Failure Tolerance in Games

In my last post, I tried to explain why I felt that accepting failure in games could produce more interesting experiences for players. Today I want to describe some methods that can be used to embrace the idea of failure, and make allowances for it both in game mechanics and story writing. I will focus on single-player or co-operative experiences - Competitive multiplayer games are very different beasts.

Videogames distinguish themselves from other media like film and literature in that they allow the audience to fail. More generally they allow the audience to take a highly active role in the outcome of a game's story.

To illustrate, you might look at the story of Mario trying to rescue Peach in Super Mario Bros. Some players may struggle through the game, using many lives, to finally overcome Bowser and rescue Peach. Other players may find all the warp zones and skip past most of the game, beating it with far less effort.

I have never beaten Super Mario Bros.

So different players have different experiences - This is also somewhat true for books and film, too (see previous post). It's pretty obvious when you watch someone play a game and they're far better at it than you. What's important here is that this is one of the fundamental elements of games, and in my opinion some more effort should be invested into diversifying player experience. How games deal with failure is a part of this, and I have some ideas on how to expand some games' tolerance of player failures.
 

Failure as a Story Element

Failure tolerance can be incorporated into videogame stories fairly easily - Simply find a situation in which your game's story could allow for the player to fail in some goal, allow the player to fail at that goal and then branch the story to accept this new element. One problem I can see with this is the additional resource cost for developing such a branch - Every new piece of dialogue, map segment or QA playthrough will inflate the game's development workload, which can be a serious issue. Making a superficial branch path could also strike players as a cop-out.

I'll provide a real example from The Witcher 2 (minor spoiler ahead). Near the end of the first chapter, there's a fight with a particular boss. I remember struggling so much with this fight that I had to load an earlier save and craft better equipment to finally win - you can imagine my frustration when after defeating the boss, a cutscene played in which my character was hurt and the boss escaped. Despite the fact that losing the fight was written into the story, if you ran out of health, you'd have to try again.

"Are you fucking kidding me?" - Geralt of Rivia

If the game had incorporated some story-based failure tolerance in this fight, the post-fight cutscene and some subsequent story elements could be changed:

• If you lost the fight, the regular cutscene and sequence of events would play out, in which the boss's actions form a major element of the game's story.
• If you won, the boss might have been wounded and escaped using a smokebomb. After following the trail of blood left by his wounds, the boss could still perform his story-changing actions, but this time the player might burst into the room just as the boss's story-changing actions take place.

Failure as a Game Mechanic

Counter-Strike's excellent money system allows the wins and losses of previous rounds to have an impact on the state of the game at the beginning of a new round. The winning team earns more money, and players who survive the round get to keep their equipment. This means that a team who loses a number of rounds may not be able to afford the weapons they want, and even a player from the winning team that dies often can find themselves unable to buy another AWP.

Failure tolerance in game mechanics can be difficult to implement, depending on genre. Real-Time Strategy games like the Homeworld series allow players to lose units in smaller battles over the course of a longer war. Homeworld goes a step further than many RTS games, carrying the player's units over to the next mission - This means that player success or failure can have a more long-term impact. 

Homeworld 2 has some fantastic sights and sounds.

On the other hand, FPS games might have a harder time allowing players to fail as part of mechanics. Health systems allow some failure tolerance, but death generally means restarting from a save point. Borderlands puts players into a 'downed' state when they run out of health, allowing them to try and finish an enemy off to rally back to life, which further extends the tolerance of failure. Player death is a frustrating interruption to play, so I think some effort should be made to allow players to mess up a reasonable amount without being killed. I don't really have any good ideas on how to allow for more failure in FPS game mechanics, but I imagine others can. Ultimately the goal would be to reduce the incidence of player death, replacing this 'hard' penalty with a softer one.

The end goal of this effort is to develop games which respond appropriately to a broader range of player actions, which I think is a desirable outcome. In a game that tolerates player failure as part of an uninterrupted play experience, players will remain immersed without frustrating 'game over' screens, and maybe we can increase the amount of people who actually finish games.