Friday, December 10, 2010

Brief definition of "game designer"

I have been trying to write a description/definition of "game designers" in 50 words or less. This is my latest:

“A game designer conceives the framework for a series of interesting challenges in the form of a ‘game’, devises mechanics (rules), creates or helps create a working prototype, and repetitively and incrementally modifies the design in the light of playtesting until it is a good game for the target audience.”

(Update: I've substituted "design" for "game" because the latter implies that the designer will create/produce the actual changes in the game, which is unlikely to be the case for a video game: someone else will actually make the modification.)

Wednesday, December 08, 2010

Game Playing Styles

(This was originally published on, 25 Jan 10. You can click on the post title to go to that version.

Some Game Playing Styles, and How Games Match One Style or Another
Lewis Pulsipher

A big obstacle for beginning game designers is the common assumption that everyone likes the same kinds of games, and plays the same way, that they do. If they love shooters, they think EVERYone loves shooters. If they like strategic games, they assume EVERYone likes them. If they love puzzles, they suppose EVERYone does. They may say they understand the diversity, but emotionally they don’t.

Sometimes the nature of the traditional video game, a kind of interactive puzzle or interactive movie for one person, obscures all the different things games can be. Today I’m going to rely on 50 years of playing games of all kinds to describe several quite different points of view. Since some aspects of these points of view depend heavily on having several human or human-like opponents, many of the examples will be from tabletop games.

The first, of course, is that some people, especially many video gamers, prefer interactive puzzle “games” that have no human/psychological component, while other people strongly prefer games involving two or more people in opposition. In fact, “multiplayer” in the tabletop game hobby doesn’t mean “more than one player”, it means “more than two, and more than two sides”. A two-player game provides some human/psychological interaction, but it’s the more-than-two-sided games where the human element, not the puzzle-like challenges set by the video game designer, becomes paramount.

Classical and Romantic
A second difference that I’ll describe in much more detail has been called the “Classical” vs. the “Romantic”, following philosophers who have discussed this difference in a variety of contexts (e.g., Nietzsche’s Apollonian and Dionysian). A more modern term for the Classical player is “mini-max”, someone who tries to maximize his minimum gain (or minimize maximum loss) in every situation—the “perfect player” of mathematical game theory, if I recall correctly. In game theory terms this player seeks the “strategy that would guarantee the highest minimal expected outcome regardless of the strategy of the opponent.” (Wikipedia)

The Classical player tries to know each game inside-out. He wants to learn the best counter to every move his opponent(s) might make. He takes nothing for granted, paying attention to little details which probably won’t matter but which in certain cases could be important. The Classical player does not avoid taking chances, but he carefully calculates the consequences of his risks. He dislikes unnecessary risks. He prefers a slow but steady certain win to a quick but only probable win. He tries not to be overcautious, however, for fear of becoming predictable. A cliche among football fans is that the best teams win by making fewer mistakes, letting the other team beat itself. So it is with the Classical gamer, who concentrates on eliminating errors rather than on discovering brilliant coups.

The idea of managing risk doesn’t lend itself to single-player video games that have just one solution. In some of these games that involve no chance element (everything is set by the designer), something like game theory calculations of the “perfect strategy” don’t come into play. There is what is called a “saddle point” or dominant strategy, a perfect way to play that will win every time. If you make the right moves in, say, arcade Pac-Man, you will go all the way through all 255 levels every time without a single death, because there is no random element. (See Inside Pac-Man, On the other hand, if the single-player game includes randomness that changes with each play, the player must manage risk, and the game becomes quite Classical. In general, single-player games are going to tend toward the Classical unless the “opposition” approaches a human in complexity.

The Romantic looks for the decisive blow which will cripple his enemy, psychologically if not physically on the playing arena. He wishes to convince his opponent(s) of the inevitability of their defeat; in some cases a player with a still tenable position will resign the game to his Romantic opponent when he has been beaten psychologically. The Romantic is willing to take a dangerous risk in order to disrupt enemy plans and throw the game into a line of play his opponent is unfamiliar with. He looks for opportunities for a big gain, rather than to maximize his minimum gain. A flamboyant, but only probable, win is his goal. He may make mistakes, but he hopes to seize victory rather than wait for the enemy to make mistakes. The Romantic is more likely to try to “get into the head” of his opponent, to divine which strategy the opponent will use and play his own strategy that best counteracts it.

