120
online anti cheats are the worst thing
(lemmy.ml)
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 24 Aug 2023
120 points (94.1% liked)
Linux Gaming
16094 readers
47 users here now
Gaming on the GNU/Linux operating system.
Recommended news sources:
Related chat:
Related Communities:
Please be nice to other members. Anyone not being nice will be banned. Keep it fun, respectful and just be awesome to each other.
founded 4 years ago
MODERATORS
But it's not. Easy anti-cheat, for example, works on Linux. The problem isn't with Linux, it's that developers don't target Linux, so their anti-cheat systems don't work on Linux.
And that's fine with me, though it would help Linux adoption if those games worked on Linux. But it's not an inherent limitation of Linux, it's just something devs need to proactively support.
Correction, EAC barely works on linux. Apex is just safer because Respawn themselves are putting in some effort.
Well yeah, you can't expect a third-party anti-cheat to solve all of your problems, each game is going to have idiosyncrasies.
I think Valve's Overwatch system is a fantastic example of ways to innovate without compromising a user's security or requiring platform-specific cheat detection. It's probably not enough on its own (those reviewers need data), but to me it's preferable to something more invasive like BattleEye. A lot of that can be done server-side, by running player movements through an AI model that detects players that fit certain patterns, or don't fit common patterns.
So EAC works, but it works at a different level than it does on windows. EAC does become less secure on both platforms when Linux support is enabled from my understanding. BattleEye, Vanguard and Riots AC don’t work on Linux either, which is a significant portion of major games right now.
I’d argue it is an inherent limitation of Linux, as it’s so open that it’s harder to validate a user isn’t using 3rd party programs to cheat.
I think it's an inherent limitation of client-side anticheat.
I think initiatives like Valve's Overwatch system is a much better approach because it relies on players who have a stake in eliminating cheaters instead of a constant war against whatever flavor of the week cheating engine people are using. Pair that with an AI model that looks for patterns (both whether players fit cheating patterns or don't fit expected patterns) and player-jurors will have enough information to make a call.
But that kind of initiative takes more effort than just integrating an off-the-shelf anti-cheat system, it forces companies to actually care about how their game runs on customer machines.
Personally I find Overwatch a horrible idea. It’s not terribly effective and relies on players, who are particularly unreliable at determining if someone is cheating. I believe those decisions should be entirely out of the hands of the players.
AI is still to expensive to run checks on every action that every player makes. Also any sort of automated system can’t be clearly banning people.
It doesn't need to be realtime, as long as the cheaters get caught after some days. So take samples from every player and run them through a learning algorithm and take more samples the more suspicious someone is.
The more important thing is how you deal with cheaters once you find them. I really like the idea of increasing lag to cheaters instead of outright banning them so you waste their time more than anything. And then if you find out they're not cheating, it's easy to just drop them from the pool.