Suggestion: rate limiting damage to buildings per player to prevent abuse in RenX

[Silv]

On multiple occasions a public game has been messed up by a cheater who somehow manages to do kill buildings in very rapid succession. From what I've heard that cheater attacks buildings from the outside with a normal weapon and they instantly go down. I don't know exactly how that works but I suspect it's either a way to increase damage or firing rate. Unfortunately I don't have specific examples or replays demonstrating the cheat.

A cheater who instantly kills infantry or vehicles is annoying but ultimately a problem local to a section of the battlefield and it's well possible to recover from having a large part of the team wiped out on the battlefield. A cheater who destroys buildings however messes up the long term balance of a match, losing just 2 buildings (particularly if one of them is the barracks/hand) is a very major blow that usually sends a team into a spiral towards defeat. Matches in which this happens due to cheating are usually vote restarted or map changed.

Assuming the buildings are killed due to excessive but otherwise normal damage done to them a fairly simple and resource friendly addition to the game logic should be able to reduce the effectiveness of this kind of cheat to a point where a halfway competent (by pub standards) team can easily defend against it. There is an upper bound to how much damage per second a single player can do to a building using normal weapons. If the server would keep track per player of how much damage to buildings has been done in the last second and would reject it when more damage is attempted this hack would be neutered.

Of course some damage types shouldn't count because they are legitimately able to rapidly destroy buildings: C4 and superweapons. The MCT can sometimes amplify damage significantly, but as it's unlikely a cheater can reliably get into a building to take potshots at the MCT having the anti-cheat not cover MCT damage likely doesn't reduce it's effectiveness much.

One downside of this approach is that it adds special case logic for a rather specific form of cheating, it would plug an active problem but doesn't do anything against cheating in general. I'm also not sure, not being used to game development, whether this approach is viable in the face of things like replication and network lag.

Another major downside is that it's quite possible that my layman's understanding of how the cheat works is off and this might not do anything at all. I'm mainly posting this based on what I've heard from others and some uninformed assumptions about how the game engine could be tricked.

The suggested approach is fairly conservative, it doesn't really prevent the cheating but reduces the impact. This is a tradeoff to allow for a simple and efficient implementation. But by limiting damage to a maximum regardless of (non-super/C4) weapon it would still be possible for a cheater to have a much more effective weapon than is intended.

P.S. I don't know if it's ok to post this in the open. I figured that as the source is available and this post doesn't contain any useful info about how the cheat actually works it shouldn't help anyone intending to cheat.

[NodSaibot]

The issue is being worked on.

[Silv]

Thanks for the info. Good to hear these kinds of issues are getting swift attention.