Originally posted by RavelloHar de har har. But I did do a decent study of the relationship
Yeah,as a BigCats' member you have built a solid experience about suspicious graphs uh?
between graphs and engine use even before our little spat in the Cats.
It was obvious the engine users were more consitent and flat,
like IM, his shot up to 2300+ and just made a more or less horizontal
line. If you look at mine or most others, its fuzzy, little peaks sticking
out, win, lose, win, win, lose, lose, lose, draw, etc.
that makes for a spikey appearing graph.
But the real IM's or an engine user would both have games like
win, win, win, win, win, draw, win, win, win, win, lose, win, win, win, draw
something like that and the profile shows it pretty clearly.
Also the profile of someone who is really, say, 1600, and plays that
way for a few dozen games or so, then gets pissed he's not advancing,
then uses shredder or something, and all of a sudden, the profile
goes from spikey like mine or yours, to win, win, win, win, win, win, etc.
That has to look suspicious.
Just as an example, look at Jean Hebert's graph. A real IM, but
even so the upward travel when he started still has spikes in it.
Of course now he left and all the games are timing out but the
spikes show human play.
Originally posted by sonhouseOh my God.
Har de har har. But I did do a decent study of the relationship
between graphs and engine use even before our little spat in the Cats.
It was obvious the engine users were more consitent and flat,
like IM, his shot up to 2300+ and just made a more or less horizontal
line. If you look at mine or most others, its fuzzy, little peaks sticking
out, win, l ...[text shortened]... spikey like mine or yours, to win, win, win, win, win, win, etc.
That has to look suspicious.
Originally posted by sonhouse
It was obvious the engine users were more consitent and flat,
like IM, his shot up to 2300+ and just made a more or less horizontal
line.
The line became horizontal because he was the highest rated player on the site, only ever lost one game and because the ratings graph only shows the last 300 games! The entire graph will show a mountain followed by a plateau since IM was facing opponents with a considerably lower rating and only gaining 1 or 2 ratings points with each win.
Also the profile of someone who is really, say, 1600, and plays that
way for a few dozen games or so, then gets pissed he's not advancing,
then uses shredder or something, and all of a sudden, the profile
goes from spikey like mine or yours, to win, win, win, win, win, win, etc.
That has to look suspicious.
How exactly is HAL's graph any different than what you've described here?
For the record, I don't think HAL is cheating (yet) as he hasn't played any substantially difficult opponents but I will say that his graph is highly suspicious upon first glance and investigation.
Originally posted by sonhouseWow.......if you bothered looking REALLY at IM's graph you would have noticed that it was flat because he played almost every kind of rating and that almost every opponent was rated a lot lower than him,so lower that he didn't get any points for beating them.
Har de har har. But I did do a decent study of the relationship
between graphs and engine use even before our little spat in the Cats.
It was obvious the engine users were more consitent and flat,
like IM, his shot up to 2300+ and just made a more or less horizontal
line. If you look at mine or most others, its fuzzy, little peaks sticking
out, win, l it.
Of course now he left and all the games are timing out but the
spikes show human play.
I too can have such a flat graph if I start playing people rated under 1000.
Not that your reasonment makes any sense however.
Edit: the canuck one told it quicker.....damn you Darv!
Originally posted by darvlayThe spikes in the curve would not be there if he was using engines.
Originally posted by sonhouse
[b]It was obvious the engine users were more consitent and flat,
like IM, his shot up to 2300+ and just made a more or less horizontal
line.
The line became horizontal because he was the highest rated player on the site, only ever lost one of them and because the ratings graph only shows the last 300 games! The ...[text shortened]... nts but I will say that his graph is highly suspicious upon first glance and investigation.[/b]
I am defending him, I hope you realize.
Originally posted by Sicilian SmaugNot badly, but looking at his win/loss ratio and the average rating of the opponents he faces, I would say his rating is pretty inaccurate and most likely higher. I could be wrong though...
That's his rating last time i looked on the MAP charts a month or so ago. Why? am i badly underestimating him?