That headline doesn't refer to the times when games break and throw up oddball bugs for our amusement, but rather when games throw so many problems at the player that they become a sort of jeopardy-based experience in crisis-juggling. Earlier today I was running through my game collection and thinking about what I might like to play. It wasn't Dishonored. Three things other stood out: Day Z, FTL, and X-Com. I began to think about what those had in common which, and what that said about my enjoyment of this year's immersive masterpiece.
And I realised it was this: peril.
Now you might argue that most games contain peril, and that would be true, but what's interesting about Day Z, FTL, and X-Com is just how badly things can go wrong and the ways in which they can go wrong. Many games offer perhaps one or two ways for things to go wrong: your health can go down, perhaps you'll die. Perhaps your hi-score at the end won't be enough to unlock achievement x. But in the games I've listed here, and many more besides, the structure of their jeopardy means there are multiple vectors for peril.
It's something people who play Roguelikes have been explicitly aware of for years: that the more systems a game has for doing horrible things to your character(s), the more interesting the situation is it possible for them to generate. I remember Kieron writing something like an eight-page article in PC Gamer about a decade ago, with accounts of just what a terrible mess he'd gotten in to, time and again, in a particular roguelike (ZangbandTK – Kieron). What was fascinating was not how successful he'd been, but how he's starved to death just metres from his goal.
From there look to Dwarf Fortress, and the power with which its intricate portrayal of dwarven dooms is imbued by the knowledge of your certain destruction, and the ways in which that might occur. The Dwarf Fortress player is a repository of anecdotal situations of emergent disaster in his subterranean kingdom.
What's thrilling about Day Z is the way in which any encounter can leave you horribly maimed, the world turning pale as your blood drains away, in desperate need of food, medical attention, and even antibiotics. Your chums might be in an even worse state, and need immediate attention to keep them on their feet. The terror of realising you need to find your way into a hospital if you're going to survive, or that you are just too weak to survive another encounter with an enemy, is where Day Z excels. While it's rare to actually starve to death, it's not impossible, and the sheer variety of ways you can end up with flies buzzing around your corpse has turned Day Z into a sort of death-simulation anecdote generator. Looking into Chernarus, in the rain, with your temperature dropping, and gunshots in the distance, is far more thrilling than any power fantasy other games could provide.
What's thrilling about FTL is that time when you were boarded, as the ship was in orbit around a sun, and you ended up venting most of the ship to try and deal with the boarders and the fires, but ended up with no oxygen supply at all. At the end of the battle you face the rest of the campaign with a single surviving crew member, who faces the next three battles on his own, running around the ship fixing systems and hull-breaches, before finally dying to a rebel fighter when he was too slow to get repair the shields. He was a rock man, after all.
What's thrilling about X-Com is coming out on top when so much is against you. Half your men are in the infirmary, half in the grave. The rookies that make up the team seem hopeless, and half the world is in the grip of panic over the alien menace. And yet you still manage to come out on top. The worse things in a game can become, the better it feels when you beat it.
And then I wonder why I don't want to play more Dishonored. And I think the reason comes from Arkane's own admission that the game is a power fantasy. In some ways, games are better when they are – and this is a peculiar-sounding phrase – a vulnerability fantasy.
Interestingly, Dishonored's own designers say that what's exciting about stealth is the feeling of vulnerability. And it's true that you can get overwhelmed by enemies if you've been discovered in Dishonoured. But what doesn't really happen – as happens in the other three games I've mentioned - is that you get put in a far worse situation that you then have to manage. The worst that can happen to Corvo (death aside) is that he gets injured, and/or the guards are alerted. He remains a superhuman killing machine. Most of the features in the game – the various powers – are about killing, rather than managing disaster. And that's what Arkane intended. It's a brilliantly pitched power fantasy – far more subtle than any run and gun. I love it for that. But it doesn't bring me enough peril.
Novelists and scriptwriters have understood this trick for years: piling problems on to the protagonist, in as many different ways as possible, is what makes for a good plot set up. How the protagonist brilliantly resolves or overcomes those problems (or not) will decide how good the story actually is.
The same kind of method can be true in games. The brilliance of FTL is that I can find myself in a situation where have no missiles left, three crew dead, a hull breach, and a fire in the medbay, and that there's still a way out. The satisfying consequence of all this is that I can still come out the other side of it alive, overcoming both my enemies and my problems through my own skill. I can be the brilliant hero-protagonist who times his repair team's race to the breach, with the venting of the air from the airlocks, with the ion-pulse which disables the enemy ship's systems and saves the day.
There's no rule of thumb here, I think, other than to make that point: the more ways things can go wrong in a game, the more ways players can have to be heroes. Multiplicities of jeopardy are not appropriate to all games, but I suspect that if you are a designer who wants a game to be the kind of game that players then go on to tell long, detailed stories about, then creating systems which heap problems on your player from many different directions might be the way to go about it.