PDA

View Full Version : Cheaters by Statistics?



Scuzzy
16th January 2007, 03:09
Some of you know I've been working on creating a stats system for my clan's TFC players and I've included a Hidden stats part of it. The system is not complete and is not ready for release at all. However, I ran acrossed some things today that I want you (admins) to look at and tell me what you think.

The main hidden page for stats is located here:

http://72.232.94.154/stats/stats_rank_hdn.php

I want to pull your attention to the last grid on that page, the IRIS/Hidden Player comparison grid.

I created this grid because I wanted to gauge a players ability to fight IRIS against their ability to fight Subject 617... I wanted to see if there was any pattern that some people see of "Player was great at finding the hidden, but stunk at the Hidden."

So, look at the top guy, Leg5122fff. As Hidden he's faced teams of 5 on average. This means as IRIS he's had 4 other guys competing with him to find and hunt down the hidden. Of the 22 times he was IRIS he killed the Hidden, himself, 60% of the time. According to his average damage on the hidden he normally did 71 points of damage on the hidden. (That's all 22 Hidden he encountered, not just the 13 he killed.) Now look at his Hidden stats, he's faced 85 IRIS, been the Hidden 17 times and only been able to kill 14 of them. He's 3 times better at playing IRIS then he is at being the Hidden. As IRIS he's incredible... but as Hidden he really really sucks.

What are your thoughts about stats like this.. and this guy in particular?

(Remember, these stats pages are very very incomplete...)
Scuzzy

ThunderDan
16th January 2007, 05:29
It might just be that he doesn't understand how to play as hidden. I played for a month thinking I had to jump and hit the pounce button at the same time to pounce - which ended up making it not work it I tapped jump first. There is a steep learning curve on the hidden especially right now when most people can't access the training and all you get is the time you happen to have in game. We could end up seeing a lot more people who can get hidden, but just not play as him.

MiasmicAnomie
16th January 2007, 06:48
I wouldn't ban on just stats.

However, imho, stats are a pretty good tool wrt who you ought to be demoing.

The thing that clinches that he ought to be watched is that his k/d ratio is over 2:1 ... I don't think I've seen anyone get a ratio like that legitimately without getting more of the kills as hidden. It might be possible, but I would definitely watch the guy (I mean, reading these stats, he's only died twice as IRIS)

The other side of this coin is that 39 rounds isn't exactly a huge statistical base. If he is cheating, then letting it get to a sample size that would be statistically valid would be letting him pick up the hidden skills that would let him hide it - so it's definitely 'demo this guy', imho.

Scuzzy
16th January 2007, 10:38
Yeah, I wasn't planning on banning based on stats, but it does give me an idea of who to watch I think. I've reset the stats often in the last few weeks, so your right my statistics sampling is very small. (I'll probably have to reset them later today again... bah!)

Scuzzy