Auto Kick for hacks is BS

As a general rule, hackers are very clever and only "hack" when they need to, and cover themselves very well, thats why its a very slim line between hacker and a very good player.

You seem to have a better graps on statistics on statistics than I do mathematically.
But my troubles are more basic.

First of all, cheaters don't have to necessarily stand out statwise. I've seen an aimbotter who barely managed 1.0 K/D because he just ran around shooting without a care for his surroundings.
The ability to see through walls is certainly an advantage, but even so the cheater will still get gunned down as soon as the fighting starts. It might boost him to 4.0 K/D, which is rare, but common enough to be encountered regularly.

Second, let's look at a couple examples what I think statistics simple can't manage.
Do you think somebody with a 76.5 K/D is a cheater?
Well, he's also got 50SPM and 0.3 KPM and basically only lives in the spawnroom.
He's actually a user here in the forum who thought it was funny to see how high he could go as a spawnroomwarrior.

Or 15.0 K/D - MAX user only...

Or a K/D of 12? Sniper specialist, barely doing anything else.

What about short bursts of almost impossible feats?
Like 50 kills in tank? In ONE minute?
I managed that in my lockdown-HE prowler once. No need to cheat, but that one certainly must stand out.

Or maybe 13 kills with an LMG before dying, in barely 15 seconds?
Of course I shot a whole crowd of NC in the back after they caught my AV grenade...


Sure, you WILL get those guys who kill 50 guys in 100 seconds with small arms.
But those are the ones that do things that are not statistically impropable, they are doing something that is simply mathematically impossible within the established rules of the game.
The more subtle ones will simply submerge in the data you get from the rest of the crowd.

In games where they have stat based software (in fact PS2 does, of course, only a very basic form of it) it is pretty amazing what stands out in large enough data sets.

It wouldn't stop or detect all subtle-cheaters, but it would detect an awful lot more than the current 0% of them.

Those are legitimate concerns, and addressing them is a matter of pointing out that just like there are physical "laws" of nature concerning gravitation, etc., there are also laws of nature in statistics as well.

Simply put, it is impossible for a human to cheat and not stand out over time, and that's the key: time. Human behavior over time is as transparent to the statistical science as a pane of pristine glass. A cheater might be able to hide temporarily, but as the data comes in, the patterns of legitimate play observed, and samples taken, the cheater will be outed by the mathematics governing their own behavior.

I think you'll be surprised how much your feat of 50 kills in a minute using an anchored HE Prowler doesn't stand out. Yes, it was quite a feat, but I'm willing to bet it doesn't deviate far from similar streaks of similarly equipped Prowlers. Please understand I'm not intending to minimize your accomplishment, but rather trying to put it in mathematical context. You're a human player. It is therefore impossible that no other human player has done something similar with similar equipment in a similar situation.

A good example is the World Record phenomenon. Every time someone breaks a "World Record" it is proclaimed as an insular achievement, that they accomplished something no one has ever done before. Well, that's scientifically incorrect. Often times "World Records" can be easily duplicated (like the world's largest pizza, or world's largest ball of rubber bands). World Records that are broken are never done so at wide margins, but just barely. When the scope of the break is evaluated they examine averages and determine relatively how great the new record surpasses the old, but those breaks are themselves a pattern, and they are assymptotic, which means there is in fact an absolute fastest mile someone can run. There is an absolute fastest time that a Rubix Cube can be solved. And on, and on. We don't actually arrive at those limits, but we inch closer, and every time inch a little less, but always closer. So there will always be a new "World Record," but never--legitimately--beyond a certain threshold.

So, while you scored 50 kills in a minute with an anchored HE Prowler (and let's assume that's a record) the guy behind you might've scored 50 kills in a 1.03 minutes. The guy before that 50 kills in 1.10 minutes. And so on, and so on. Assuming yours is a record it'll be broken when someone scores 50 kills in 0.99 minutes, or something.

When we collect data on player behavior we want to know everything. Not just about their kill streaks, but what equipment they're using, what continent they're fighting on, in what hex they're fighting, how large a battle it is, how many shots they fire, how many shots land, and where, and even how they manage recoil, and we take that information and use a special piece of software called SPSS (Statistical Package for the Social Sciences) and build massive aggregations of data. Then we organize that data into Standard Normal Distributions (bell curves) describing the various variables--which are not infinite because there are a finite number of variables that describe the game state. Then we begin tracking players along those distributions, literally from second to second.

