@n3c1k you have some discord or whatever? i am building an analytical tool too, thought we might share some insights or just generally useful stuff. been grinding my head around various positioning valuation metrics for a while now.
@n3c1k Ah, I see. I asked because my current method is much simpler: I sample 9 weighted body points, raycast them against the static map mesh, apply FOV, then calculate visible body coverage.
So I was trying to understand how much more accurate full GPU visibility method would be
@n3c1k what i meant by LOS raycast is using tri static map mesh + raycasts in order to determine visibility of players. u previously said u use "GPU vision detection with animation reconstruction", so i wondered how much more accurate is that compared to tri mesh+raycasts
@n3c1k thanks, got it. I also once asked a question under your other post, but i'll ask here too: have you compared the actually accuracy of GPU rendering method vs LOS raycast? Since rebuilding a scene from demo's ticks is quite heavy, i wondered if it yields substantial improvement
@SiNister1080p What are you talking about? You can perfectly cheat on both XBOX and PS5. Multiple emulation devices and AI cheats exist there and usually it's even safer to cheat on those platforms than PCs.
@ottrTweets I think a pricetag on a game that clearly feels like it would benefit from a battle pass type of monetization is weird. It will gatekeep a lot of players.
@n3c1k also I am developing a platform for analysis too, but on more ephasis on aiming metrics and positioning. would be glad to have a chat, maybe share some findings :)
@n3c1k have you tested it against the geometric LOS approximation from demo state (using tri static map mesh + raycasts), how big of an advantage is your approach? would be interested in seeing numbers.
Because your approach is magnitudes heavier i believe.
@ArcRaidersInfo I am not saying it's what the game needs or whatever, but if it had pve mode, i'd for sure jump on it and play it with friends. (I love pvp in games, just hate it in extractions)
@Kerovski I mean apart from manual reviews, you can't really detect AI aim assist. They just need to hire like 5-10 people with strong analysis tools to ban those obvious cheaters, there's no way around it. 2 PC AI Aim Assist is almost undetectable if done right
@Kerovski I did something even better :) I had no-recoil script running on both old and newer patches. Same script, same mouse pattern. On older patch everything lands in the same spot, on the new patch everything is semi-random. Idk what they changed, but they in fact did.
@fzvjwdsty@ggsarzi@ZPostFacto If they arenโt replicated over the network like real players, theyโre already distinguishable to any decent cheat developer