July 2006 Rating List

Comments

To feel bad for Vishy because all he ever accomplished is being the second strongest player in chess for roughly ten years is stupid. True, everybody who likes him wishes Vish had been a champion, but to me his accomplishments dwarf those of Spassky and Petrosian. It's fun to support our favorites in competition for meaningful titles, yet ultimately other things can speak to their achievement as a chess player a lot more.
Posted by: Yuriy Kleyner at July 2, 2006 23:17

Good news all around - Aronian is consistent as never!
Mig have you talked about Aronian with Kasparov? What is Garry's opinion about him?
Posted by: Ando at July 3, 2006 00:29

I think it is better to let the GM,IM,FM titles as they are now.

A very good effect would be to have two new titles, both should be very difficult to get:

1)SGM (Veselin, Vladimir, Gata, Michael Viswanathan, both Peters & Co)

2)CL (Bent, Svetozar, Boris, Ulf, Viktor, Lajos Jan & Co)

where of course a CL could be a SGM too
(Anatoly, Vassily, Garry of course when he'll return) but a SGM still not a CL (Levon).
Posted by: marcolantini at July 3, 2006 04:41

I think just raising GM norm to, say 2650 will do the job.
The main idea should be : keep number of GMs at nearly 200 ppl point.
Also, they could get back the title from long-term inactive players, or from those who drop their rating below , say, 2500.
Posted by: Ando at July 3, 2006 05:15

they will _never_ raise norm requirements. What do you say to those who complain that their tournamen result would have gotten them a norm a month ago, but one month ago there wasn't any tournament within 500Km in which they could have participated. Just "tough sh***"!?

Creating a title of SGM sounds good to me, as it is common practice already with regards to the top 20 or so...
Posted by: Albrecht von der Lieth at July 3, 2006 06:55

Albrecht, they can just announce 1 year beforehand that theyre going to raise the norm, then no one will complain like you said.
Posted by: Ando at July 3, 2006 07:12

I was a little surprised to see Afromeev in at 2620, despite being "only" an FM. I know he would have had to get over a couple of rating norms to get his IM or GM title, but to have been up in the 2500s for the last 4 years, and to have risen like this, and NOT to have gotten either of these titles seems a little careless. He is also practically geriatric (52), so I think it would be a nice gesture for Ilyumzhinov to just give him the title out of sympathy. :)
Posted by: Costello at July 3, 2006 07:19

Ando,
yes, really a fantastic idea to get back the GM title to Larsen and Gligoric as they drop their rating below 2500.
Posted by: marcolantini at July 3, 2006 07:35

marcolantini, I do not insist on that, that was just a suggestion. :)
If you consider that a the title "GM" is something that means "he is good or he has been good some time in the past", then you should not take back titles.
If it means "this guy is good", then you should.
I like the second way. Larsen or Gligoric have their permanent place in chess. Nothing, including taking the title "GM" from them, will erase their great games or will put shadow in them.
It's like removing inactive players from rating lists. You can argue that Fischer with his 2780 should be in second place now. Not showing him in list is somehow equivalent to taking his rating from him, saying "it is not actual now". So why should you not take title, isn't it gained by gaining the rating, which has became "not actual"?
Anyhow, raising the norm AND taking titles away are orthogonal. You can do both, none, or any of them.
Posted by: Ando at July 3, 2006 08:12

Why mess around with the titles?

The simple fact is there are far more stronger players around these days than there used to be. For many reasons - computers, increased competition and preparation.

Just because there are more GM's around now does not mean a 2500 player these days is weaker than a 2500 ten years ago. If anything, the level of chess has increased due to the reasons stated above.

I'm an IM by the way, made my title ten years ago, have been around 2400 ever since, and that seems about right to me...

Maybe there is a case for a new title higher than that of Grandmaster.
Posted by: sa at July 3, 2006 08:13

Ando you'll agree that the meaning "this guy is good" is already in his ELO rating.
So the GM title should mean something different, may be exactly your "he is good or he has been good some time in the past".

:-) If you got time try by yourself to make a pattern for the FIDE mail that could be sent, something of kind and polite like
"Dear mr.Larsen or dear mr.Gligoric you surely have permanent place in chess but please, from tomorrow do not speak of yourself as a chess grandmaster becouse..."
I imagine how happy they will be to receive...:-)
Posted by: marcolantini at July 3, 2006 09:26

It's always reassuring to know that whenever Kasparov is surpassed, even in an arbitrary statistical sense, even by a single point, we'll be able to heal the pain by reading reminders of his past greatness.

