Catching Online Game Cheaters

Apreche

I'm sure you've all seen the recent news about the teacher who caught his class cheating because of the huge difference that cheating made in the test result statistics.

I'm thinking, what if you used those same statistics to catch cheaters in online games? In any game from Pac-Man to Counter-strike you would expect to see a bell curve when looking at the overall performance of players. If you see some weird spike above the bell curve to make two camel humps, then that's probably a sign that there is cheating going on. You can then look for players who are in that hump and figure out which ones are cheating.

If you can succeed at doing this, it will be impossible to circumvent. Right now they try to use technological counter-measures, but they are defeated constantly. This system wouldn't be defeated unless a cheater intentionally played at their own skill level despite cheats. You would also be able to detect people cheating by methods that technology can not find. For example, someone once made a robot arm with a webcam that watched the monitor and played Quake. That could theoretically be caught post-hoc if it performed consistently above average.

Is anyone out there even doing this? Even if you didn't actually ban people out of fear of false positives, it would be really interesting to run the numbers on this for a game like Counter-Strike.

Axel

I think the big issue here is the difference between people who are "cheating" and "exploiting." Someone could find an unintended way of playing the game, some way the programmers didn't design. However, the game naturally allows for this to work. This then makes them superior. They are using no external programs, hardware, nor are they breaking any rules. They simply found something in the game that helps them win, and no one else knew about it, so they used it. This person isn't really cheating, are they? That's how things are in most fighting games, top players use glitches/exploits to play. At least, I know it's that way in Smash Bros.

So, yeah, you could find cheaters just by looking at sudden increases in skill. But false positives would be greater than you think, and what you might wind up defining as a cheat isn't really a cheat.

Jack Draigo

Is anyone out there even doing this? Even if you didn't actually ban people out of fear of false positives, it would be really interesting to run the numbers on this for a game like Counter-Strike.

There are people who don't cheat at Counterstrike?

Omnutia

If you cheat to make yourself rediculously better at a game than anyone else [That was a horrible sentence.], then it's pretty easy to spot. If you just cheat to bring yourself into the top 80%, like being able to see through walls, then this idea falls appart.

MrRoboto

But your system could only detect that there is cheating going on, not who the culprits are.

Rym

There are people whodon'tcheat at Counterstrike?

Honestly, there is basically no cheating going on in CounterStrike anymore unless you play on really crappy servers.

Andrew

http://www.polygon.com/2015/9/15/9311497/how-a-hearthstone-player-broke-the-rules-during-the-games-biggest

Axel

Austbuck should've been DQed. Without knowledge of how he acted in previous matches, and his breaking of a key rule only revealed in the last round of a major tournament, it is faulty to assume that he didn't break the rule just because "there's no proof he used a sideboard deck." No previous players knew his submitted decks, and so had no reason to check. The reason you force them to delete extraneous decks is to guarantee they can't cheat.

It's a failing of the administration to handle it in this way.