One of my all-time favourite first-person shooters, Counter-Strike: Global Offensive, is about to tell naughty players to sort their attitude out.
In a new blog post on the official site, the Valve team write that while you can already tweak what you see in-game like turning off players' avatars, names, and voice/text chat—they're a bit of a nuclear option as it affects everyone. To help with this, Valve has been working on a new system based on reports.
Yes, reports. User-made reports on behaviour will now have a bigger effect on players who are deemed to be repeat offenders. People who get a lot of reports, will end up seeing a warning message and if they continue being crap they'll become muted by everyone by default until they earn "enough XP to remove the penalty". However, people can still go and unmute them manually.
Valve said they're already tracking the reports, and said to get into the habit of using the reporting system with the "Abusive Communications or Profile" option if you come across someone being terrible.
Interestingly, the report system takes into account the person reporting as much as the person being reported. If a regular player who doesn't report often makes a report, it will be weighted more heavily than someone who doesn't play often or reports constantly.
What are your thoughts on such a system?
You can find CS:GO free on Steam.
Should prevent the "I'm so offended by everything!"-people from effectively mass-reporting everyone.
I also like that it still lets people play, they are just being muted.
More games should do that.
On CSGO I do some "overwatchs" and there's a lot of obvious cheaters, with wallhacks, aim, speed hacks, etc, really easy to catch, so an IA can do the job. Looks like it's really easy to bypass VAC, but I think that it's a lot harder to bypass a AI.
Maybe an IA can detect a toxic player too
At least if cheaters were put aside it would help this great game. Yesterday, I stopped to play after 4 different rounds I was in front of cheating people with no fun. Replay VERY helps and it should be used by full time moderators after reports.
The inventiveness of cheats is also incredible: the classical guy that only killed by head shoot in one bullet whatever the gun he had, the guy that can take 3 awp shots in a raw with only 80hp decreased (meaning that he can probably tweak his cheat? or with adding more armor? don't know how) -seen yesterday, the guy that can view all hidden people ...I don't know how also :D
This game is a hack hole for hackers XD
Quoting: fagnerlnA little off topic, I was talking with someone about Proton and how intrusive anti cheats blocks it, and I was thinking about why there's no anti cheat based on AI.
On CSGO I do some "overwatchs" and there's a lot of obvious cheaters, with wallhacks, aim, speed hacks, etc, really easy to catch, so an IA can do the job. Looks like it's really easy to bypass VAC, but I think that it's a lot harder to bypass a AI.
Maybe an IA can detect a toxic player too
Valve is already using one. The problem is that AI requires training and will always have false positive detection's (but to be fair, humans do have them as well), so I doubt that it will eventually get to be the unique security measure to be used.
Quoting: fagnerlnA little off topic, I was talking with someone about Proton and how intrusive anti cheats blocks it, and I was thinking about why there's no anti cheat based on AI.To a degree an AI could be used to detect these sorts of things and such AI approaches have already been used in the past. Hell, who knows, maybe you doing overwatches is already training an AI model behind the scenes? But AI isn't a silver bullet, despite how much some marketing teams in tech may want you to believe it is. Particularly in terms of toxic language, AI will have serious trouble deriving context to make determinations on what is and isn't toxic. It'll either let toxic stuff pass or you end up with a Scunthorpe problem. You can ask YouTubers what they think of Google's monetization AI and how well it works (or rather, how poorly it does). Some problems are just best dealt with by humans.
On CSGO I do some "overwatchs" and there's a lot of obvious cheaters, with wallhacks, aim, speed hacks, etc, really easy to catch, so an IA can do the job. Looks like it's really easy to bypass VAC, but I think that it's a lot harder to bypass a AI.
Maybe an IA can detect a toxic player too
Quoting: SirLootALotIsn't the entire point of cs:go to swear at each other in russian?or in spanish, depending on time you play ^^
See more from me