Updated Ratings!

[Chexhero]

Hello everyone. Below you can view the new updated ratings as of July 2011. The Southern Open and GAYP Nationals are the next two tournaments to be calculated. I am still debating a little if whether or not I should include the blitz tournament, but expect it to get put in there. Just to note- you may see a second PA 2010 calculated on the tournament table, but that was only created to fix a typo and update a change. You will see rating changes at the bottom, but no one's rating from that tournament was actually updated on the ratings table. Anyway, happy New Year and enjoy!


http://icheckers.net/ratings/

[Chexhero]

Actually John, my PA performance did not affect anyone because I was under 20 games. But I did get a pretty good increase

[waynegober]

I am very confused now. The 2011 Dist 8 was a round robin tournament in top division. I tied for first with Albert. My rating stayed at 2007 but Albert's gained 13 points and we played the exact same opponents?? He won our round and with a win and draw but I won against Ed King? He was 1997 before tourn then 2010 after , I stayed without change at 2007 with us both having 19 and tied for first. A 13 point swing with us seems pretty much even though we had different scored with diff opponents. But round robin tie can make that much difference?

Wayne

[Eric Strange]

your rating is calculated round by round, the difference in how u do vs opponents round by round changes your ratings and will effect scoring for the next round.

Although you did end up with the same points at the end the difference in how u did vs each opponent is bigger. Given that you and Larry were two of the highest rated players and he took 2 wins from Larry and a win and draw from you I can see where an extra 13 points would come from.

So in answer to your question yes there can be a swing in points in a round robin where you both tie at the end. 13 points really isn't very much, but when your fighting for placement in the rankings it can make or break you, I like the competitive edge Elo puts on people!

[Chexhero]

I noticed the same thing with me and Jonh Acker in the Ohio State. We both finished with the same amount of points, but because I played a few more higher rated opponents and did well against them, I got a much bigger rating boost, like 40 points more even though I actually had 20 more points than he did coming into the tournament. I tied 2 games against Beckwith, a high rated player and he lost both games. He did better against Millhone who beat me, but Millhone is much lower rated than Beckwith. I also tied Alex a round and he did not get to play Alex at all. So who you play is very critical toward your rating change.

[waynegober]

I understand if we played different opponents. But this was round robin.
When you say each round is scored and an immediate change of rating takes place for the next round? If so, in that type scenerio, I could win say 2 games in round 1 off of a player, knocking him down several points, then another player beat him in round 2 when he has fewer points and get less credit? Seems like ratings should be calculated from same vantage point during the whole tourney, especially in a round robin.

Anyhow thanks for the quick response and your hard work

Wayne

[Alex_Moiseyev]

your rating is calculated round by round, the difference in how u do vs opponents round by round changes your ratings and will effect scoring for the next round.

Very interesting, I didn't know this. If this is true, then it is not ELO at all !!!

Classic ELO system calculate ratings "event-by-event", not "round-by-round". All ratings for all players are static until you finish processing event.

Lets say before event started, my rating is 2430, average rating of all my opponents is 2240, and finally I should have 75%+ (or whatever) in order to keep my rating at the same level or higher. If I play 14 games - this means I should have 10.5 points to break even. If I've got 11 points - I increased my rating, if I've got 10 points - it's dropped.

Thats it. If you "re-calculate" rating from game to game or round-to-round - this is full complete mess, not ELO !

Please, do me a favor -in continuation of this conversation do not call this system ELO -this is wrong and not true at all. I don't know - where you read this and who told you to do this.

Regards,

Alex

[waynegober]

hope this doesn't double post

If 2 players enter a round robin rated even and then finish with even points and one gains 15 rating points (which means quite a bit in the narrow range of Elo) then something is wrong.

I suspect it is the round by round change of rating. It seems like some of the Elo software kept the rating the same for calculations over the course of the tournament and only change rating at end which makes sense. A player is not becoming a worse or better player over the course of each round in a 7 round tourn. and the luck of the draw would seem to be very important in round by round changes.

Wayne

[Alex_Moiseyev]

I suspect it is the round by round change of rating.

Yes, and this has nothing to do with ELO, this is "Strange system" We can use this perfect terminology in future conversations.

[waynegober]

I am not trying to start a war.

An official rating system needs review by several people to catch this
type of thing.

If the ratings have been based round by round it is a very huge flaw.

However, I still appreciate all the effort, but we must get it right if it is to be anything more than a "meaningless list"

Eric and Joe, no criticism of you personally but something is wrong and I have suspected it from rating changes I have seen (not my own, I'm still rated too high according to those I know rated lower)

Best regards,
Wayne

[Eric Strange]

I suspect it is the round by round change of rating.

Yes, and this has nothing to do with ELO, this is "Strange system" We can use this perfect terminology in future conversations.

You are wrong on this Alex, I am following the current ELO system used by FIDE and MANY other organizations. I would do this game by game if checkers is willing to revamp to single game by game scores instead of round?? but round is decently accurate enough.

The current Elo rating of the chess player changes after each game.

This comes from http://www.chesselo.com/

I also read the link to ELO on ACF page for wikipedia and there is NOTHING on full event scoring but also about single game calculations. I would appreciate if you did not come off so aggressive and rude if you don't know what your talking about champ.

I did find this published which makes it easier to understand why Alex is freaking out, Alex understands the actual published 1960's version of ELO which has been adapted by many organizations.

The phrase "Elo rating" is often used to mean a player's chess rating as calculated by FIDE. However, this usage is confusing and often misleading, because Elo's general ideas have been adopted by many different organizations, including the USCF (before FIDE), the Internet Chess Club (ICC), Yahoo! Games, and the now-defunct Professional Chess Association (PCA). Each organization has a unique implementation, and none of them precisely follows Elo's original suggestions. It would be more accurate to refer to all of the above ratings as Elo ratings, and none of them as the Elo rating.

