Lessons Of The Mario Kart CheatersS

The more people play Mario Kart online with their friends, the less often they cheat — at least under their own name. That's the finding of an expert on the psychology of games and the group tracking online Mario Kart.

Last week I spoke to Jamie Madigan, a blogger and gamer with a PhD in psychology who teamed up with his former employers at GameSpy to crunch some numbers that would help explain why people cheat when they play games online.

Madigan had a theory, borne out by some data, that people are more likely to cheat when they are anonymous and that they'll cheat less the more they are connected to other gamers — the more, in other words, that they are known.

GameSpy handles the networking for the online play of Mario Kart Wii and provided Madigan data on the number of cheaters, people who didn't exploit in-game tricks like snaking but glitches that send erroneous data about race completion times. He found that, by far, those who had no friends tied to their accounts cheated the most.

Here's the chart Madigan presented at the Login conference in Seattle displaying those results:

Lessons Of The Mario Kart CheatersS

"There's a couple of possible explanations for something like that," Madigan told me. "One is that people who have more friends or who play more games with their friends cheat less. The other possible explanation is that people use an alternative account or profile when they cheat and they have fewer friends associated with that profile because it's not their true one."

This is logical, of course. People tend to be jerks to strangers more readily than they are to their friends or family. At least, that's the theory, right? "When people are in groups, when they are anonymous and can protect themselves from reprisal, they tend to engage in more antisocial behavior," Madigan said. "In the context of gaming, that means cheating, exploiting glitches and being antisocial."

Gamespy's Sean Flinn, who helped supply Madigan the data, said the findings were as expected but believes they still help support some compelling theories. In an e-mail to Kotaku, he wrote: "What we found was pretty strong evidence in favor of our initial hypotheses: cheaters tend to have lower 'friend counts (#s of game-based friends) than non-cheaters; incidents of cheating tend to occur among players who have low #s of – or zero – buddies. The incidents of cheating fall off rapidly as friend counts rise. Moreover, the people who DO cheat tend to do so quite frequently … not only do they cheat, but they cheat a bunch."

One of the most destructive occurrences in Modern Warfare 2 was the notorious "javelin glitch" a multiplayer match-ruining exploit that compelled Madigan to re-work a famous psychological conundrum and name it the Glitcher's Dilemma. Simply: "For any given normal person they have the choice of: 'Glitch to dominate the map' or 'Don't glitch and have a more fair outcome.' But they're faced with a dilemma because the other players have the same choice. If player A decides to play fair and not glitch, and player B decides to play unfairly and use the glitch, then player A is dominated in the game."

The original, classic example of the dilemma is called the Prisoner's Dilemma and involves two people, interrogated separately, given the chance to betray their partner in crime or stay silent and hope their partner does the same. "The logical, rational thing to do is to do the antisocial thing and either squeal on your partner in crime — or in the glitcher's example, to abuse the glitch," Madigan said.

Lessons Of The Mario Kart CheatersS

Ideally, though, a gamer wouldn't want to have to take that glitching path, which can be considered by fans of fair play, the low road. Madigan believes there are ways out of this: "You can defuse those dilemmas by having things like players who know each other and are friends, or players who know each other outside of that game, or players who expect to play more than one match in a row with that person."

Gamespy and Madigan both advocate an online culture of connected friends, one in which you play with less anonymity and feel more of the social pressure to behave and treat your peers as you'd have them treat you. "The more socially you play when you play online, the better the experience that you're likely to have," Flinn said.

Consider how much better things would go for you, the gamer, if you knew who you were playing with and if they knew a little something about you, Madigan reasons. "When you have situations where anonymity is reduced, where people expect to have a reputation on the line, people tend to not choose the antisocial option. They tend to cooperate. Or at least on the first round they'll say I'll cooperate and see what you do, and if you cooperate as well, then we're good to go. But if you abuse the glitch or lie, on the next round I'm going to pay you back for that." Doesn't that sound good?

For more of Madigan's work, check out his The Psychology of Games, including a full run of his slides based on the GameSpy data, as presented at last week's Login Conference.