> I would love to see a modern competitive game with optional anticheat that, when enabled, allows you to queue for a separate matchmaking pool that is exclusive to other anticheat users. For players in the no-anticheat pool, there could be "community moderation" that anti-anticheat players advocate for.
This is roughly what Valve does for CS2. But, as far as I understand, it's not very effective and unfortunately still results in higher cheating rates than e.g. Valorant.
Huh. When you say that "it's not very effective" do you mean the segmentation between the pools, or the actual anticheat isn't very good? (I'm assuming the latter - I've heard that VAC is pretty bad as far as anticheat goes)
Oh sorry - I misread your suggestion! I thought you were talking about separate matchmaking logic for known cheaters, but you're asking about opt-in matchmaking for those willing to use invasive anticheat.
The example still kind of applies. In the CS world, serious players use Faceit for matchmaking, which requires you to install a kernel-level anticheat. This is basically what you're suggesting, but operated by a 3rd party.
Hmm, I guess that since VAC is not a kernel-level anticheat, the comparison between it and Faceit for CS is pretty close to my idea. Thanks for pointing that out.
VAC is actually an AI based anticheat. I guess IF (a big if) it ever gets good enough it will be better than any kernel level AC, because it analyzes the gameplay, not the inputs, meaning a DMA cheat would also be caught.
But so far that still seems to be miles away.
"VAC" is a catch-all term for all of Valve's anti-cheating mechanisms.
The primary one is a standard user-mode software module, that does traditional scanning.
The AI mechanism you're referring to is these days referred to as "VAC Live" (previously, VACNet). The primary game it is deployed on is Counter-Strike 2. From what we understand, it is a very game-dependent stack, so it is not universally deploy-able.
I don't think that's what VAC is. I think VAC just looks for known cheat patterns in memory and such, and if it finds indisputable proof of cheating it marks a player for banning in the next wave. Maybe there is some ML involved in finding these patterns but I think it's very strictly controlled by humans to prevent fase positives. That's why VAC bans are irreversible, false positives are supposed to be impossible.
Valve has some AI detection stuff for CS2, but it’s remarkably ineffective. VAC itself delivers small DLLs that get manual mapped by Steam service, do some analysis and send that to Valve (at least to the best of my knowledge, there may be more logic implemented in Valve’s games or in Steam/Steam service).
Community alternative (faceit) requires kernel level access. The actual anticheat matchmaking is essentially unplayable
Wait, so the "community alternative" is also kernel-level anticheat? I think that's different from what I'm proposing - I'm suggesting a comparison between an anticheat and no anticheat (with community policing of lobbies and handing out of penalties).
Why would a player knowingly choose to play on matchmaking that is advertising no anti-cheat?
But anyway counterstrike did have community policing of lobbies called overwatch - https://counterstrike.fandom.com/wiki/Overwatch
It was terrible as it required the community to conclude beyond reasonable doubt the suspect was cheating, and cheats today are sophisticated enough to make that conclusion very difficult to make
Because their (or their friend's) computer can't run the anticheat, but they're interested in playing with friends? My sister and mom wanted me to play Valorant with them a free years back, but apparently it needs kernel anticheat, so I just can't run it. I'm not going to buy a new computer for a game.
And the way community policing worked in the past is that the "police" (refs) could just kick or ban you. They don't need a trial system if the community doesn't want that.
> Why would a player knowingly choose to play on matchmaking that is advertising no anti-cheat?
I guess I didn't exactly make that clear...
A few of the arguments advanced by the "anti-anticheat" crowd that inevitably pops up in these threads are "anticheat is ineffective so there's no point to using it" and "anticheat is immoral because players aren't given a choice to use it or not and most of them would choose to not use it".
I don't believe that either of these are true (and given the choice I would almost never pick the no-anticheat queue), but there's not a lot of good high-quality data to back that up. Hence, the proposal for a dual-queue system to try to gather that data.
Putting in the community review of the no-anticheat pool is just to head off the inevitable goalpost-moving of "well of course no system would be worse than a crappy system (anticheat), you need to compare the best available alternative (community moderation)".
> Why would a player knowingly choose to play on matchmaking that is advertising no anti-cheat?
My understanding of the proposal is that it advertises no invasive anticheat (meaning mostly rootkit/kernel anticheat). So, the value proposition is anyone who doesn't want a rootkit on their computer. This could be due to anything from security concerns to desiring (more) meaningful ownership of one's devices.
VAC (the valve anticheat) is not kernel-level. The community alternative is. The official matchmaking is pretty full of cheaters.
VAC is essentially no anticheat with how easily it is bypassed.