The very moment a cheater turns on an aim bot his accuracy is going to momentarily jump. He'll turn it off, and turn it back on, and off, and on, and each time he does his accuracy jumps. That's a human pattern, not random, impossible to make random. It's predictable. He'll get banned in minutes, and never see it coming.

I think it's doing OK as is, but obviously there's always room for improvement. I for one LOVE seeing those alerts about "DouchePantsNancy has been banned for cheating!", I see it as a victory and get a good chuckle whenever the admin's get creative with the alert (there's been some pretty funny ones, lol). Secondly, they aren't blindly just banning everyone that gets reported (I get reported constantly, because whoever does the reporting usually sends me a tell, talking about how screwed I am because I just got reported), but my confidence in the admin team's logic and dilligence assures me that I have nothing to worry about.

Good job, thus far, Admin team!

KDR 4 to 1 and up is a generally a hacker or a sniper camping and not helping his team! SO the hackers are run and gunners! And use tanks as well!

4 to 1 is just your average (or below average) heavy assault player.

We can already stop here. Not all hackers are smart and the most effective shortcut to deal with hacking is to exclude the not so smart hackers from the game.

Well I have seen super fast invincible Libs, ESFs and Harrassers this week (that all could hit you 100% and see you even cloaked and under a bridge!) and a flying MAX (that could hit you 100%).

sometimes they are clever about it sometimes they don't care if their umpteenth account gets banned and sometimes they get caught out due to a mistake on their part.

i've only ever encountered two hackers, one was super obvious and didn't give a rats that we all saw him doing it openly(terrain collision hacking, possibly teleporting).

another dude was also terrain hacking just after a base cap as an infiltrator, he was gunning down my allies one by one in an open area, he was cloaking to get under the terrain after killing someone unseen and then would find someone else to shoot at. his ultimate undoing was when a friendly lib came into land and he fired off a few rounds from his PDW while under the terrain. tracers don't jsut happen to sprout out of the ground 30m or away from players after all.

i taunted the dope to ensure he would kill me so i could get his name. i reported him and moved on.

If encontered a few but most didn't influence the tide of battle.

They are becoming more bold.

I was on an alert and the fight came down to one base vs to win or tr could hold and force the tie with nc

The control point is in this valley with high rocks overhead. And the spawn is across a bridge . You know the one I am talking about.

Well long story short. 12ish tr are on the control point. Vs heavy warps through the wall guns everyone down in 2 seconds and dissappears through the floor.

Needless to say vs won the alert , because we couldn't get back into the point .

I don't care if they hack on some backwoods point. It's when they influence the outcomes of alerts and large battles with bs.

If you youtube videos of people hacking, you can see there's at least one hack program that gives you the ability to turn off 100% accuracy. His aim average would blend in with the rest of the legitimate players, and nobody would be the wiser, except for all the obvious headshots. Trust me, for all the people out there saying that hackers aren't that clever, they don't have to be. The hack programs do everything for them. You'd be amazed what some of those programs can do, and how subtle they can make it.

My brother has used a ESP hack in Battlefield 4 for almost 2 years has never been caught/banned. Since the hack's overlay doesn't show on video captures his YouTube followers think he's a Pro player. Whenever someone complains in-game he just tells them to go git gud.

this... there's no such thing as a careful and focused hacker... hackers cheat because they're lazy scuts

Honestly, when you see that large and more than rich companies like Blizzard-Valve having troubles to deal with hacking...Well, you can't really blame PS2 Admins who work in a smaller company with (i guess) less ressources.

Perhaps should they communicate a bit more about their anti-hack software if they have one. Shouting loudly "WE UPDATED OUR ANTI-HACK in the last patch" may be good. Good for their customers who will feel a bit more confident about what's happening around them, and it may make some people hesitate to use their hacks/scripts.
Perhaps should they also monthly give the total number of cheaters who were banned.

I suspect that is a reality common to an awful lot of youtubers in an awful lot of games.