In the standard single player interactive puzzle video game, there is no human opponent to “psyche out” or to fool. Yet some of the more sophisticated modern games are designed to provide a “computer opponent” that behaves in some ways like a human, and clever players figure out ways to take advantage of the programming to “fool” the opponent. When playing a multi-sided game such as Civilization or Warcraft III against several computer opponents you can find ways to “make the opposition look foolish”: in fact, this may be easier than when playing against good human opponents. A political victory in Civilization, in effect persuading the computer players to give up, can be seen as a Romantic goal. Further, real-time games tend toward the Romantic simply because there isn’t time for the Classical player to make careful calculations. Under great time-stress some people will still try to play Classically, it will simply be harder for them to do so effectively.

In the single-player video game with no chance element, the Romantic very likely has no opportunity to “take the path less trodden” in order to fool the computer.

Here’s a simple comparison of these two types of players. The Classical player, in Tic-Tac-Toe, will always play to the center square when playing second if his opponent doesn’t take it—and will always take the center if he moves first. The Romantic may try to fool his opponent into playing badly by making a less-than-optimal play, in order to try for a win rather than accept the otherwise-inevitable draw.

To further generalize, playing against the computer tends to encourage the Classical, playing against people tends to encourage the Romantic. However, when the stress of limited time is introduced, it becomes difficult or impossible to play Classically as you have less and less time to calculate risks.

Many good players depend on intuition rather than study and logic to make good moves, yet the moves can be either Classical or Romantic. A Romantic player can also be a very cerebral or intellectual player who happens to prefer the Romantic style. Nonetheless, the Classical player tends to use logic while the Romantic tends to use intuition. Some people would refer to Classical players with derision as “mathematical” players. It is true that Classical players are concerned with odds and expected losses and saddle points (though this alone doesn’t identify or qualify a person as a Classical player). Nonetheless, Classical players do quite well in non-mathematical games.

Games sometimes tend to favor one playing style over the other. Chess is clearly a Classical game. Single-player video games are often Classical. Poker tends to favor Romantic play, because so much depends on bluffing. Most shooters (the frenetic kind) are Romantic, while stealth shooters tend to be Classical, as far as you can categorize single-player games. A game like two player Street Fighter can be played either way. It seems that the very best players, though, play Street Fighter Romantically, somehow reading their opponent’s intentions and beating them to the punch, the ultimate in playing the opponent rather than playing the game. For more about this see David Sirlin’s book, Playing to Win: Becoming the Champion ( ) (

Diplomacy, though without any overt chance factor, is a good game for both Classical and Romantic players. The negotiations and alliance structures give both types plenty to work with. The Classical player tends to be better at tactics and strategy; he prefers long alliances to continuous free-for-all, for there are too many risks and incalculable factors inherent in a fluid situation. The Romantic tends to prefer the fluid state, and his big weapon is the backstab.

It’s hard to say whether an extreme form of Classical play, in a typical one-player video game, would involve rare resort to reloading a saved game, or would involve frequent saves and attempts at all kinds of different tactics to find out which one is best. I tend to be a Classical player, and I prefer the former, but I’m not going to make the mistake of assuming I’m typical!

While “Minimaxers” are usually Classical players, I have known gamers who apply minimax methods to characters or unit mixes, to more or less tactical concerns, but play the overall game Romantically. “Yomi” is David Sirlin’s term for reading the opponent’s mind; the best Romantic players probably have “Yomi”, but this is not necessarily so, and it’s possible that a Classical player may be able to read opposing intentions but still relies on attaining the minimum maximum gain.

Nonetheless, you’d expect most Classical players to be mimimaxers, and most Romantic players to rely on Yomi.

Reaction to Chaos and Randomness
But this is only one way of looking at game playing styles. The third and last for this article, is to look at a player’s reaction to fluidity and randomness. I’ll call the three points of view:
• the “Planner”,
• the “Adapter” (who tends to represent the middle ground) and
• the “Improviser”

The Planner likes to plan ahead-well ahead. He loves it when things he did long ago in a game come together to give him a big success. He is likely, though not certainly, going to prefer a game where much if not all of the information is always available, e.g. chess. He’s likely to prefer turn-based rather than real-time games. When it’s time for him to make a play, to execute a strategy, he doesn’t want to find that the game has changed drastically owing to a recent move by someone else, or because of the nature of the game itself. The Planner will often be a Classical player as well, though this is not necessary.

The “Improviser” does not like to plan ahead. He wants to react to circumstances at the time he makes his play, and he doesn’t mind at all if circumstances change drastically between one play and the next, or in a short time (in a real-time game). Games with limited information availability aren’t going to bother him, while games with perfect information aren’t likely to be attractive. Such players tend to be Romantic, obviously.

