The chess column in today’s NY Times has a game played in the U1600 section of the recent National Chess Congress. It is won by the winner of that tournament, David Nguyen. He won all six of his games. His rating going in was 684. His performance rating for the tournament was 1911. He had previously played 11 rated games in 1994. The article states that Mark Glikman, in an email message, wrote that the odds are about “one in a trillion.” It also says that “Goichberg said he would also ask the federation to raise Nguyen’s rating, which was about 1250 after the tournament. The federation changed it to 1600.”
Mike Nolan has stated that he has “found evidence” that my rating was over 2000. My rating ‘floor’ is 1700. It should be 1800. I have been accused of using my 1700+ rating to gain advantage, although I have previously asked the USCF about this, being told it was BC-Before Computer. No one wanted to take the time to do the research, so I let it go. I have now written to Bill Hall, Mike Nolan, and, the man Mike said takes care of these things, Walter Brown, asking that my rating be changed to the 1800 floor. I have heard nothing from anyone, yet Bill Goichberg can ask that this fellow’s rating be arbitrarily and capriciously changed and it happens immediately. Why is that?
I know nothing other than what I have read in the NY Times, but it seems unfair to this person to be given a 350 point rating boost based on one tournament. Thirty something years ago a young man by the name of Charles Daniel entered a tournament in Atlanta for those rated under 2000, or unrated. It was his very first tournament, and he won it, and the, I think it was $1500, big money in those days, that went with it.
In my four decades of playing in USCF tournaments I have seen many players work hard, and play chess that was not rated, and improve their game. Some have won money, in some cases, a considerable amount of money, in each section as their rating caught up with their playing strength. Why should this man NOT be allowed to do the same thing?
Your rating floor was changed to 1800, retroactive to January 2004. I think that change went into the database last Monday, but it took Friday’s rerate ahead of the creation of the January 2010 rating list for it to be reflected in your rating. Your rating for the January 2010 list will be 1800. Your rerated events are now posted to MSA, it shows that your peak rating in the last 12 months is 1808.
I believe that the reason David Nguyen now has a 1600 rating is that he won more than $2000 in an Under 1600 section, which earned him a money prize floor of 1600.
The NYT report was inaccurate in a couple of ways. Of course, Bill Goichberg is no longer USCF President. More importantly, the way McClain worded it suggested that Bill made a special request to the USCF to raise the guy’s rating. In fact, there’s a rule (which applies to everyone) that if you win a prize of $2000, you get a floor which will make you ineligible to win that class prize again. You can certainly debate whether this is a good idea or not, but you should start with the facts. (I have reservations about “money floors,” but I can’t work up much indignation since I don’t believe class players should be winning prizes like that in the first place.)
That’s the trouble with believing everything you read in the newspapers. Every time there has been newspaper or TV coverage of an event of which I have had any personal knowledge, there have always been factual errors in the story. In this case, the request made by Bill Goichberg probably took the form of simply following the rules, which state that organizers of big-money events are supposed to furnish lists of major class prize winners so that USCF can give them a rating floor in the next higher class.
It could be that, in the cases you mention, either (1) the amount won was less than $1000, in which case the rule does not apply, or (2) the organizer did not follow the rules by submitting a list of major prize winners, or (3) these examples occurred years ago, before the rule went into effect.
To answer nocab’s question purely on principle, Nguyen had just showed evidence of 1900 playing strength. That’s different from saying that one had a certain rating, or a certain strength, 20 years ago. Both may be adequate reasons to put in a floor, but they’re different reasons, not directly comparable.
Hold on now, Arti. On several occasions I have shown ‘strength’ far above my rating, an example would be when I won the Atlanta Championship 5-0 as a class’B’ player. I have also shown, I’m sad, but honest enough, to report, ‘strength’ far below my rating on many occasions…
This is only ONE tournament. While on duty at the ACC I had one player who had just won his section tell me that he was ‘busted’ in every game! Fortunately for him, his opponents all blundered, making the proverbial ‘last mistake’…This player showed considerable ‘strength’ in that tournament, no doubt. But just like the bloop hit in baseball that ‘looks like a line drive’ in the boxscore, this gentleman knew he did not play particularly well…
If a players rating has to be changed, it would seem to show there is something very wrong with the rating system, would it not?