Give a chance for defense before you assume someone making bold statement is 100% right.

[waynegober]

Eric,

The following is taken from the site your posted about Elo after the discussion of one game Elo changes. Am I missing something?
Wayne


The Expected Result for chess tournament is calculated individually as the sum of all expected results with each opponent. The FIDE rule of 400 points has impact on the Expected Result. Before, FIDE calculated the Expected Result in another way. The following 3 examples explain how the Expected Result is calculated.

Example 1: (the FIDE rule of 400 points not applied): Player A rated 2000, played 2 games against Player B rated 2100 and Player C rated 1750. The Expected Result for player A is therefore calculated as this:
Expected Result 1 (rating difference 100, Player A weaker) = 0.36
Expected Result 2 (rating difference 250, Player A stronger) = 0.81
Expected Result for Player A (for 2 games) = 0.36 + 0.81 = 1.17

Merry Christmas and Happy New Year!

Example 2: (the FIDE rule of 400 points applied): Player A rated 2000, played 2 games against Player B rated 2450 and Player C rated 1750. The Expected Result for player A is therefore calculated as this:
Expected Result 1 (rating difference more than 400, the rule applied (for weaker 0.08), Player A weaker) = 0.08
Expected Result 2 (rating difference 250, Player A stronger) = 0.81
Expected Result for Player A (for 2 games) = 0.08 + 0.81 = 0.89

Example 3: (the FIDE rule of 400 points applied): Player A rated 2000, played 2 games against Player B rated 2100 and Player C rated 1500. The Expected Result for player A is therefore calculated as this:
Expected Result 1 (rating difference 100, Player A weaker) = 0.36
Expected Result 2 (rating difference more than 400, the rule applied (for stronger 0.92), Player A stronger) = 0.92
Expected Result for Player A (for 2 games) = 0.36 + 0.92 = 1.28

Calculating New Rating and K-factor

New Rating = Old Rating + Rating Change ... or ...
New Rating = Old Rating + K-factor * (Result - Expected Result)
The K-factor is assigned, and it may range from 10 to 45 for different chess organizations. FIDE uses the following rules for the K-factor:

• The K-factor is 30 for players new to the rating list, until they have completed events with a total of at least 30 games;
• The K-factor is 15 for players with a rating under 2400;
• The K-factor is 10 once the player has reached 2400 and been registered for at least 30 games. Thereafter it remains permanently at 10, even if the player’s rating is under 2400 at a later stage.

Calculating Rating Increase/Decrease

Rating Increase/Decrease is very important for your Rating Change. If your Rating Increase is 0.15 and your K-factor 20, then your Rating Change will be 3. If your Rating Decrease is - 0.20 and your K-factor 10, then your Rating Change will be - 2.

Rating Increase/Decrease is actually not called like this. Officially, it is called in the FIDE Players' Database as "change". But let's name it Rating Increase/Decrease for better understanding. It is an absolute value which can be plus, minus, or zero and is calculated as this:

Rating Increase/Decrease = Result - Expected Result

Example 1: Player A in a 9-game chess tournament scored 5.5 and his Expected Result was 4.0. The Rating Increase/Decrease for player A is therefore calculated as this:
Rating Increase = 5.5 - 4.0 = 1.5

Example 2: Player A in a 5-game chess event scored 1.5 and his Expected Result was 3.5. The Rating Increase/Decrease for player A is therefore calculated as this:
Rating Decrease = 1.5 - 3.5 = - 2.0

Example 3: Player A in a 9-game chess tournament scored 4.5 and his Expected Result was 4.5. The Rating Increase/Decrease for player A is therefore calculated as this:
Rating Increase/Decrease = 4.5 - 4.5 = 0

Calculating Performance Rating

Performance Rating = Opponents' Average + Performance Change

Performance Change is based on the Performance Ratio. If a player scored 9 in 9 games, his or her Performance Ratio is 1.00 and Performance Change +800. If he scores 4.5 in 9 games, it is correspondingly 0.50 and Performance Change will be 0. The needed value is taken from the Table of Performance Change.

Please notice that your Performance Rating does not depend on your own rating but does depend on your Opponents' Average and "how you performed" (Performance Change). Performance Rating is very important for getting the Grandmaster and International Master norms.

In some Chess Organizations (but not in FIDE), the Performance Rating is calculated with "the algorithm of 400": If you win, add 400 to the opponent's rating; if lose, subtract 400, if you make a draw, no change. Then find the average.

[Danny_Alvarez]

I agree with Eric that scoring by round is more accurate than for each tournament at its end.

Scoring after every game would be even more accurate but i suggest we leave them be for now....

cheers
Danny

[Alex_Moiseyev]

I also read the link to ELO on ACF page for wikipedia and there is NOTHING on full event scoring but also about single game calculations. I would appreciate if you did not come off so aggressive and rude if you don't know what your talking about champ.

I did find this published which makes it easier to understand why Alex is freaking out, Alex understands the actual published 1960's version of ELO which has been adapted by many organizations.

Eric, the system which we used in USSR calculated the whole event, they called it ELO in 1980's and it was very accurate system. I do certainly beleive that calculating rating "per game" or "per round" (don't see much difference, but per round could be better) will generate paradoxes.

However, indeed I am not aggressive or rude and always take it easy on anything. You are good

[Eric Strange]

I understand your concerns but the ELO system that I am using is the same system yahoo, kurnik, pogo, and other websites with two player games use.

This may not be the historic version but it is the adopted version for online gaming and gives an accurate depiction on skill level in my experience. No weird paradoxes