Curious about ping delay/baked simulation/impact about esp-analysis

[ryanjpwatts]

Yes, they have definitely tried and seemingly succeeded before.

During my research I came across this which suggests Valve were taking similar steps to prevent it and players claimed it worked with great results but footage of functional cross-map ESPs in CS2 seem to paint a different picture unfortunately, even in competitive matchmaking.

I would definitely expect Valve to consider the approach of baked occlusion culling data as they all have static, unchanging maps. It would likely be a lot more performant but effectiveness could be questionable for the reason you proposed with these baked areas potentially being too large. My solution aimed to support changing world environments like you would see in games like Fortnite, Rust and Overwatch.

I did some expirementing with latency (as yes, mirror does thankfully provide latency simulation) and and the results weren't too different with 100ms, regardless this could be alleviated by sending rays to positions ahead of where the player may move, a technique already done in a lot of server-authorative movement systems where the clients player will move on their screen based on the inputs provided despite the server being the true assigner of position.

Factoring in the requirement to network the position of both player and player camera I may compare the performance between engine camera culling techniques and my raycast approach, it would be good to see in numbers how both solutions could affect a server.