5 ms·

I never understood why anyone would ban cheaters or go after the cheat creators. Pay 40$, reverse engineer the cheats and get a cheap list of security issues y

by tuco86 8y ago

I never understood why anyone would ban cheaters or go after the cheat creators. Pay 40$, reverse engineer the cheats and get a cheap list of security issues you ought to fix anyways. Thus raise the bar so it becomes hard. Hire the remaining few very skilled crackers for your reverse engineering team (or bug bounty them). In contrast to lobbying the/a government for overreach laws, have the government find the offenders and then litigate. There has to be a better and cheaper way. Also bring back dedicated servers so the community can police itself.

MaulingMonkey 8y ago
Blue team can't always keep hackers out of bank servers, how the hell are they going to keep hackers out of the software running on the hacker's own hardware? They have physical access! Even if you turn the game client into a perfect dumb client that merely consumes video (ala game streaming services), in theory there are still opportunities to cheat - image recognition is an ongoing area of research into machine learning, no reason you couldn't apply that towards creating aimbots or other tools that replace player skill. In practice, game clients are way more hackable than a dumb client for performance reasons. RTSes use lockstep networking models, reducing bandwidth but allowing map hacks. FPSes predict movement and render locally to reduce perceived latency, enabling the extraction of accurate 3D position information and replacing camouflage with blinking pink neon "shoot here" textures. Reducing server costs by using P2P networking models allows players to DoS each other instead of the game servers, and gives them more authority over the game state to tamper with. And that's just the tip of the iceberg of intentional design decisions - we haven't even gotten into actual bugs in the implementation yet. You can do a cat and mouse game of detecting and banning and obfuscating and so on, but you're only going to stem the tide and contain the damage, not actually stop it any more than you're suddenly going to stop piracy. And even doing that involves spending a lot more resources than $40 and a couple of FTEs if your game is even remotely popular. And runs the risk of understandably pissing off your legitimate users if your game appears to be spying on them. And if you avoid that, it will still be more resources you can't spend on making new content, new games, QA, etc.
- tuco86 8y ago
  You are right, it's an unwinnable cat and mouse situation. I was focusing too much on the printing of in-game currency in my head and didn't communicate. I wasn't proposing to stopping it, tho. Make it hard and make them hide, cheats aren't all powerful if you try to stay undetected. Also, having no security around the currency for years, when it's such a problem, just seems lazy. Bringing law enforcement into the equation won't stop cheats either. And it involves huge costs for the company and much more for the society, I think. Also: ML can be used on both sides as well. A scaling Autoadmin which detects odd behavior doesn't seem harder than a bot playing for you without showing odd behavior.

by tuco86 8y ago

MaulingMonkey 8y ago

Blue team can't always keep hackers out of bank servers, how the hell are they going to keep hackers out of the software running on the hacker's own hardware? They have physical access! Even if you turn the game client into a perfect dumb client that merely consumes video (ala game streaming services), in theory there are still opportunities to cheat - image recognition is an ongoing area of research into machine learning, no reason you couldn't apply that towards creating aimbots or other tools that replace player skill. In practice, game clients are way more hackable than a dumb client for performance reasons. RTSes use lockstep networking models, reducing bandwidth but allowing map hacks. FPSes predict movement and render locally to reduce perceived latency, enabling the extraction of accurate 3D position information and replacing camouflage with blinking pink neon "shoot here" textures. Reducing server costs by using P2P networking models allows players to DoS each other instead of the game servers, and gives them more authority over the game state to tamper with. And that's just the tip of the iceberg of intentional design decisions - we haven't even gotten into actual bugs in the implementation yet. You can do a cat and mouse game of detecting and banning and obfuscating and so on, but you're only going to stem the tide and contain the damage, not actually stop it any more than you're suddenly going to stop piracy. And even doing that involves spending a lot more resources than $40 and a couple of FTEs if your game is even remotely popular. And runs the risk of understandably pissing off your legitimate users if your game appears to be spying on them. And if you avoid that, it will still be more resources you can't spend on making new content, new games, QA, etc.

tuco86 8y ago
You are right, it's an unwinnable cat and mouse situation. I was focusing too much on the printing of in-game currency in my head and didn't communicate. I wasn't proposing to stopping it, tho. Make it hard and make them hide, cheats aren't all powerful if you try to stay undetected. Also, having no security around the currency for years, when it's such a problem, just seems lazy. Bringing law enforcement into the equation won't stop cheats either. And it involves huge costs for the company and much more for the society, I think. Also: ML can be used on both sides as well. A scaling Autoadmin which detects odd behavior doesn't seem harder than a bot playing for you without showing odd behavior.