Posted by: greg koster at July 3, 2006 09:57

I for one am outraged. A historical change in the ratings which has a lot of relevance to Kasparov and which no information on Kasparov's past accomplishments has any relevance to and yet Mig still mentions Kasparov in his post. Some things never change.
Posted by: Yuriy Kleyner at July 3, 2006 10:55

Hey, here is a relevant question: at what point did the use of ratings become so commonplace?
Posted by: Yuriy Kleyner at July 3, 2006 10:58

The FIDE rating list has been corrected.

Magnus Carlsen and other participants were missing some games from the Bosnia 2006 tournament.

Carlsen is now #31 with 2675.
Posted by: Xaurus at July 3, 2006 12:33

Carlsen's rating is absolutely unreal. If I'm not mistaken, Kasimdjanov was rated 2676 when he participated in San Luis. Tromso was another 2700+ performance. Taking all that together (which more or less summs up to a 2720ish rating of Carlsen at the moment)leaves one wondering how much of an advantage Aronian actually has in their candidate match(assuming those ever will take place)...
Posted by: Albrecht von der Lieth at July 3, 2006 15:01

Inclusion of Kasparov's achievements seems reasonably relevant. The man casts a shadow over modern chess, what are you going to do? Every outstanding achievement is going to be measured against Kasparov's achievements. He is the gold standard, and will likely remain so for some time to come.
Posted by: macuga at July 3, 2006 16:09

Inclusion of Kasparov's achievements seems reasonably relevant. The man casts a shadow over modern chess, what are you going to do? Every outstanding achievement is going to be measured against Kasparov's achievements. He is the gold standard, and will likely remain so for some time to come.
Posted by: macuga at July 3, 2006 16:09

Last November 30, Magnus turned 15. How does he compare with other great players at the same age?

Top chessmetrics-rated players

On their 15th birthday:

2608 Kramnik
2594 Kamsky
2532 Bu
2514 Judit
2514 Pono
2513 Malakhov
2512 Kasparov

On their 16th birthday:

2672 Kasparov
2652 Kramnik
2613 Kamsky
2596 Fischer

(Per Jeff Sonas' site.)
Posted by: greg koster at July 3, 2006 16:20

If a person goes to school and earns a PhD or doctorate, he never loses the title.

DNA was not really discovered until 1952. so very little was known about it. say someone earned a PhD in Biology in 1950. he knew nothing of DNA. now everyone in high school knows all about DNA. should all the high school kids get a PhD in Biology. Should all the old PhD lose their doctorate title.
Posted by: tommy at July 3, 2006 17:30

Just to clarify, as of my Chessmetrics calculations through 1/1/2005, the all-time strongest players on their 15th birthday were:
#1 Judit Polgar 2626
#2 Sergey Karjakin 2614
#3 Bobby Fischer 2614
#4 Vladimir Kramnik 2612
#5 Garry Kasparov 2603
http://db.chessmetrics.com/CM2/SingleAge.asp?Params=199510SSSSS3S000000000000111000000000000015180

And the all-time strongest players on their 16th birthday were:
#1 Garry Kasparov 2701
#2 Bobby Fischer 2680
#3 Vladimir Kramnik 2661
#4 Ruslan Ponomariov 2656
#5 Judit Polgar 2654
http://db.chessmetrics.com/CM2/SingleAge.asp?Params=199510SSSSS3S000000000000111000000000000016192

These numbers are intended to be directly comparable to present-day FIDE ratings, so the current FIDE ratings of Karjakin and Carlsen should be roughly the same as their Chessmetrics ratings. I know lots of people would like to see the Chessmetrics lists brought up to date with the present, so you can see where Karjakin and Carlsen (among others) currently stand. My main excuse is that Chessmetrics doesn't bring in much income and so it has to take a backseat to more mundane details like a paycheck. But I am hoping to set it up soon so that it can be easily updated each month with newer data.
Posted by: Jeff Sonas at July 3, 2006 18:18