From my reading of statistical baseball books, for example: Curve Ball: Baseball, Statistics, and the Role of Chance in the Game by Jim Albert and Jay Bennett, I realize that it is possible for a lifetime .250 hitter to enjoy a month of hitting like Ted Williams. Just because he hits .400 over a short span of time does not mean he is a .400 hitter. The same is true far a chessplayer…
I could understand if, because the rating he obtained was from 15 years ago, USCF were to ‘throw it out’ and give him a rating based only on this one tournament. Then he would be a ‘provisional’ 1900. Although I do not have all the details, I have nothing to go on other than what I read in the NY Times column.
Bill Hall has sent me an email in which he writes, "Also, there may be more information available that may make such a decision not so “arbitrary.”
Until that information is made available, we will have to reserve judgement.
I look forward to ‘cooler heads’ than I tackling this problem!
Well I didn’t take your other tournament history into account because I didn’t know about it; I was just using the info you posted in this thread. But if your performances are “up and down” then they probably average out in the rating formula to your current rating.
When we implement the “title” system that is in the pipeline, your “up and down” performances may qualify you for a good title, whereas someone with the same rating but more consistent performance would not get such a good title, because the title calculation only considers your good results and ignores the bad ones.
If we have someone who’s been away for a long time, there’s a thought that he should get a higher “k” because it’s likely that his strength has changed in the meantime, either up or down. That’s an idea that is not currently implemented in our rating system, but it might be sometime. It sounds like in this case, the player received a “money floor” which is intended to prevent someone from winning repeated big money prizes because they are under-rated.
Floors have two effects. They prevent the player from entering low class sections and winning low class prizes, which the player might not like. They give him the guarantee that his rating will not go below the floor, which a player might like. Whether the player likes them or not, they are not optional, they are not prizes or rewards. They just … are.
The effect of a money prize floor is to raise someone’s floor to a certain level as of a designated event. If that person’s rating is currently below that level as of that event, then it also raises that person’s rating to the floor, since that’s what floors do, establish a minimum below which that player’s rating cannot go.
I think that the person receiving a money prize floor winds up above that floor as of that event more than half the time just due to the ratings gain from the event.
Haven’t we all? I’ve beaten 2300s and lost to 1200s. “Showing strength far above rating” means much more than that – it means demonstrating that one’s average strength is far above the rating. A player who, out of the blue, wins big money in a class tournament would be a likely example.
That’s absolutely typical for players below, say, 2000. I have heard it said that, for a C player, one move in every ten is a blunder. There are no doubt similar stats for A, B, D, and E players, with the percentage being worse the lower the rating.
Years ago, a friend of mine who was in the middle of a typical “quantum leap” from about 1400 to about 1800, lamented that all of his recent wins were due to his opponent’s blunders. I pointed out that a player’s opponent must make a mistake in order for the player to win. My friend responded that there seemed to be a difference in degree between his opponents’ blunders and my opponents’. I would win because my opponent had a backward pawn or a weak square, he would win because his opponent dropped his queen or walked into a mate in one. We finally agreed that, during a quantum leap, a player’s opponents’ blunders do not actually increase in number. Rather, the player just notices them more, because he overlooked them when he was a weaker player.
By the way, that friend is now rated 200 points higher than I am.
Nope. It just shows that no rating system can catch 'em all. Making sweeping changes to the rating system because of occasional anecdotal evidence would be a huge mistake.
Arpad Elo once compared chess ratings to the problem of measuring the vertical distance from a ship deck to a cork floating in the ocean below, when the ship is rocking back and forth and the cork is bobbing up and down in the water.
Better leave it up to the ratings committee to decide whether changes to the system are advisable. They have tons of data available and can easily perform tests to see what the total effect of various suggested changes would be.
I would add that the player might have changed in strength. Since the rating system is not assisted by brain implants measuring one’s chess strength (we had to curtail development during the current USCF lawsuits but it’s scheduled for implementation, sometime after the title system), we currently have to wait for actual game results before updating a rating.
So there’s historical strength, current strength, today’s performance, all the same three things for the opponent, and the intangible factors that if you’re both playing terribly today, the first person to drop a rook will probably lose because then the game may become technical the scope for future tension and blunders will be reduced. There are many gaps between the inputs to rating (historical game results) and the outcome of today’s game.
I found that an improving player, or any strong player, can cause blunders by the opponent because his play exerts a sort of psychological pressure. Fischer’s opponents have said they felt he was literally pushing them off the board, pushing them backwards away from the table. I’ve had the same feeling against some strong opponents, and then I learned to push back, too.
