Ok, there’s something I want to make clear, as I feel like it has been falling out of favor.
Loss is story.
As a player, it’s natural to want to win. People play games for an myriad of reasons, but I think it’s safe to say that a large amount of people play games to win them. But the thing is, the difference between a game and a puzzle is that you can lose a game. A puzzle, for instance a sudoku or a jigsaw puzzle, you can sit in front of and systematically try every possible solution until you finally succeed. Given enough time and effort there is no way to lose a puzzle. Not so with games. Games tend to require a player to compete, either with other players or with the game itself, and it is possible to reach a section of a game you cannot beat without getting better at the game.
The problem, in my mind, comes from turning a game into a puzzle. Using an example from my own life, I’m guilty of this with regards to the Fire Emblem games. For the unfamiliar, it is a turn based strategy game series chock full of swords, sorcery, and interesting characters. Fire Emblem has a number of interesting facets, but the two that are of interest for this discussion are the support conversations, that unlock additional story and dialogue between units, and the fact that units who are eliminated from a battle are permanently removed from the game. Together, this could make for interesting and poignant story, as characters develop interpersonal relationships and have to deal with the tragedies of war. The problem arises with the trivial ease with which a level can be restarted if things go wrong, to the point where it is more a matter of dedication and patience to beat the game rather than skill. When you can try any action an infinite number of times, the statistics that determine success lose all meaning.
Loss is meaningless if it can be immediately reverted.
This philosophy is already prevalent in most pen and paper role playing games, but it occasionally gets subverted. Let’s look at D&D. In, say, 3.5 D&D, if a character is killed after a certain level, it is actually more likely for them to come back than not. Once the players have access to high level divine spells, they will not stay dead. They just won’t, unless their enemies go significantly out of their way to make it so, either by disintegrating every part of every party member, or killing everyone who might eventually raise them, or any equally outlandish solution. Once a player knows for sure that if they die they’ll come back the next day, what is death but a nap, occasionally paired with XP loss? It serves a story’s tension to have some measure of chance for failure, It serves a character’s humanity to have some fear for their own mortality. Life or Death engagements have little meaning when Death is just time out. By the same token, the story itself should be able to survive the death of a character. I’m not saying there should be no means of reviving a character, if it serves the story you are all telling, but there’s a difference between finding the right horn of the fabled blue minotaur to trade to the queen of the fairies to get her to fulfill the death oath sworn between her court and the father of the PC and going to see Jim’s brother, who’s a decently high level cleric of Pelor and will probably even spot us the gold if we tell him the quest is epic enough.
Video games find it harder to implement this idea, simply because a player who loses will restart, most of the time. That said, players have been trained to accept a win ending and a game loss, or at most a plethora of win endings based upon their morality choices. But what about basing the ending on their qualitative choices? The S.T,A.L.K.E.R. games embrace this idea to an extent, offering the player a number of epilogue scenes depicting the consequences of their actions, ones they may not have thought twice about. Fallout games also generally do this, although the game itself generally makes it clear which actions are choices that will effect the course of the game. The trick is to give the player choices during game play, and not show the effects until after the game is finished. A game wherein you fight the final battle and that decides whether you win or lose can have an epic finish, but a game in which the entire shape of the final battle is determined by the outcome of a hundred skirmishes you breezed through earlier is a game whose story the player has an impact on, and vice versa. It’s also a where the player will be incapable of simply skipping back a checkpoint and saving that one NPC who matters to the story.
It’s worth noting that this advice can easily be applied to traditional writing as well, as writers should not be afraid to kill characters, break objects, or destroy places that haven’t necessarily served their narrative purpose if it serves the story. It’s much more jarring to lose a character in the middle of his quest than once he’s finished all his relevant story functions.
So remember, plot armor is your enemy, and make your story strong enough to survive breaking.