Look at the User 141849's graph. His name is irrelevant, he hasn't done anything wrong. This is only one example that I've found here and there.
Look at he right part of the graph. Some fluctuations of more than 150 rating points up and down. This shouldn't be possible in our current rating system. But there it is.
Can we trust the rating system, when we see this kind of example?
Even I have had mystical fluctuations but of more modest kind.
I would be glad if I got an explanation for this...
Originally posted by FabianFnasHis graph looks bugged.
Look at the User 141849's graph. His name is irrelevant, he hasn't done anything wrong. This is only one example that I've found here and there.
Look at he right part of the graph. Some fluctuations of more than 150 rating points up and down. This shouldn't be possible in our current rating system. But there it is.
Can we trust the rating syst ...[text shortened]... fluctuations but of more modest kind.
I would be glad if I got an explanation for this...
He only lost games to time outs the last part, so his rating should go down, down, down, down. In stead its going down, up, down, up, down, up....which is wierd and incorrect.
Originally posted by FabianFnasFeedback sent? That's important.
Look at the User 141849's graph. His name is irrelevant, he hasn't done anything wrong. This is only one example that I've found here and there.
Look at he right part of the graph. Some fluctuations of more than 150 rating points up and down. This shouldn't be possible in our current rating system. But there it is.
Can we trust the rating syst ...[text shortened]... fluctuations but of more modest kind.
I would be glad if I got an explanation for this...
Not sure why, but sometimes graphs get bugged and Russ or Chris are quick to fix with feedback.
P-
Originally posted by rvvThat's not an example of the same type.
Another example :- user 435074
He jumped to the ratings all in a sudden and is able to keep that .... Any comments?
My example in the first posting is probable a bug in the rating procedure in the RHP core system. It has nothing to do with the member himself.
If you think it's something fishy with the member - try the fair Play ticket: http://www.redhotpawn.com/fairplay/createticket.php
Originally posted by Daemon SinI also play over 200 games at a time the rating does this. Of course you wouldn't know that playing 5 games at a time.
You spell it: V-E-R-Y R-U-S-T-Y
You're throwing accusations at other people when you're sitting nearly 200 points below you highest rating and at least 100 points below a rating you easily sustained in the early part of the year.
I have been over 600 games, so don't even go there about a 200 point difference. You try and you will find staying with in 200 points is not an easy thing to do.
The trouble with most of you who don't know what it is like to play more than a dozen games or so, is you think you know everything about ratings and how they work.
He Sandbags his numbers purposely, something I don't do. It is just too bad it is not against the TOS!! I don't think it would be all that hard to prove.
I ask you one more thing MR-KNOW-IT-ALL how does he manage to beat two players over 1900 and one player over 2300....I think I was very nice calling it SANDBAGGING!
We are talking about a player here who claims he can't get out of the 1300-1400 range...I say B.S
Look at yourself you've had close to 300 point different in the rating, I don't think you're sandbagging though. In your case I think your like me, just not that good.
11 Jul 10
Originally posted by FabianFnasAlthough I have nothing to do with the development of this site, I'm a software developer and have an idea of what might be going on here.
Look at the User 141849's graph. His name is irrelevant, he hasn't done anything wrong. This is only one example that I've found here and there.
Look at he right part of the graph. Some fluctuations of more than 150 rating points up and down. This shouldn't be possible in our current rating system. But there it is.
It looks to me as if the bars are out of order. They all relate to games that finished on the same day. I'm guessing they may even have ended at the exact same time, if they were automatically timed out by the system (at least one was a tournament match). My guess is that each finished game is associated in the database with the player's rating at the end of the game (the height of the bar that is plotted in the graph.) To show the rating graph, the system retrieves the games in the order that they finished. If they all finished at the same time, the games might be retrieved in a different order to the one in which the end-of-game rating was calculated.