RHP was the first site I'd played on and I always suspected that timeouts and random wins by people resigning games on mass (Dustnrogers / Trackhead21), inflated people's ratings. Plus the fact that you have got the top 100 players here now rated over 2000.
I have the same win / lose ratio at Chess Knights, aprox. 60 / 40%, but because there are less timeouts and the top rating seems to be naturally capped around the 2000 mark, I think that site's rating system is more accurate.
I never thought I ever played consistently to 2000 strength at RHP, I think my rating less timeouts / early resignations here would be much the same as it is at Chess Knights, hovering between 1700-1800.
How do you feel your rating at RHP compares to your rating elsewhere?
This is unimportant because the ratings of players at RHP are there to compare themselves with other players within the context of RHP.
Your rating is not like a golf handicap that is transferable to from one course to another. It is more like an average score for this course or chess site.
you can't really compare ratings between different pools of players. a rating exist only in relation to the pool it is drawn from. -for example, BCF-ratings differ considerably from FIDE-ratings, but they are just as indicative of a player's (past!) performance within the pool they're drawn from. the reason why RHP-ratings are within the same ballpark as FIDE is a pure coincidence. there is NO relation between the two.
you could compare current RHP-ratings against former RHP-ratings over some period. well not individual ratings, but statistical averages. in that context you could talk about inflation of ratings, which would be relatively small in it's effect. that's because new players have a gaussian distribution rating-wise. so do old players. -there is no reason to expect that those two groups would differ in statistical sense.
the only thing to affect averages is the issue of rigid initial rating of 1200 for a newcomer. which is actually quite close to the 'real' mean, which would likely be the same as the median which is 1287. that difference of 87pts will create some inflation, but as the provisional ratings are averaged during the first 20 games AND effect of playing against provisional players is halved AND the provisional rating converging within a few games to the hypothetical 'real' rating, the effect should be small. you could measure that effect by monitoring the median through time. I believe that could be easily done by gathering medians from older threads. this topic arises about once a month in one form or another.
having the initial rating not fixed at 1200, but adjusted to the current median, would pretty much stop any further inflation.
I just think its note worthy that Chess Knights has no players rated over 2000 and that it uses a vacation system where each member is entitled up to 30 days off a year without fear of time out. This means that loss by time out is very rare and consequently there aren't the artificial spikes in ratings you see at RHP when a 2000+ player loses by time out to a 1500 rated player.
Regardless of whether you believe the accusations of cheating levelled against me, I have just completed my 100th game at Chess Knights. I have been playing there since my ban here. I have a current rating of 1694 and my highest rating to date was 1823. I have won 76 games [59%], lost 49 [38%] and drawn 5. I had the same win / lose ratio at RHP as Exy, won 1391, lost 701, drawn 180.
However, my rating here was often a good 150-200 points higher because I regularly beat higher rated players by timing them out, or the likes of Dustnrogers or Trackhead21 would suddenly resign all their games on mass and hand out free points. It just seems to me that the time control discipline in place at Chess Knights helps to keep the ratings more realistic.
Originally posted by Steve ExeterFirst and foremost, as has been noted by others, ratings have meanings only within a specific pool of players, not across pools. Nevertheless, there is some interest in comparing ratings, and it would be nice to believe that players over 2000 anywhere have exhibited some degree of expertise at the game.
I just think its note worthy that Chess Knights has no players rated over 2000 and that it uses a vacation system where each member is entitled up to 30 days off a year without fear of time out.
Perhaps no one is over 2000 at Chess Knights because no one there is that skilled. 😲
I play at a lot of sites, and RHP is one of few where notions of rating inflation seem absurd. Indeed, through my first 50 games, it was terribly difficult to get my RHP rating up to my USCF OTB.
Were you banned for cheating?
I think it would be interesting if Russ added a stat to the users' profiles showing the percentage of wins / losses by time out.
I received a section (b) ban by the Game Mods for suspected engine use, if you want to know more then read my other forum posts as this user.
Originally posted by Steve ExeterThanks for the reminder. That discussion remains fascinating, so I bumped it.
