AI Anti-Cheat coming this year

Nerror

With the recent talk of how widespread cheating is, I thought this might be of interest. This can actually combat the type of cheat that is based on an AI "watching" the livestream of the player playing the game, and sending inputs to whatever controller is used. A cheat that current anti-cheats can't combat.

https://www.youtube.com/watch?v=LkmIItTrQP4

In the video they mainly discuss FPS games, but I see no reason why the AI can't make a biological profile for MMO gameplay as well. It seems universal. That goes for aim-botting and ESP and basically anything else involving the player moving around.

If the AI can detect botting in its various forms, including farming, which it seems like it can be configured to, it'll be an amazing tool for the live GM team. Intrepid would be much less reliant on player reports to detect them.

The AI anti-cheat is going to be implemented in big games this year, so there is time for it to mature a bit for Ashes release. I hope Intrepid takes a look at it at least, and determine if something like that is worthwhile. With over 99% accuracy, it looks really promising

NiKr

I'd still prefer this to be report based, just because tracking 10k+ players on every server at all times would probably require a ton of resources and might introduce lag or something similar to the game.

But this definitely looks promising and could definitely make GMs' lives way easier.

Arya_Yeshe

Good video, but for shooters, using anti-cheat will cost money.... in the other hand letting people cheat and ban multiple accounts will reward the company with a lot of money.

In WoW, I saw so many people multicasting multiple druids I can't even imagine how much money Blizzard made out of this, but if Blizzard went after them then it would cost money

I simply left a few games because cheaters ruined the game for me, that's what players should do, they should put the companies on their knees by leaving the game, then maybe devs will work on this... it's not hard

Noaani

Arya_Yeshe wrote: »

Good video, but for shooters, using anti-cheat will cost money.... in the other hand letting people cheat and ban multiple accounts will reward the company with a lot of money.

In WoW, I saw so many people multicasting multiple druids I can't even imagine how much money Blizzard made out of this, but if Blizzard went after them then it would cost money

I simply left a few games because cheaters ruined the game for me, that's what players should do, they should put the companies on their knees by leaving the game, then maybe devs will work on this... it's not hard

This is basically the point.

Games have had means of detecting contemporary cheating methods since the 90's. However, many game developers and publishers simply dont consider spending money on anti-cheats purely to ban accounts so they have less income to be a viable business decision.

What needs to happen in order for tools like the above to be widely used, and for accounts found to be cheating to get banned, is for gamers as a whole to reject the notion of playing games with cheaters, rather than simply accepting it as we do now.

If gamers simply do not accept cheating in games (especially games with a subscription model), then developers will quickly understand that detecting and banning cheaters is the only viable option.

Until gamers all make this decision though, developers have little in the way of incentive to even use tools like the above.

BaSkA_9x2

There already exist MMORPGs with systems in place that start recording players' inputs after they are flagged as potential cheaters. The flagging system can be automated, report driven or both, but I'm sure Ashes will have a good player flagging system in place.

Regarding reviewing the collected data (players' inputs), that is an expensive process throughout, so implementing AI to do it or to help with it would probably make it cheaper, but then I wonder how many mistakes it would make.

A mixed anti-cheat system where the AI gives its verdict but still requires a human to validate them might work, but I don't know if it's possible or even worth it to completely remove humans from this process.

Finally I wonder if that type of automated anti-cheat would detect something like this (blatant aimlock during a tournament which wasn't detected by the regular anti-cheat) in FPS games, and even if it did, I wonder if people would believe its verdict.