Many Players Will Just Have to Get Better at Counter-Strike
Cheaters – Nobody likes them, not even cheaters, which is why Valve has developed a remedy for dealing with more of them on Counter-Strike: Global Offensive. Considering the popularity of CS: GO, the game witnesses hackers constantly, but it looks like the number of ne’er-do-wells will decrease, thanks to Valve’s new deep-learning algorithm.
Thus far, Valve has relied on a self-regulated CS: GO community to weed out the cheaters. ‘Overwatch,’ as it is called, uses active members called Investigators to identify occasions of cheating or general misconduct. Unfortunately, not every human can spot the moments of hacking. According to Beebom, the conviction rate for cheaters is around 15-30%. And so we have Valve’s new deep-learning neural network designed to cross-examine all variable events.
According to Valve programmer John McDonald, the new network was trained using aimbots. As you can imagine, aimbots have distinguishable patterns whose recognition will inevitably lead to faster bans. In conjunction with VACnet, the algorithm apparently increased conviction rates to 99%. The fortunate result will be a more enjoyable Counter-Strike. Better yet, Valve wants to make the neural network available to all competitive games on Steam.
“We think that it is really helping us get developers off of the treadmill without impacting our customers in any way. Our customers are seeing fewer cheaters today than they have been, and the conversation around cheating has died down tremendously compared to where it was before we started this work,” said McDonald.
Time will tell how Valve’s deep-learning neural network will affect the games industry at large. Stay tuned for updates.