Player Death in Games: Towards a Critical Vocabulary For Assessing Difficulty (Part I)

In Part I of this article, game designer Ian Schreiber proposes a methodology to classify and measure what is "death" in games.

Commonly in games, failure is framed as “dying.” This is a little misleading for a few reasons. First, obviously the player can start the game again, so it is not permanent death in the same sense as dying in the real world. Second, “dying” is a story element; in some games you die, in others you are “defeated” or “incapacitated” or you “retire” or you “fall down” or any number of other things. Mechanically, there may or may not be a difference between these, so singling out “death” from the others is incorrectly treating the back story as a game mechanic. Third, there are many different ways “dying” can be handled in a game. We need a better way to classify what “death” is and how it works.

Dying is a setback as a direct result of player failure

I think it is more constructive to think in terms of “setbacks” than “death” when thinking of penalties for failure. If a player must redo an entire level, does it matter if the cause was that they “died” at the end or that they need to repeat the level to gain extra resources that they lost due to their carelessness later on? In either case the end result is the same: replay of the level.

Is this different from if the game forced the player to replay the same level later in the game, for example as part of a special quest? Yes, for two reasons. First, if the player must cover old ground, they are still progressing through the game, so there is a feeling of advancement rather than setback.

How about a game where, as part of a story event, the player is forced to lose something permanently? In many RPGs, for example, the player will permanently lose a party member, or is robbed of their gold and possessions. This is certainly a setback that takes some time to rebuild; why does it not feel like “dying”? The key difference here is that the event is caused by the game itself, and not as a result of the player failing to perform a task.

The concept of “dying” in a game, then, requires two elements:

• The player must have a setback, either in terms of time or resources.
• The setback must be a direct result of the player’s failure in the game.
  

Dimensions of death

Dying is worse in some games than others. In old-school arcade games, you had to start over from the beginning. Later on you were given the chance to “continue” from the last level that you reached. Some games allow you to save (either anywhere, or at predetermined save points) and dying simply knocks you back to your last save. In other games there is a penalty for dying, such as loss of cash or equipment, but otherwise no lost time (other than the time it takes to regain what you lost).

In some of these games, death is worse than others. How do we compare the severity of dying in one game versus another?

I think death has two dimensions that can be measured:

• Probability of dying
• Intensity of setback when a death event occurs
  

Probability is important. If a game is so easy that one never dies, the level of setback is irrelevant. On the other hand, if a game involves the player dying constantly, a slight change to the penalty for dying will affect the player experience a great deal.

How can probability be measured? You could express it mathematically as either a percentage of attempts to overcome an obstacle that end in failure, or an average number of failures required to succeed, or even the mean time between failures. If hard data is not available, it can generally be categorized as high or low based on the overall perception of difficulty in the game.

How can the intensity of a death event be measured? The easiest way to express this is in terms of lost time, i.e. how much additional time must the player spend in the game to get back to an equivalent (if not identical) position before the death occurred.

By multiplying these two things together – the probability of an event and the severity of the event – we can get a measure of just how much of a threat players will feel from failure/death.

Ian Schreiber has been making games professionally since 2000, first as a programmer and then as a game designer. He currently teaches game design classes for Savannah College of Art and Design and Columbus State Community College. He has worked on five shipped games and hundreds of shipped students. You can learn more about Ian at his blog, Teaching Game Design, and website.