The story goes I was in the back seat of the car continually asking “Why?” which seemed to disconcert my father. His mother said, “Well, Ronald, how else is he going to learn?”
The great baseball writer, and thinker, Bill James, made a career out of questioning accepted truths. He was castigated by the so-called ‘establishment’, but later was asked to join and accepted a job with the Boston Red Sox, who later won the World Championship.
I have many questions concerning chess. Today I will ask about statistics.
Why is it that when I looked at the pairings for the last round of the recent US Open and saw my friend IM Ron Burnett paired with GM Dimitri Gurevich I could not ascertain which player had White and which had Black? There was a letter, ‘A’ in front of the pairing number. The letter ‘A’ was before every pairing. Why could it not be the letter ‘W’, or ‘B’?
Why is it a player has only one rating? Like a baseball hitter, who has a different batting average vs either a right-handed pitcher, or a left-handed pitcher, it would seem only logical for a chessplayer to have a rating when playing White and one for when he plays Black, would it not?
I would like to know how some of my chess playing friends have done against GM’s as opposed to how they have fared vs IM’s. I would like to know how they, and, for that matter, I have faired against Masters; Experts; class ‘A’, etc.
The great GM Walter Browne, is called ‘Mr Six-Time’ because he won the US Championship a total of six times. I would like to know how he fared against his opponents, not only in the US Championships, but in other tournaments as well. Many years ago there was an article in CHESS LIFE before the US Championship detailing how each opponent had fared aginst the other participants. I thought it one of the most interesting articles to appear in our magazine. It would have enhanced my enjoyment of this years Championship to have known how, for example, Hikaru Nakamura had fared vs Josh Friedel before the last round battle. It would have also have been nice to have known how they have fared against each other, with each color, and how they have done ‘lately’.
As a fan of baseball, I can go to retrosheet.org I can check on a statement like, for example, “Whitey Ford was used by Casey Stengle mostly against the better teams.”
Some years ago while working at the Atlanta Chess and Game Center I met a young couple, Ed and Jillison Parks, who were real chess fans. I gave them lessons and Ed even came to the House of Pain to play in a few tournaments. They enjoyed getting online and keeping up with the Royal Game, just as fans of other games do. Chess needs more fans of the Royal Game. I believe better statistics would enhance the prospect of having more chess fans, and the more fans of chess, the better for those who play the game!
We chess fans can have more, and better, statistical analysis if we demand it.
baconlog.blogspot.com
blog.chess.com/nocab
Actually a player could have as many as three ratings: regular OTB, quick OTB, and correspondence. Some people think there should be additional rating categories; I have proposed Scholastic as a separate rating, but that idea seems rather universally unpopular.
We don’t provide as many statistics as some other rating systems do, and I don’t know what it would cost to provide them. Some detailed reports would require a complex user interface or a staff person publishing interesting statistical analyses.
Mr. Bacon,
At least part of the reason that the USCF doesn’t have separate ratings by color is because color information isn’t submitted when the event is submitted for rating. If you and I were to play a match, and I had Black for all the games, USCF wouldn’t know (or care).
The second part of your question can be answered by the fine program written by Wayne Zimmerle, available at this URL: http://www.gpcf.net/MSA/msa.htm. To entice you to download this program, I merely state that one of the assistant TDs for the 2007 U.S. Championship used this program to figure out the record of each potential matchup in the event. Please note, however, that this program only uses regular rated (or dual rated) games, and only games since 1991.
Alex Relyea
I believe FIDE will soon require color history. which makes me wonder how and when it will be captured by WinTD and SwissSys and the USCF rating system?
FIDE does currently require color history. For a FIDE-rated tournament the TD has to send in a wallchart with colors to the office.
Alex Relyea
I believe that if the event is run with WinTD, Walter can use the PTO file to create the FIDE format file with color information, though he still has to edit it to add some other information that WinTD doesn’t collect.
Adding an option to upload the PTO file is under consideration.
If color information gets added to MSA crosstables, I think the following format would be best (in fact, an improvement on the existing MSA format, in other ways as well):
[code] rtg:pre:post score colors rd1 rd2 rd3 rd4 rd5
- Doe, Jane R:1872:1903 3.0 wbbwb W10 L15 D24 W19 D21
NJ 12345678 Q:1765:1787 - Public, John Q. R:1748:1741 2.5 bwbww D18 W16 L20 L13 W22
IL 13456789 Q:1699:1692[/code]
The following would also be acceptable (though a little inferior to the above, in my opinion:
[code] rtg:pre:post score rd1 rd2 rd3 rd4 rd5
- Doe, Jane R:1872:1903 3.0 W10w L15b D24b W19w D21b
NJ 12345678 Q:1765:1787 - Public, John Q. R:1748:1741 2.5 D18b W16w L20b L13w W22w
IL 13456789 Q:1699:1692[/code]
Bill Smythe
I think I prefer the first version as well, since we can drop the colors column if colors aren’t present in that event. However, the average length of a name in LN, FN format is 16 characters and over half of the active players are provisionally rated, so it might need to look more like this:
[code] rtg:pre → post score colors rd1 rd2 rd3 rd4 rd5
- DOE, JANE ELIZABETH R:1872/15-> 1903/20 3.0 wbbwb W10 L15 D24 W19 D21
NJ 12345678 Q:1765 → 1787 - PUBLIC, JOHN QUINCY R:1748/20-> 1741/25 2.5 bwbww D18 W16 L20 L13 W22
IL 13456789 Q:UNR → 1699/5[/code]
I think Mr Bacon’s basic point – that chess would benefit by the development of richer and more easily accessible statistics – is a very good one. Several sports writers have remarked that it was statistics that “saved baseball” and helped it compete with other more modern sports in the mid-1900s through its unique affinity with fascinating statistics.
Fans love statistics, and the same is true for chess fans. It gives them things to talk about. The popularity of the “Live Top List” of the top rated players (http://chess.liverating.org/) is an example. It would be useful to do more to develop this aspect of chess. Of course, it is not so easy as compiling the information costs money. But perhaps simple improvements could be made here.
Collecting color information is a good start. Information about “streaks” (i.e., Player X hasn’t lost with white in over a year) is also something helps “create storylines” that frame an upcoming important game, etc. For instance, it would be very interesting to be able to look up quickly who has the longest streaks going of various kinds in the country (i.e., most wins in a row, most games without a loss, etc.), as well as other list-type information (most rated games this year, most money won this year, greatest point gain, greatest upsets, most wins against GMs, etc.) These would all be wonderful to have available.
TOM
Better yet, include the number of games only in the post-rating, not in the pre-rating (the reader can easily calculate the pre- from the post-). Makes for a much better visual:
[code] rtg:pre:post score rd1 rd2 rd3 rd4 rd5
- Doe, Jane R:1872:1903/20 3.0 W10w L15b D24b W19w D21b
NJ 12345678 Q:1765:1787/10 - Public, John Q. R:1748:1741/25 2.5 D18b W16w L20b L13w W22w
IL 13456789 Q:1699:1692/9[/code]
Bill Smythe
[/quote]
FIDE tracks colors. Here’s Florin Felecan’s page:
ratings.fide.com/individual_calc … 2009-09-01
And on this page we find out that Felecan is more bloodthirsty with Black:
FIDE has only been requiring color information since early 2007.