenish forces as they perish. I have all the admiration in the world for these players while at the same time readily admitting that I cannot do the same. I learned back in the days of Warcraft 2 that I could never compete with players who had faster reflexes, better clicking skills, and more actions per minute (APM) than I could manage. In League of Legends terms, playing Starcraft is something like controlling all five champions on the map at the same time, plus all the minions, plus all of the towers. Is there any wonder why Starcraft players were (are) frequently so contemptuous of League players? "You only control ONE unit on the map? LOL!"

DOTA is a game much more similar to League of Legends, which makes sense as the original team that worked on the Defense of the Ancients mod for Warcraft 3 now has its designers split between DOTA (Icefrog) and League (Guinsoo). Whether one game is "better" than the other is a source for fanboy debate, and I don't particularly care one way or the other. However, like Starcraft, there's no arguing that DOTA is a much more mechanically complex game than League. The obvious example is the denying mechanic, wherein heroes in DOTA have the ability to kill their own minions to deny gold and experience to the other team. This is hardly the only such example, however. League restricts champions to a universal recall back to base to recover health and buy items, while DOTA allows rapid movement around the map with teleport scrolls, and has flying couriers that can bring items to heroes in the field. League has one place to buy items, while DOTA has multiple stores, including secret hidden shops, located in several different places around the map. Some items can only be bought in certain stores, make sure to remember which ones! League and DOTA both have "jungles" with neutral monsters to kill, but in DOTA it's possible to face double the number of monsters by pulling and stacking the creeps at the proper moment on the clock, something that factors into many farming strategies. It's always the same time of day in League, while DOTA has a day/night cycle that affects visibility and certain hero skill kits. Champions in League always have four skills, while DOTA heroes can have as many as six skills (if not more in some circumstances), and many skills involve the control of pets (or even multiple pets) for very lengthy periods of time. Anyway, hopefully the point of all this is clear. DOTA unquestionably has more mechanical complexity than League of Legends, and this factors into the feud between the fanbases. "How can such a simple noob game be so popular?!" seems to be the common refrain, or something along those lines.

Of course, the reason why League became so popular is precisely BECAUSE it stripped away so much of the complexity of its competitors. League did an impressive job of removing many of the things that people found so tedious or unfun about Starcraft (building workers and managing the economy) or DOTA (very complex champion mechanics and a bizarre shop/item system) only to replace them with more basic, easier to understand systems. Your non-gaming buddy likely can't muster up the skill level needed to play Starcraft, but he/she can probably run around as Teemo and have fun putting down mushrooms and annoying people. (This is why DOTA2, whenever it officially releases, is never going to overcome the League of Legends playerbase. The game is simply too complex. The game that does surpass League eventually will do so by being simpler and easier to play than League. Less complexity, not more.) This is the reason behind League's success, for good or for ill. And for many fans of other esports, they believe it's for ill.

The argument about League of Legends in the esports scene shouldn't be about the mechanical complexity of the game. It's clearly inferior to its competitors in this regard. This means... what, exactly? Does it even matter that there's not as much complexity in League compared to other competitive games? The argument should be instead about the DEPTH of the gameplay, and whether it holds up to close scrutiny. That's a subjective question, obviously, but as someone who has watched an inordinate amount of professional League of Legends teams compete... well, I do think that the game holds up just fine. There are plenty of real strategic dilemmas to solve in terms of team composition, summoner skills choice, level 1 strategies (which are getting more and more intricate all the time), laning setup, jungler route and level of jungler aggression, buff/dragon/baron control, and so on. There's a lot going on when these top teams play, and I would argue that it makes for gameplay every bit as entertaining and deep as what viewers get from watching Starcraft or DOTA. Viewership numbers on tournament streams seem to agree.

League of Legends: a casual game for noobs. (Follow that link, sarcasm intended.)

Think for a moment of how silly the alternative point of view sounds: because League of Legends is a mechanically simple game compared to its competitors, because it has a lower "skill ceiling", that means it is less deserving of attention and makes for an inferior esports experience. This is the most common argument floating around out on the Internet. Does that even make sense? I mean, does making a task more complicated somehow make it more deserving of our attention? If so, then we should all be out creating Rube Goldberg machines. No one has ever tried to apply this logic to the world of sports, for example. Soccer/football/futbol is by far the world's most popular sport. It's also one of the simplest sports in the world too. You can essentially boil the rules down to two things: score the ball through the goal, can't use your hands. The very simplicity behind the sport is the reason for its success, since it doesn't require any kind of equipment whatsoever. All that you need is a field and a ball. Does this somehow make soccer a "less deserving" sport because the game has (mostly) simple rules and can be easily understood by anyone? Or is American football somehow "better" due to its incredibly intricate ruleset and arcane series of instant replay rulings by referees? Obviously that kind of logic is nonsense. The complexity of a sport or game has nothing to do with its entertainment value, or its intrinsic worth as an event - however we would even try to measure that!

In the end, therefore, complexity in esports proves to be just as much of a dead end as complexity in game design. The need to micromanage a flying courier or Fungal Growth a swarm of marines while simultaneously injecting larvae at three bases doesn't make DOTA or Starcraft inherently more interesting or deeper than a game like League of Legends. Nor does stacking endless convoluted game mechanics make for a better game during the design process. What ultimately matters is whether strategic depth and interesting choices can be created during gameplay, and that's something that is irrelevant to the level of complexity in the final result. Sometimes adding more complexity can be a good thing: many people prefer Starcraft or DOTA precisely because they are very complicated games, and that's absolutely fine. Everyone has a different tolerance for how much complexity in game mechanics they're willing to stomach. One person's tedious bloat is another person's engaging masterpiece. But one thing that I hope more people can learn to embrace is that gameplay with depth can be found everywhere, in both the simple and the complex. There's a reason why Super Mario Brothers is still going strong three decades after the series began. Simple, rich, deep 2D platforming greatness will probably never go out of style. Adding a bunch of complexity to that gameplay system would only weaken the overall design. That's the kind of lesson from which we can all learn.