A reader writes:
I read with interest your post on the failure of many great books to come up with a satisfying ending. It's hard enough to end a book, but I'd wager that it's even more difficult to end a video game.
Video games are a young medium, but they've always had a strange relationship with their endings. Early arcade games don't end so much as break. If you get far enough into, say, Pac-man, and you're confronted with a "kill screen" where the game no longer functions properly. Your reward for getting through an NES game could be hugely disappointing. For instance, my brother and I spent one Saturday afternoon toiling away at Rampage only for the game to end with "CONGRATULATIONS" and nothing else splashed across the screen. There's a sense that these games could be excused for their lame endings, as they were monstrously difficult to complete and were suffering under crushing limitations of budget and memory.
Today, you've got games with budgets equivalent or greater to that of Hollywood movies. Naturally, expectations have risen.
We'd like some kind of narrative in our games – at least the ones which can support a narrative, mind you. The problem is that a game is not being watched; it's being played. You have to take into account player input, because in a sense, you (as a developer) are sharing storytelling responsibilities with the player. Thus, your ending has to respect the themes your story has set up in the previous hours, but it also has to respect the player's accumulated gameplay experience. It's difficult to convincingly kill off the main character in a first-person shooter, for instance, because the player has seen his avatar carve his way through thousands of bad guys. Oftentimes the solution is to kill off the character in a cutscene, where control is taken away from the player, but this generally comes off as something of a cop out. One of the only games I've seen pull this off is Halo: Reach; your character dies, but does so as a function of actual gameplay after completing his mission.
It's even worse if your game allows some kind of choice system. The most pertinent modern example is the ending to the Mass Effect series of games. The central promise of the series was that choices the player made – which faction to support, which system of morality to endorse, etc. – would end up mattering greatly in the end. I'll spare you the aggrieved nerdrage (Worst. Ending. Ever.), but when the various endings to Mass Effect 3, the final game in the trilogy, ended up being largely identical [illustrated in the above video], the backlash was so severe – the overall critical score of the game is a 93, while the user score is 40 points lower – that the developers released a free add-on essentially adding cosmetic differences to the ending.
The strange thing was that the ending to Mass Effect 2, the previous game, was widely praised – it's No. 19 on this list (Halo: Reach is No. 2). It's as if developers practically luck into making particularly great or horrible endings. There seems to be less thought being put into this than there should be, but given that word-of-mouth can end up costing a game much-needed sales, I'd wager that it's a topic being discussed at most major developers.
Anyway, sorry to nerd out about this, but the election is over so I figure I can indulge in a bit of frippery.
The Dish 