How close are internet ratings to otb ratings? I’ve heard different things.
Are they at all similar?
Generally, Internet ratings are inflated around 300-600 points, especially if comparing regular OTB ratings to quick chess online, where the person with the best manual dexterity wins.
Yea…I agree, mines inflated.
Gambits RULE!!
yea, nightmarish game the other day when I found myself transpositioned into a Queens Gambit Accepted variation…sigh
Internet ratings are inflated, as a person with a higher internet rating can have a faster computer with faster internet speed; if a USCF expert plays on a slower computer with dial-up and a USCF class C player with a faster computer with DSL, then they play each other online with a G/5 game, the class c player has a better chance (as it is a online game then a OTB game) to win the game then the expert. As the expert computer still needs more time to upload the information, then sending in the information on dial-up, the experts computer will take a greater amount of time off the clock then the class c computer.
The other major reason that the rating is more inflated, as some internet chess groups start the player out with a rating of 1500: as the last time of the debate with new players, with a group of unknowed rating players being in a tournament or section: the new players if they meet each other for the first round – they are given a base rating of 1000. This information is a little old as this was talked back in the 1980’s, with the new programings the base for first time players without meeting anyone before should be the same.
The other reason that the ratings are higher, as most games on the internet are rated as a one-on-one game, when a person can cherry pick the person he/she wants to play. Like yahoo and other internet rating groups a person can start his/her own board, then always give themselves white. So a person with a fast computer, with a fast internet server, able to cherry pick the person they want to play, then able to force the other player to play black, then if the other player plays well – then start to insult the other player, so they quit, or get mad then blunder.
Earnest,
Douglas M. Forsythe, local td
I’ve been giving this a lot of thought lately. My internet rating is, in fact, about 300 points above my USCF rating. This does seem to be the standard. However, I have to disagree with most of the reasons that Douglas has presented.
First, as far as computer and internet provider speeds are concerned. The difference between a slow and fast computer and the difference between a slow and fast connection is measured in milliseconds. A slow connection may have a lag time of 100 milliseconds behind a fast connection. After 50 moves thats only a difference of 5000 milliseconds. Also, most chess servers use Timeseal which monitors and compensates for unresponsive connections. Differences between computers is apparent when crunching a lot of data, but pretty insignificant when transmitting one chess move.
Personally, I always let the chess server pick the colors and I take on anyone who wants to play. Also, I never cheat. My rating difference is still 300 points, so I don’t think these are factors either. At least not in my case.
One factor that does come to mind is the comfort of playing from home. When I play on the internet I can play exactly when I want to (when I’m in the right mood). I can also sit in a comfy chair, turn off the lights, have a fresh cup of coffee and have no distractions. Of course, my opponent is free to do the same. Even so, a 300 point difference in ratings is quite a lot.
A chess rating is actually the result of a statistical analysis of previous games. All statisticians will agree that statistics become more accurate as the size of the database is increased. Because the internet games are free I have over 500 games on FICS (Free Internet Chess Server). Because USCF games are costing me an average of about $10 per game I only have 22 USCF games. On this basis you would think that the internet rating would be potentially more accurate.
After all of this I can only think that the real difference between USCF and internet rating systems must be in the formulas they are using. Maybe the formulas should be adjusted, even if it’s only by adding a fudge factor of +/- 300 points? Since USCF was here first I’d say it’s up to the internet chess servers to bring their rating systems in line with the USCF rating system.
Tony
There is the other reason the ratings can be higher then the USCF rating. Lets’ take a chess server like “yahoo”, they have more members and more players, that start too play chess and then drop out. Very few of these chess players are also active or former members of the USCF. As like ‘CaliforniaTD’ is a active member of the federation, would be with people that would not even give the thought of going to a real USCF over-the-board tournament. If taking yahoo with membership that are not USCF members, they are looking to make the memberships have very common ratings as the federation.
The factor the ‘yahoo’ members have high ratings, if they stay with it for years: they have been in more games with people in the end are just there for only a few games then drop out of the system. When someone like ‘CaliforniaTD’ play with the members that have been in the system for years, he should be able to hold his own, should be able too win more games then lose.
Having ‘yahoo’ with millions of accounts with a chess rating, and the USCF having 90,000 members with only 2/3 of the membership having a rating – asking ‘yahoo’ to be like the USCF in ratings would not be fair. The fact that ‘yahoo’ has more members with chess ratings, would be asking a larger company to be following a start-up company on product service. The other factor, as ‘yahoo’ is a international company – why mimic the USCF elo rating system – as other countries have different systems of ratings then the USCF.
Douglas, you may have a point in that the average internet player has probably never played in a tournament. Still, it seems that the rating systems should be structured so that a player who goes from one system to another should maintain a fairly consistent rating. If this were the case it is true that an average Yahoo player would probably have a lower rating than an average USCF player. There’s nothing wrong with that, as long as the same player ends up with a similar rating regardless of where they play.
If USCF wants to change their rating system to be more in line with the internet chess servers that wouldn’t bother me. I’ll take another 300 points. 
Tony
The rating even between fellow federations are different. Even with nations that use the elo rating system, the USCF is the higher rating based system of them all. With Russia the rating is 250 points lower and most other nations the rating is 200 points lower – does it matter that the rating of a master would be 2000 or 1950. Just give everyone 300 points, would be inflation. For the regular rating of all members (2003 Annual Rating List, page XXV) is 1050; making it 1350 would be a 28.5% inflation, making then a class C player be 1700 - 1900.
The rating system was designed for the help of the directors and organizers. If it was not for the 'Swiss System" there would be little use for a rating as we understand it. If the federationg gives everyone each and every year a 10% inflation in their rating, with inflation at 10% (77 / 10 = 7.7 years) it would be 7.7 years before doubling your present rating. Anyone thats as old as myself know what the value of a dollar was in the 1960’s to what the value of a dollar in 2004.