Just to add to Greg Koster's and Yuriy Kleyner's posts: I mean, lads, it's time to switch up the medication here. Surely putting the current spectacular form of Topo into an historical perspective (versus Kaspy) is of relevence. Or should we indulge our sensitivities by completely revising the last thirty years? Jeez.
Note to Tommy, just to extend your riff on academic titles, there are titles higher than PhD, in many countries it is "Professor", in some Germanic countries "Prima Dozent" (interestingly, when I lived in Switzerland I was Dr. Dr Costello (MD PhD), and my wife was Fr Dr Costello, despite not owning the title; perhaps spouses of GM title holders should be Mrs GM Kasparov; one for Ilyumzhinov to consider....). It may be reasonable to add an extra layer of removable title; Candidate for example, could be awarded to a top 10 placer, and removed when they fall out of the top twenty. I also like the idea of CL nomenclature as described above (presumably Chess Legend). We do tend to focus on people's current rating as a measure of their status, an acknowledgement of someone's past status might not be unreasonable.
Posted by: Costello at July 3, 2006 18:33

Jeff--

On your Chessmetrics site I clicked on "View Age-based lists" and posted the results above. Your post evidently contains updated info.

Yes, you should be on a payroll!
Posted by: greg koster at July 3, 2006 18:37

For more than 10 years I have heard the term "Super Grand Master" used in chess literature. It seems past due for FIDE to actually create the title for the very strongest chess players. The criteria for awarding it is open to debate, but clearly it should be based on rating and longevity, and there should be no more than a couple dozen.

For example - award SGM to any player who goes above 2700 and stays there for 50 games, or 2725 for 50 games, whatever.

Point of trivia - I don't think Fischer actually made GM on points, I think he got the title automatically when he qualified as one of the final 8 zonal candidates in the World Championship cycle.
Posted by: superpatzer at July 3, 2006 20:04

It makes no sense to tie requirements for the SGM title to a fixed rating. Tying a requirement to a fixed rating, whether 2600, 2650 or 2700, ensures that it will eventually be eroded and devalued by rating inflation.

What we should do, instead, is use relative ranking. So, GM could correspond to the 95th percentile of active tournament players (norms could be set to 95th percentile performance, plus 100 points). The SGM title could correspond to the 99.5th percentile. This ensures that title requirements will withstand rating inflation over time.
Posted by: macuga at July 3, 2006 20:28

I think the time is at hand for a new category of Grandmaster that would be applicable to 2700+ players. I guess a title like "Super Grandmaster" would be allright, but I just can't get around a 2800+ player being called a Super Duper Grandmaster?
Posted by: chesstraveler at July 3, 2006 20:46

Man, Navara came out of nowhere. I've barely heard of the guy and he gains 61 points in one quarter!! Anyone know anything about him? I don't even know what he looks like. His name sounds...Spanish?
Posted by: dutchdunce at July 3, 2006 22:32

All this Hyperventilating about rating points. Elo ratings were never intended as anything other than a very rough tool.

An elo rating-point spread of twenty points predicts that Player A should defeat Player B in a TWENTY-GAME MATCH by ONE GAME. (1-0 with 19 draws, 2-1 with 17 draws, etc.)

And folks really want to talk about a one-point elo rating advantage!? What does a one-point elo advantage work out to? [Math help requested.] A predicted ONE GAME edge over a 400-GAME MATCH?

Topalov is currently 70 elo points ahead of Kramnik; but bettors wagering that Topalov will dominate Kramnik-Topalov by the elo-predicted margin of three games [3-0 with 9 drawn, 4-1 with 7 drawn, etc.] will get more action than they can handle.
Posted by: greg koster at July 3, 2006 22:45

Navara--

On Olympiad Board One for the Czech Republic Navara scored 8.5/12, including wins over Svidler and Beliavsky. A Chessgames.com poster says David Navara is a logic student in Prague.

"Navara certainly puts the definition crazy genius into another dimensions.... Seeing his behaviour, you would send him to a hospital, but reading his articles you see how intelligent and sensible person he is."

"...and the oddity of the boy's behavior, his clothes, his walking up and down with one finger streched out in one hand and two in the other, his freaky benign smile frozen on his face, was only matched by the overall ineluctable impression of "niceness" one got from him."
Posted by: greg koster at July 3, 2006 23:00

Jeff Sonas has a great site, but I have a couple of quibbles with Kasparov's numbers.
First, his performance at the 1978 USSR Championship was 50%, not +1, which should lower his rating by a few points.
Second, the 1979 Banya-Luka tournament, which propelled Kasparov into stardom, was held April 11-25 according to Jeff's site, while Kasparov's bday is April 13. Now, since the heading says:
"the all-time strongest players on their 16th birthday were:",
the Banya-Luka tournament shouldn't be considered in the calculations.
Posted by: mikey at July 4, 2006 00:19

A one-point or even a twenty-point advantage roughly translates to: "I managed to play one extra game somewhere which you didn't, and therefore I will stand ahead of you in the ratings this month. Next year, I will play five more games or do slightly better than expected and as a result my rating will be higher. Which will lead to an infinite number of monkeys thinking that our relative chess strength has changed."

