Anticheat - A product of trash developer, IQ = ♥♥♥♥♥♥♥♥

[75HZ 30tap machine]

The vac that you guys created is a pile of brainless ♥♥♥♥; it works so well it's like having an IQ worse than an animal. You flag normal people in vac, without an apology, without providing any information to refute the claims, completely one-sided. It's an old game, and all you do with anti-cheats is occasionally add information to updates for entertainment, without any effort, and the result is a pile of ♥♥♥♥.

[Lisariné]

If VAC Net doesn't react — Valve is terrible.
If VAC Net reacts to suspicious activity — Valve is terrible.


CS2 Community is strange.

[もこにゃん]

Originally posted by Lisariné:

If VAC Net doesn't react — Valve is terrible.
If VAC Net reacts to suspicious activity — Valve is terrible.


CS2 Community is strange.

Please learn some knowledge in statistics.
Null hypothesis (H₀): The inspected player is not suspicious (the gameplay is regular)

• Type I Error (False Positive): You reject the null hypothesis (H₀) when H₀ is actually true. (Risk level is set by α, the significance level).
In layman's term: The system thinks the player has irregular gameplay, cancels the match and forces the whole team with the player into cooldown, but the player is not cheating at all. This is the most annoying case for the team with the unfairly suspected player.

• Type II Error (False Negative): You fail to reject the null hypothesis (H₀) when H₀ is actually false. (Risk level is β; statistical power is 1 - β).
In layman's term: The system thinks the player has regular gameplay, but the player is cheating, like those spinbotters that are still freely cheating. This is the most annoying case for the team facing the cheating team.

• Correct Inference (True Negative): You do not reject H₀ when H₀ is true.
In layman's term: The system thinks the player has regular gameplay, and the player is actually not cheating, so it is correct. This is supposed to be the majority.

• Correct Inference (True Positive): You reject H₀ when H₀ is false.
In layman's term: The system thinks the player has irregular gameplay, cancels the match and forces the whole team with the player into cooldown, and the player is actually cheating, so it is correct. This is the most satisfying case.

Now the AI model (which is a statistical model) is being aggressive in an attempt to eliminate Type II Errors (false negative), but it instead generates too many Type I Error cases, while still fails to eliminate Type II Error cases.

Sorry, but it is not that strange after you study statistics properly, even without actually studying advanced machine learning or deep learning.