The “Adapter” likes to impose order on chaos, he wants to be able to see ahead a couple moves (or a short while in real-time) and then adapt to them, that is, arrange to “take control” of what’s going on. As you can see, this falls somewhere between the other two.

Once again, some games favor one of the three styles or another. Team video games, if the team actually tries to plan and work together, can be for Adapters. Real-time strategy games may attract Adapters, who can plan ahead some, having gained some information about what’s going on. Two multi-sided boardgames that fit the “Adapter” mindset are Vinci and RoboRally. Vinci is a game with perfect information, and with little overt chance, yet you can’t plan far ahead because the rise and fall of empires and selection of new empire capabilities results in great changes on the Europe-like board in a few turns. RoboRally requires players to program movements of their Robot in a violent race through several checkpoints in a bizarrely-dangerous factory. Each player is dealt nine movement cards, and must lay five face down to be executed in order one at a time. You can plan a route, but you won’t always get the cards you need. Chaos sometimes results from player mistakes, yours and mistakes of others.

Civilization (board or video) tends to be a game for the Planner. Card games tend to be for the Improvisers, though some can favor the Adapter. Poker is a game for Improvisers, except that there can be long-term bluffing plans that are characteristic of a Planner.

Diplomacy could attract Planners, Adapters, or Improvisers, depending on how it’s played.

In Tetris, if you’re just reacting to each shape as it appears, you’re playing as an Improvisor; if you’re trying to calculate which shapes will go well, so that you’ll know where to put one when it shows up, you’re playing more as an Adaptor. Because of the time stress and uncertainty about what will appear soon, it’s hard to play Tetris as a Planner.

Because arcade Pac-Man is ultimately predictable, a Planner may have been the first to notice the patterns and find ways to take advantage of them. Insofar as video games tend to conceal a lot of information, they’re not fruitful ground for a Planner, rather encouraging Improvisation.

Platformers reward short-range planning of the kind common amongst Adaptors. Some RTS games (the ones that are short on time-stress and long on strategy) are good for Adaptors. Survival Horror games with limited ammunition available are good for Adaptors. But something like Left4Dead, with practically unlimited ammo and a Director that increases the challenge as necessary, fits an Improviser point of view.

Depending on circumstances, a Planner or Adaptor should be a good leader in a team deathmatch or capture the flag using maps that are well-known.

Race games can favor any type depending on how much information is known to players when the race begins.

Role of Chance
People might tend to assume that these playing styles are closely related to the role of chance in the game. But it’s not a matter of “how many dice rolls”. Some chance can be managed. Tabletop or video Dungeons and Dragons, on the face of it, is full of dice rolls or equivalent, but a player can do his best to minimize the number of times he must rely on dice to save his bacon, or he can “go with the flow” and rely on the dice.

If there are few dice rolls or equivalent, and some are very important while many are not, then chance is very hard to manage. Randomness is largely unmanageable chance. The Planner doesn’t like randomness, while the Improvisor won’t mind at all. Adapters like some fluidity as a result of what other players do, but don’t much like randomness. Classical players tend to hate randomness, while Romantics may welcome it.

In general, games that provide difficulty by requiring quick reactions tend to favor the Improvisor style and make Planning difficult. You don’t have time to plan a lot in Halo or Combat Arms; you can in the “stealth” shooters such as many Red Storm games like Rainbow Six. Real-time games tend to be better for Improvisors, turn-based games for Planners. Games with most information hidden from the players make Improvising much easier than Planning, hence the AAA video games that usually use “fog of war” (hidden information, even the map is hidden to begin with) tend to be games for Improvisors more than Planners.

In other words, “traditional” one-player video games tend to favor the Improvisor rather than the Planner. But this will gradually change over time: as the market for video games continues to expand, many new players will dislike being time-challenged, they’ll want to relax while they play their games, they’ll want to play a little bit (one turn) at a time. The trend is already obvious in casual games.

These are only three spectra of game-playing styles, out of many. For example, I know someone whose main pleasure in playing games is in helping someone else win! I suspect this is such a tiny minority view that designers need not worry about it—though cooperative games have become quite popular this year--but it helps illustrate how many different “favorite ways to play” exist among game players.

(Parts of this were originally published in Dragon magazine, September 1982, and in revised form in The Games Journal, February 2005, and revised again on GameCareerGuide, 26 November 2009.)