Koster, where is your quote from? Navara sounds like Bobby Fischer and Misha Tal's love child.
Posted by: Yuriy Kleyner at July 4, 2006 00:40

Jeff,

Thanks for all your help. You mention that Carlsen and Karjakin's current FIDE ratings should be the same as for chessmetrics. I heard that FIDE recently changed their system and was wondering what major difference in the ratings remain. For example, I was under impression that in your ratings, unlike FIDE's, number of games played over 2 years matters, as well as the fact that the games expire.

Greg I believe got the information from your old chessmetrics site--the new one does contain exactly the info you posted.
Posted by: Yuriy Kleyner at July 4, 2006 00:46

Carlsen's unofficial FIDE rating has already risen to a current ~2685 (at 15 years 8 months). With a good performance in the Norwegian championships and at Biel he could hit 2700 before his 16th birthday.
Posted by: simsan at July 4, 2006 04:00

marcolantini,
your arguments are ridiculous.
Do you HAVE to send an insulting letter to someone as he loses his title? Do they send letters when Larsen is dropping his rating? When Larsen loses ELO points, do you send him letter saying that he is out of top ten, then top 100, then send him a letter saying that he is playing like a patzer and is already 2500, etc.?
You might say that you don't like a system like that, but, before posting ARGUMENTS against it, remember that if gaining title is done via performing good in terms of rating, then, losing the title because of a bad performance cannot be logically contradictious.
Posted by: Ando at July 4, 2006 05:31

Yuriy--
Quotes are from posters on Navara's Chessgames.com page. Not the Bible, perhaps, but who would make this stuff up?
Posted by: greg koster at July 4, 2006 05:32

Sidenote: Why is it so damn difficult to copy and past in this blog???

@Ando:

"...remember that if gaining title is done via performing good in terms of rating, then, losing the title because of a bad performance cannot be logically contradictious."
True, but it doesn't follow from it either. Calling upon the rules of logic is double-edged anyway because noone really knows how we should construe our everyday-kind of notions about valid inference into formal language.
E.g. "ex contradicito quotlibed" usually strikes one as quite contraintuitive but is perfectly valid...
Posted by: Albrecht von der Lieth at July 4, 2006 06:16

I understand Navara is autistic and his behavior should be seen in that light.

Why is it hard to copy-paste? Never had any trouble myself. ctrl-c, ctrl-v.
Posted by: Mig at July 4, 2006 06:25

Albrecht,
I am not creating a metaphysiscs system, so I don't need to give absolutely correct syntatical proofs of what I say. My statement means that whatever LOGICAL argument you bring, I will reformulate it in terms of ratings, and, as ratings are point of change, while titles are not, there will be a contradiction. I did it twice with marcolantini's arguments-I think my arguments make sense. Meanwhile, if you say "we like to think of a GM as of someone who has been/is a great player, not of someone that is playing good at this moment", nobody can say anything against this.
I really don't like introducing SGMs, because, as someone mentioned, Super-Duper GMs will be needed next. Making SGM dynamic(only top-ten) and FM, IM and GM static doesn't make sense as well(to me at least).
Well seems like almost everybody likes the idea of just raising the norm, so maybe it is the most politicaly correct solution.
Posted by: Ando at July 4, 2006 07:34

@mig:

If I have reached the point of hittin ctrl-c/v, I'm fine as well.
The difficult part is marking a paragraph (or marking anything at all in a controlled fashion). If I try that with the mouse, usually either the whole document is marked, or some "Posted by: whoever at July 1, 2006 1:23" footnote from the very top of the page...
It's quite confusing as there seems to be no recognizable pattern in what actually gets marked. Sometimes nothing at all, sometimes everything _above_ the point I start marking (even I want to mark down), somtimes just a footnote from somewhere above...
Maybe something to do with Internet Explorer?

