VACNET Anti-cheat
Channel:3kliksphilip · Length:13:03 · Published:2018-04-10 · Views:573853
Summary
3kliksphilip summarizes Valve's GDC talk on VACNet, a deep-learning system that flags CS:GO cheaters and feeds the Overwatch review queue. It covers Trust Factor, VAC, and how the data is collected and analyzed.
Why HN recommends it
HN cites it as a source for how Valve uses machine learning on player behavior to detect cheating.
Valve uses deep learning model called VACnet to "watch" every player on their servers — HN, 2020
Machine learning & deep learning · Gaming culture & analysis · Speedrunning, glitches & game hacking
Subtitles (excerpt)
At GDC 2018 Valve ran a presentation about Deep Learning and how they’re using it to combat cheaters. I really suggest checking it out- the whole thing’s fascinating. The talk is 45 minutes long and then the rest is for questions and answers. But if you don’t have the time, then I’ve tried to condense it in this video along with any other interesting bits of information they revealed. The project began in 2016 when Valve looked at what the community wanted the most. But they couldn’t hear anything because the problem of cheating was too loud. It drowned out any other suggestions! But was cheating the problem, or was it just that people thought it was? Apparently, VAC’s results made it clear that there was a cheater problem. This is partly because CS:GO’s engine is old. It’s 75-80% the same as Half Life 2’s and any cheats that have been made for any Source game since then can probably be ported rather easily, though it was stressed that they keep an eye out for these and ban anything posted publicly. CS:GO is also a good target for cheats because its combat is short and precise, making cheats very effective. There are 2 main categories: aim, and ‘info’ assistance, which is stuff like wallhack and
On Hacker News
If you're interested, Valve uses deep learning model called VACnet to "watch" every player on their servers[0][1]. If any dodgy gameplay is detected, it gets submitted into a moderation queue called Overwatch[2], which is basically delegates the decisions to ban cheaters community. It's really cool how they've gamified this. [0] [LINK] [1] [VIDEO] [2] [LINK] — cameronbrown, 2020-07
That's not what I said - my point is that companies with a relaxed security baseline (ie. not having a dedicated team of people tasked with investigating cheats and improving their anti-cheat system) tend to invite an increased number of cheaters into their games, and I back this up with real events. - These companies will send cease-and-desist letters to cheat makers instead of working to detect… — judge2020, 2023-04
Worth noting that the VAC games on that list, CS:GO, TF2, and Dota 2, have effectively given up on actually detecting and preventing cheaters and now rely on server-based or community-consensus methods of finding and punishing cheaters (with TF2 being the exception with how little maintenance Valve performs on it). Both CS:GO and Dota 2 have a report+overwatch system where users watch a replay to… — judge2020, 2021-09
[VIDEO] Valve is already using machine learning to analyze player mouse movements across all games to look for "inhuman" patterns and send those directly to the community-staffed manual review process called overwatch. — kllrnohj, 2021-07
> They also have a neural network setup that constantly scans match logs for cheaters For those wondering what this is referring to, it's a system called VACNet[0] presented at GDC last year. [0] [LINK] [1] [VIDEO] — cameronbrown, 2019-07
Related videos
- The spelled-out intro to neural networks and backpropagation: building micrograd · Andrej Karpathy · 2:25:52
- In Search Of A Flat Earth · Folding Ideas · 1:16:16
- But what is a neural network? | Deep learning chapter 1 · 3Blue1Brown · 18:40
- SM64 - Watch for Rolling Rocks - 0.5x A Presses (Commentated) [OUTDATED] · pannenkoek2012 · 24:38
- Reverse emulating the NES to give it SUPER POWERS! · suckerpinch · 22:21
- This Is Phil Fish · Innuendo Studios · 19:05
- AlphaGo - The Movie | Full award-winning documentary · Google DeepMind · 1:30:28
- Transformers, the tech behind LLMs | Deep Learning Chapter 5 · 3Blue1Brown · 27:14
- USENIX Security '18-Q: Why Do Keynote Speakers Keep Suggesting That Improving Security Is Possible? · USENIX · 51:22
- Lecture 1 | Machine Learning (Stanford) · Stanford · 1:08:40
- MarI/O - Machine Learning for Video Games · SethBling · 5:57
- Deep Dive into LLMs like ChatGPT · Andrej Karpathy · 3:31:23
- The Engoodening of No Man's Sky · Internet Historian · 53:51
- Computer program that learns to play classic NES games · suckerpinch · 16:17
- RL Course by David Silver - Lecture 1: Introduction to Reinforcement Learning · Google DeepMind · 1:28:12
- Let's build GPT: from scratch, in code, spelled out. · Andrej Karpathy · 1:56:20