Recently, a team of psychologists led by Niklas Ravaja at MIND Labs in Helsinki, Finland, decided to study precisely what sort of emotions people experience while playing games. So they took a bunch of gamers in their 20s and had them play Super Monkey Ball 2 bowling, competing amongst each other (the top scorer won free movie tickets). While they played, the gamers were wired up to a bunch of biosensors — including skin-conductance meters, cardiac monitors and facial electromyographs. Psychologists have long found that by detecting spikes in one’s physiological activity, they can pinpoint the precise moment you find something fun or frustrating.

As the subjects played Monkey Ball, their pleasure spiked upward when they knocked down a lot of pins. On the other hand, if the ball closely missed the pins and landed in the gutter at the end of the lane, it produced frustration. This is pretty much what you’d expect.

But then something odd happened: When the players aimed really poorly and sent the ball zooming off the edge into space, their brains didn’t register frustration. They registered pleasure. “Although the event in question represents a clear failure,” the researchers wrote, “several physiological indices showed that it elicited positively valenced high-arousal emotion (i.e., joy), rather than disappointment.”

Sucking, it seems, can be fun.

This is a totally counterintuitive result. Gamers are utterly obsessed with success — who’s l33t, who’s the suXX0r. Indeed, much like the Inuit with their 40 different words for snow, gamers have created a sprawling lexicon of slang designed to quantify — with surgical precision — exactly how much you suck or rock. (Dude, I totally pwned that n00b llama in pvp!) In theory, totally failing at a game ought to bring nothing but the sting of defeat, right?

Sure, except in one case: When you’re playing a game so well-designed that it is delightful even when you screw up.

That’s the secret behind Super Monkey Ball 2. The lose condition — spilling off into space — is unusually hilarious: a monkey plunging through infinite Euclidean freefall, his scream slowly vanishing into oblivion. As I bowled gutter ball after gutter ball, sure, I was cursing myself — but I was laughing, too.

And when I think about it, some of my favorite gaming experiences have occurred not when I’ve been triumphant and victorious, but when the game has gone horribly awry, and the ensuing chaos is a hoot. At one point while playing Grand Theft Auto: San Andreas, I botched a drive-by shooting, and in my panic to escape, accidentally ran over a police officer. As I tore off down the highway with about 20 cop cars wailing behind me and a police helicopter hovering overhead and my engine on fire — fiddling with the radio to try and find a suitable soundtrack for what was rapidly turning into a suicide drive — I knew I was doomed. But it was still a blast.

Indeed, some games have elegantly transformed their “fail” conditions into core game mechanics. When the first version of the Burnout racing game emerged, players discovered it was totally fun to watch their crashes in slo-mo. So the designers switched course — and in the next few versions, made crashing a central part of the play. When you spin in Burnout Revenge, you can jump into bullet-time mode and try and steer your tumbling car into enemies. They harnessed the classic failure psychology of the kamikaze: If I’m going down, I’m taking you with me.

Mind you, there are limits to this approach. There’s a fine line between sucking that’s pleasurable and sucking that just, well, sucks.

When I’m playing a mission-based game that has irregular save points — like, say, Far Cry Instincts — and it forces me to redo a difficult quest over and over and over and over again every time I screw up, that’s not fun. That’s an affront to decent people of all nations. Same goes for games where camera angles are so lousy you lose track of a battle (like Jaws) or where the puzzles are meaninglessly obtuse (like the third and fourth Tomb Raider games). It’s only fun to fail if the game is fair — and you had every chance of success.

So clearly this isn’t easy stuff to design. But the genius of Ravaja’s research is that he suggests it is possible — and indeed ought to be a goal for every game designer. (And as he proved, it’s even empirically testable: Slap the wetware on your beta testers and measure how joyful they find it when they go down in flames.)

So consider this the new way to tell when a game is truly great: It’s fun even for the l@merz.