And I _know_ to handle my computer - I'm not some 90-year-old who, being told to move the mouse around the screen of his 'puter, lifts up the mouse to actually move it around the physical screen :o)
Posted by: Albrecht von der Lieth at July 4, 2006 09:18

Here's something that went mostly unnoticed: 12-year-old Hou Yifan gains ONE HUNDREN AND NINETY points, to shoot up out of nowhere into the 8th place on the women's list. She's rated 2488, ahead of both Kosintseva sisters, Lahno, Galiamova and Zhu Chen.
Posted by: Alex Shternshain at July 4, 2006 11:11

If my memory does not fool me (and correct me if it does) then Tal, besides Karpov, have also crossed the 2700 mark in the early 80s, although for a short period of time.
Posted by: edu at July 4, 2006 11:11

I don't know if this applies to a Mac, but on a Windows XP computer using Internet Explorer there is something unusual about the selection method on this blog. (Copy/paste works fine, but selecting is normally done simply by holding down the left mouse button and dragging the mouse across the lines one wishes to select. On this blog, that usually highlights the section above where the mouse is located. Odd.)

There is no problem on the home page of http://www.chessninja.com --that works just like most pages on the Web.

But the Daily Dirt pages do work differently.

duif
Posted by: Duif at July 4, 2006 12:29

The strange mouse select behaviour in the blog text starts immediately below the point where the links on the left hand side of the page end.

Strangely those links are placed at the bottom of the html file. It's beyond me how the layout of this page is achieved.
Posted by: simsan at July 4, 2006 14:27

Thanks for the Navara info--

"A possible future star is David Navara from the Czech Republic, already a GM at age 17. David is autistic and before each round you can see him nervously pacing up and down the hotel car park while waiting for the bus, apologising to anyone who comes near him and avoiding eye contact by always looking down at the ground. However, when he gets to the board his demeanor changes. Although his head remains bowed he stops looking nervous and plays confidently and fast."

Markorr.net
Posted by: greg koster at July 4, 2006 15:43

Sounded like the writing came from a very well written piece on Navara and I wanted to read it in its entirety. Unfortunately, it is from various random postings. Fascinating stuff nonetherless.
Posted by: Yuriy Kleyner at July 4, 2006 16:52

I want to confirm the crazy behavior when attempting to select text on this blog. Those who have complained about copy/paste difficulties are correct in their complaints.
Posted by: Arthur at July 4, 2006 19:45

It's a publishing system with a hand-configured template. The source code you see has little to do with the actual code that produces the page. I see that Internet Exploder has trouble with the text. It even appears blank sometimes, interestingly enough. If you use a compliant browser like Firefox everything works fine, however.
Posted by: Mig at July 4, 2006 23:06

Indeed, seems to have something to do with the links to the left...

... a very, very VERY crude workaround...

Mig, you could try this: you can read out the height of the comments-body div with documents.all.comments-body.offsetHeight. (if that doesn't work, give the comments-body a unique id="something" id and try again documents.all.something.offsetHeight)
This should return a value in pixel. This value should be a multiple of "something" since the comments-body is composed of x lines of text, each the same height (possibly with some x because of this form here at the bottom. But you can figure this out via pre-calc math --> linear system of equations)

Then, you can write a very simple javascript function, to be inserted at the of the links-div that document.writeln("")s as many times as you found ot that multiple of offsetHeight is. Manually putting in a few hundred 's works, at any rate.
Posted by: Albrecht von der Lieth at July 5, 2006 05:33

whoops, that tag didn't come out right. It should be document.writeln("");
Posted by: Albrecht von der Lieth at July 5, 2006 05:35

whatever - you know, that br-html tag ;o)
Posted by: Albrecht von der Lieth at July 5, 2006 05:36

MIckey back in the top 10! Oh yeah!

Ah, that feels better...
Posted by: Babson at July 5, 2006 08:48

"The minimum Elo for a GM title should be 2600 (more?) and the norms raised correspondingly. It's gotten to the point that fewer than 10% of all GMs ever even reach the top 100."

Nice, but it doesn't solve the underlying problem--rating infation. It looks like it takes the same rating to get into the top 100 now than it did to get into the top 60 in 2000 or the top 10 in 1970. This can easily be solved by averaging the ratings of everyone in, say, the top 50, and keeping this average constant from list to list.

Of course, since FIDE's policy seems to be more GMs and ever-higher ratings, even something that simple is probably a pipe dream.
Posted by: Daniel at July 14, 2006 00:31