I want to help !!

[chewymole]

Hi, I play warzone, and its 30% or higher infested with hackers/cheats. I have thought of this same idea, but on a smaller scale and only with public data available from Activision. Problem is its very hard to get access to it. I found 1 source awhile back but its not longer working. and had some code actually pulling my game data live on my website. But again that fell through and stopped working.

My plan was to get my lobby, then poll for all players in my lobby (with public data) which kinda sinks the ship, anyway then get any who have a 90% or higher headshot average and have a low rank say 50 and below or low time played against average headshots on all games played for that user. I am sure there are better metrics to use to identify hackers/cheaters. but this is a good starting point.

Cheaters get banned all the time, they just start up a new account and go again. So looking for new accounts with high accuracy is a good place to start.

if i found any i would log it, then go look up their public record on cod.tracker.gg but i never got that far because again my data source fell through.

Now, i like your idea but i am afraid you would have a hard time getting the data sets you need. not sure if the public data tracks hit box or the like. And if that's the case, you would need a client, which would be considered a cheat to use by the game. For instance PunkBuster (https://punkbuster.en.softonic.com/) which a lot of game engines use, but is basically useless in todays games. It has been around for years.

Ideally the game makers would allow Your anti-cheat client app to be installed with the game as a plugin/add-on, then your API service would do all the work after the clients push you the data sets. The game doesn't matter, as long as they have a plugin / add-on architecture system that can push game data through to the plugin. Then the plugin would just push the data to the ML/AI system. Once the system detected a cheater in the designated lobby, it could alert the pusher client which could then display something on screen. If enough players had the app, they could all team up and destroy the cheater. In a perfect world, the cheater would be highlighted on the screen for the entire lobby to see.

Game makers need to (if they have not already) generate a unique installation fingerprint for each game based on the hardware that is being used. So if its uninstalled / reinstalled it would end up with the same fingerprint. Then once a cheater is identified, that fingerprint is then stored in a db of bad guys. so if they try again, the insta-ban process will kick them to the curb.

[Nimdy]

@chewymole

Thanks for jumping on this train with me. I am sure many others way smarter than us will jump on it as well in the future.

The whole idea here is to get folks talking and marrying up our ideas into some type of framework that companies will want to jump in on.

Ultimately, these issues can be fixed... or to the point like I suggested in the original write-up "wont be worth them cheating"

The data is the hardest part and I also love the the way your started taking the data you could find and compare it to your analytical idea.

I believe one of the end state concepts is the dev work that will need to be added to the coding of FPS games. For example logging of metrics of:

Average PDTH    (Player + Distance to Target Hit)
Average PDTHW   (Player + Distance to Target Hit + Weapon)
Average PTTKD   (Player + Target + Time to Kill + Distance)
Average PTTKDR  (Player + Target + Time to Kill + Distance + Reports)
Average PTTKHHL (Player + Target + Time to Kill + Hitbox Hit Location)
Average Weapon Distance Hit
Average Weapon Distance Kill
Average Weapon Hitbox Hit Location
Average Weapon Distance Hit + Hit Box Location Hit
Average Weapon Distance Kill + Hit Box Location Hit
Average Weapon Time to Kill
Average Weapon Accuracy

These listed above are just ideas and can expand in all types of different directions of course.

Until then, we are stuck here coming up with our own idea.

So based on what you said... maybe we could design a https://www.warcraftlogs.com/ like application to be ran with any game. However, the big boulder is what data would be able to collect?

I dont think we would get the data we really are aiming for.

Yes, your thought behind:

My plan was to get my lobby, then poll for all players in my lobby (with public data) which kinda sinks the ship, anyway then get any who have a 90% or higher headshot average and have a low rank say 50 and below or low time played against average headshots on all games played for that user. I am sure there are better metrics to use to identify hackers/cheaters. but this is a good starting point.

Is what we might just be stuck with unless, we can get one of these companies to give us some data samples.

All we need are some data samples... and if the data we need is not there, then we ask a company to work with people like to design metric collections that need to be added to the game.

Please keep dumping your thoughts down and I will continue to do the same.

I am sure others will come and help out.

This is a huge problem... we are just trying to help solve a problem instead of complaining about it. :)

Keep in touch and hit me up on discord.