I think it would be interesting if Russ added a stat to the users' profiles showing the percentage of wins / losses by time out.
I received a section (b) ban by the Game Mods for suspected engine use, if you want to know more then read my other forum posts as this user.
Originally posted by Steve ExeterPersonally - after 100+ games - I have never had a 2000 rated player or much of any good rated player (1400 and beyond) timeout to me. The only timeout I've received in 100+ games as one occasion to a Mark or Martin something who was rated at 786. I got no points for the win. I think if a player plays near his rating range and maintains his rating at say 1500 or any other rating, and doesn't pick weak players to beat up on to gain replacement points, but only those players in his rating range, and can hold it for 100+ games... In my oppinion he is assured he could play at that rating anywhere (USCF, CCF, etc...) That's without using outside sources, opening books and databases, and computers... That is what I do. And, I beat up 1700 OTB players all the time and it's not that difficult when I take 1 hour and 20 minute timed games. On here, I make all my moves in 20-30 (it varies) in 30 minutes each day. I'm maintaining 1500s without the use of any aid as is unusual in correspondence play. So, I think RHP ratings (with the 10,000 users) is very accurate if utilized correctly. Or else why would I consistenly be beating 1700 players OTB in real life and carry a 1500+ rating on RHP playing basically quick chess? I think these ratings are accurate. Indeed. I know a hand full of players on RHP who hold CCF, USCF and FIDE ratings. It's odd that these players seem to fall within their rating ranges OTB, with only slight differences in rating. As for that other correspondence site, RHP is the most accurate because it has more users.
I just think its note worthy that Chess Knights has no players rated over 2000 and that it uses a vacation system where each member is entitled up to 30 days off a year without fear of time out. This means that loss by time out is very rare and consequently there aren't the artificial spikes in ratings you see at RHP when a 2000+ player loses by time ou ...[text shortened]... the time control discipline in place at Chess Knights helps to keep the ratings more realistic.
Originally posted by Steve ExeterTime outs are very rarare. Out of my 100+ games on RHP, I've only had 1 or 2 players timeout on me. So, statistically, that's a moot point in most cases. Ask other players. You'll find out it's a moot point, or maybe I've just been players who stubborn and stay up on their games. hehe
I just think its note worthy that Chess Knights has no players rated over 2000 and that it uses a vacation system where each member is entitled up to 30 days off a year without fear of time out. This means that loss by time out is very rare and consequently there aren't the artificial spikes in ratings you see at RHP when a 2000+ player loses by time ou ...[text shortened]... the time control discipline in place at Chess Knights helps to keep the ratings more realistic.
Originally posted by Steve ExeterThe chess knights are dominated by computers and players using them for assistance.
RHP was the first site I'd played on and I always suspected that timeouts and random wins by people resigning games on mass (Dustnrogers / Trackhead21), inflated people's ratings. Plus the fact that you have got the top 100 players here now rated over 2000.
I have the same win / lose ratio at Chess Knights, aprox. 60 / 40%, but because there are les ...[text shortened]... between 1700-1800.
How do you feel your rating at RHP compares to your rating elsewhere?
Exy how come your not playing any games here but are complaining about the inflated rating system buggered by time outs and junk?
Fish, Cut bait, or get out of the boat!
RTh
Originally posted by RingtailhunterRTh, I didn't come here to whinge about time outs at RHP and I'm not playing any games because whilst I don't agree with the decision to ban me from playing here, I have to respect it. However, I see no harm in my commenting in the forums from time to time.
The chess knights are dominated by computers and players using them for assistance.
Exy how come your not playing any games here but are complaining about the inflated rating system buggered by time outs and junk?
Fish, Cut bait, or get out of the boat!
RTh
Actually, I was just curious to know if the differing ways in which the two chess sites (that I have experience of playing at) handled vacations had an impact on ratings.
If it's the case that the Chess Knights are dominated by engine users then why are the ratings there so low? If what you say is true I would expect to see at least one player rated over 2200 and many around the 2000 mark but it's not so. 😕