GD Column 25: When Choice is Bad

The following was published in the May 2013 issue of Game Developer magazine…

Nothing defines video games more than the importance of player choice. Interactivity is what separates games from static arts like film and literature, and when critics accuse a digital experience like Dear Esther of being “not really a game,” it is usually from a lack of meaningful player choice.

However, because choice is held up as such an ideal among game designers – armed with phrases like “enabling player agency” and “abdicating authorship” – the downside of choice is often ignored during development, hiding in a designer’s blind spot. In fact, every time a designer adds more choice to a game, a tradeoff is being made.

The game gains a degree of player engagement as a result of the new option but at the cost of something else. These costs can commonly be group together as either too much time, too much complexity, or too much repetition – all of which can far outweigh the positive qualities of the extra choice.

Too Much Time

If games can be reduced to a simple equation, a possible equivalence would be (total fun) = (meaningful decisions) / (time played). In other words, for two games with similar levels of player choice, the one which takes less actual time to play will be more fun. Of course, usually the comparison will not be so obvious; a new feature will add a meaningful decision, but is it worth the extra time added to the play session?

As an example, Dice Wars and Risk are similar games of territorial conquest which answer this question differently. In both games, players attack each other by rolling dice, and the victors are rewarded with extra armies at the start of their next turn. In Risk, the player decides where to place these armies, which can be a meaningful decision depending on the current situation. In Dice Wars, however, the armies are placed randomly by the game, and the result is a much faster game.

Which design is right? While the answer is subjective, the relevant question to ask is whether the combat decisions become more meaningful if the player takes the time arrange her new armies – or, as is likely, how much more meaningful they become. After all, the player can pursue a more intentional strategy in Risk, but is that aspect worth the not insignificant extra time taken by the army placement phase?

The answer may depend on the audience (Dice Wars is a casual Flash game while Risk is a traditional board game), but the designers should understand the ramifications of their decisions. Sometimes, army placement in Risk can be a rote decision, and sometimes, reacting to an unexpected arrangement in Dice Wars can lead to a new, more dynamic type of fun. Ultimately, the aspects of Risk which lengthen the play session must justify the time they cost to the audience.

Too Much Complexity

Besides its cost in time, each choice presented to the player also carries a cognitive load in added complexity that must be weighed in the balance. More options mean more indecision; deciding between researching five different technologies feels much different than choosing between fifty. Players worry not just about what they are choosing but also about what they are not choosing, and the more options they decline, the more reason there is to worry.

Each type of game has a sweet spot for the number of options that keep play manageable, enough to be an interesting decision but not too many to overwhelm the player. Blizzard RTS’s have maintained a constant number of units per race for decades; StarCraft, Warcraft 3, and StarCraft 2 all averaged 12 units per faction. For the third game, the designers explicitly stated that they removed old units to make room for new ones.

Indeed, RTS games as a genre are under assault from their more popular upstart progeny, the MOBA genre, best exemplified by League of Legends and Dota 2. The original MOBA was a Warcraft 3 mod entitled Defense of the Ancients, which played out like an RTS except that the player only controlled a single hero instead of an entire army.

This twist broadened the potential audience by radically reducing the complexity and, thus, the cognitive demands placed on the player. Instead of needing to manage a vast collection of mines and barracks and peons and soldiers, as in a typical RTS, the player only needed to worry about a single character. Consider the UI simplifications made possible by allowing the camera to lock onto the player’s hero instead of roaming freely across the map, which forced the player to make stressful decisions about managing his attention.

Of course, this change did take away many of the meaningful choices found in an RTS. Players no longer decide where to place buildings or what technologies to research or what units to train or even where to send them; all these choices were either abstracted away or managed by the game instead. Again, the relevant question is whether these lost decisions were worth the massive amount of complexity they added to a typical RTS.

The success of MOBA’s demonstrate that although players enjoy the thrill and spectacle of the large-scale real-time battles pioneered by RTS games, they do not necessarily enjoy the intense demands of trying to control every aspect of the game. Designer Cliff Harris discussed a similar point for his successful alt-RTS Gratuitous Space Battles, which does not allow the player any control of units during combat: “GSB does not pretend you can control 300 starships in a complex battle. It admits you can’t, and thus doesn’t make it an option. Some people hate it. Over 100,000 enjoyed it enough to buy it, so I can’t be the only person with this point of view.“

Too Much Repetition

The final way that too much player choice can negatively affect the game experience is perhaps a bit surprising, but games with too much freedom can suffer from becoming repetitive. A typical example is when a game presents the player with an extensive but ultimately static menu of choices session after session; players often develop a set of favorite choices and get stuck in that small corner of the game space.

Sometimes, a fixed set of options can work if the player needs to react to a variety of environments; the random maps in a Civilization game can prod the player down different parts of the technology tree. However, almost all games could probably benefit from reducing some player choice to increase overall variety.

Consider Atom Zombie Smasher, a game in which players use up to three special weapons (such as snipers or mortars or blockades) to help rescue civilians from a city overrun by zombies. However, these three weapons are randomly chosen before each mission from a set of eight, which means the player reacts as much to the current selection of weapons as to the city layout or zombie behavior. Instead of relying on a particular favorite combination, the player must learn to make unusual combinations work, which means the gameplay is constantly shifting.

Similarly, in FTL, the crew members and weapons and upgrades available change from game to game, depending on what the randomly generated shops provide. Thus, the game is not about discovering and perfecting a single strategy but about finding the best path based on the tools available. Put simply, the variety of gameplay in Atom Zombie Smasher and FTL emerges because the designers limited player choice.

At the opposite end of the spectrum, games with hefty customization systems usually devolve into a few ideal choices, robbing the flexible systems of their relevance. In Alpha Centauri, players used the Unit Workshop to create units with different values and abilities. However, the most effective combinations soon became obvious, marginalizing this feature.

Thus, giving the player too much control – with too many options and too much agency – can reduce a game’s replayability. Indeed, would Diablo be more or less fun if players couldn’t actually choose their skills? The game would certainly feel different as the loss of intentional progression would turn off many veterans, but the new variety might attract others looking for a more dynamic experience. Randomly distributed skills might force players to explore sections of the tree they would have never experienced otherwise. The important fact is that this loss of meaningful player choice would not necessarily hurt the game.

Ultimately, game design is a series of tradeoffs, and designers should recognize that choice itself is just one more factor that must be balanced with everything else. Even though player control is core to the power of games, it does not necessarily trump all the other factors, such as brevity, elegance, and variety.

5 thoughts on “GD Column 25: When Choice is Bad

  1. There’s no such thing as emergent complexity. Whatever complexity a game has it is inherent to its rules, it was always there. Basically your statement is nonsensical, and so is the difference you’re alluding.

  2. Pingback: Clippings — 31 July 2013 » Matchsticks for my Eyes

  3. There may be no such thing in reality as emergent complexity, but it’s still a great way for designers to assess their rulesets and see how much depth they have.

  4. Pingback: 翻訳記事:選択が善ならざる時 | スパ帝国

Leave a Reply

Your email address will not be published. Required fields are marked *