Any chance we see new devs pick up this anticheat to make their games work on windows desktop pcs and Linux handhelds alike? With win 10 death and steam deck 2 certainly in their forecasting, I wonder if valve is gonna push Microsoft around a little.
One could argue windows wouldn’t have the market share it does without steam. Maybe steam Linux support becomes a priority over windows now.
Valve Anti-Cheat (VAC) is available to all developers using the Steam Platform, but it’s not particularly effective unless you apply personal tweaks or use it in conjunction with another anticheat system (such as sigh kernel level anticheats).
Many of the big PVP games on Steam employ this hybrid system, with some cheaters getting VAC’d and others receiving “Game bans” because of the different tools employed.
Valve is very vague about their anti-cheat systems. They didn’t say they updated VAC. It could’ve been an update to VACnet, their machine-learning anticheat. I’m not sure if VACnet is available for other devs but I imagine other devs also aren’t particularly interested because ML-based anticheat would be even harder to implement.
John McDonald (Key engineer behind VACnet) has offered to extend the service to other game developers, to the point of even publicly requesting partnerships at a GDC conference.
Valve is intentionally vague about its anticheats, yes, but that is because each implementation is somewhat unique. TF2’s, for example, was almost cracked completely open through analysis of the source code leak alongside pattern detection of bots and previously banned players until recent events.
It’s believed that Dota and CS both have their own unique spin on how they do things, with I believe CS being the primary arm for VACnet to be tested on, and I would imagine it’s not a drop-in solution for any game developed outside of Valve. They are open to partnerships though, so who knows?
I imagine they’re vague about how their anti cheat works in an attempt to obscure the methods of detection so people can’t use that information to get around it.
Absolutely. I wasn’t trying to imply Valve wouldn’t offer VACnet to others. I’m sure they’d be happy to partner up with other devs. My point was more that other devs probably aren’t interesting because of how hard it would be the implement. Like you said Dota and CS have different implementations and I imagine deadlock does too. I imagine most games would end up with either a custom implementation or a custom model, both require a significant work on the developers side. It’s probably easier to add something like EAC than VACnet so devs most likely go that route.
One solution to maintain Linux compatibility with good protections could be similar to what Embark Studios is doing with their games (The Finals, and Arc Raiders): Windows version gets Kernel-level anti-cheat and additional layers, Proton branch gets a user-level version of the anticheat instead.
Personally, I would argue that the majority of individual cheaters (who are seeking to undermine their opponents and win with minimal effort) are likely to be on Windows platforms where kernel-level protections would be most effective in detection. With an Embark-style dual solution - you can both catch the most cheaters and score W’s with the Linux community.
Any chance we see new devs pick up this anticheat to make their games work on windows desktop pcs and Linux handhelds alike? With win 10 death and steam deck 2 certainly in their forecasting, I wonder if valve is gonna push Microsoft around a little.
One could argue windows wouldn’t have the market share it does without steam. Maybe steam Linux support becomes a priority over windows now.
Valve Anti-Cheat (VAC) is available to all developers using the Steam Platform, but it’s not particularly effective unless you apply personal tweaks or use it in conjunction with another anticheat system (such as sigh kernel level anticheats).
Many of the big PVP games on Steam employ this hybrid system, with some cheaters getting VAC’d and others receiving “Game bans” because of the different tools employed.
Valve is very vague about their anti-cheat systems. They didn’t say they updated VAC. It could’ve been an update to VACnet, their machine-learning anticheat. I’m not sure if VACnet is available for other devs but I imagine other devs also aren’t particularly interested because ML-based anticheat would be even harder to implement.
John McDonald (Key engineer behind VACnet) has offered to extend the service to other game developers, to the point of even publicly requesting partnerships at a GDC conference.
Valve is intentionally vague about its anticheats, yes, but that is because each implementation is somewhat unique. TF2’s, for example, was almost cracked completely open through analysis of the source code leak alongside pattern detection of bots and previously banned players until recent events.
It’s believed that Dota and CS both have their own unique spin on how they do things, with I believe CS being the primary arm for VACnet to be tested on, and I would imagine it’s not a drop-in solution for any game developed outside of Valve. They are open to partnerships though, so who knows?
I imagine they’re vague about how their anti cheat works in an attempt to obscure the methods of detection so people can’t use that information to get around it.
Absolutely. I wasn’t trying to imply Valve wouldn’t offer VACnet to others. I’m sure they’d be happy to partner up with other devs. My point was more that other devs probably aren’t interesting because of how hard it would be the implement. Like you said Dota and CS have different implementations and I imagine deadlock does too. I imagine most games would end up with either a custom implementation or a custom model, both require a significant work on the developers side. It’s probably easier to add something like EAC than VACnet so devs most likely go that route.
One solution to maintain Linux compatibility with good protections could be similar to what Embark Studios is doing with their games (The Finals, and Arc Raiders): Windows version gets Kernel-level anti-cheat and additional layers, Proton branch gets a user-level version of the anticheat instead.
Personally, I would argue that the majority of individual cheaters (who are seeking to undermine their opponents and win with minimal effort) are likely to be on Windows platforms where kernel-level protections would be most effective in detection. With an Embark-style dual solution - you can both catch the most cheaters and score W’s with the Linux community.