< Earlier Kibitzing · PAGE 201 OF 203 ·
|Feb-10-18|| ||zanzibar: I have a math/computational problem...
It turns out that given the complete set of games for a tournament, without round numbers, one can reconstruct the Berger table schedule or Berger list.
That means that one can determine the full pairings/round schedule just from the tournament games, even when no round numbers are available for any of the games. No additional information is needed so long as all the colored pairings are available.
The problem with my approach is that I'm not smart about it, instead relying just on brute force. That means the algorithm scales with the number of permutations in the Berger list.
So, for a 4-RR (4-player RR), there's 4! = 24 combinations, which is no sweat. But for a 10-RR there's 10! = 3628800 (3.6M) combinations to try, and there's a noticeable computation time for the solve.
Believe it or not, if one is determined/patient enough, a 12-RR can be inverted:
Bad Kissingen (1928)
That's 12! = 479001600 (479M) combinations to try, only one of which is the solve:
Capablanca, Jose Raul
Yates, Fred Dewhirst
Marshall, Frank James
But that's getting close to the limit of the approach.
This must be a problem for group theory. Is there a better approach?
I don't even know if it's mathematically true that the problem is invertible in principle.
|Feb-10-18|| ||zanzibar: <Fusilli> RE: Averbakh photo|
One also has to beware of copyright issues - <CG> really can only use Public Domain (PD) sources without gaining explicit permission.
So, I usually use PD chess periodicals for historic figures (plus, then I can identify the source of the illustration/photo).
Google image search has a tools feature, which allows a selection on usage rights (where "Unrestricted Usage" = PD). You'll find this setting can sometimes restrict your choices severely. (C'est la vie).
WikiCommons is always a safe choice to check, and here's a selection (however not all the Averbakh's are the chess player of interest):
|Feb-10-18|| ||Fusilli: Thanks, <zanz>, that's very helpful.|
|Feb-10-18|| ||Fusilli: <zanz> Actually, from your Wiki link, there is a photo of Averbakh playing Averbakh vs Parma, 1963|
Beautiful photo, no copyright issues!
|Feb-10-18|| ||zanzibar: I agree <Fusilli>, and funny, I was just looking up the game myself - just by using the position on the board!|
(Black to move after 17.Be3-g5)
click for larger view
(Black to move from the plunger on the clock, the Black king + dsb from guesswork, the pawn on a3 can barely be seen peaking out over the clock)
|Feb-10-18|| ||zanzibar: BTW- the photograph's attributes
<This is an image from the Nationaal Archief, the Dutch National Archives, donated in the context of a partnership program.>
The Dutch have an enlightened attitude as concerns copyright. Their newspaper archive is also wonderful.
|Feb-11-18|| ||zanzibar: RE: <22e Latvian Ch - Riga (Mar 1965)> (and Berger tables)|
From <Shahkmatny Bulletin (June 1965)>'s heading we get our lede:
<Победа М. Таля -
Первенство Латвии, 1965>
<The Victory of M. Tal - Championship of Latvia, 1965>
RUSbase only has 31 of the games (out of the 15-RR's 105 games), and I was able to find 4 more fragments in Tal's magazine. The tournament has bracket dates (Mar 4-26, 1965), a complete RR xtab, but no round numbers (either from the RUSbase games (or CB or 365), or from the Russian literature I found).
My first goal was to create a stubified version which could match the known games colored-pairing, but which would admittedly be uncertain on the colorings of missing games. That is an interesting problem, I think, but very hard to do other than by brute force, which is computationally intractable for 15-players. It may not be possible to do otherwise, but I can't say for sure.
Luckily for me, it turns out that <JFQ> has a page on Tal's games from the tournament, all of which have been preserved. We are both fortunate that <chessical> provided the round numbers for these games.
Game Collection: Tal at the Latvian Championship 1965
Now, having all the games of any given player allows the Berger ordering to be quickly determined, either using the colored-pairing or bye to obtain the Berger number of the player and backtracking.
Having the Berger order allows a full stubified version of a tournament to be generated, with the correct coloring/round number for all the pairing. This can be merged with the available games to get a world-class version of the tournament made for public distribution.
Which you can find here:
|Feb-11-18|| ||zanzibar: RE: <22e Latvian Ch - Riga (Mar 1965)> |
I sourced all the games from the tournament except for these Tal games:
A23 32 (R1) 1-0 Tal -- Zhuravlev
E55 91 (R4) = Mileika -- Tal
A48 34 (R6) 0-1 Grant -- Tal
D35 16 (R7) 1-0 Tal -- Kirjakov
C84 26 (R10) = Kirilov -- Tal
B17 32 (R11) 1-0 Tal -- Skuja,H.
B59 50 (R12) = Klovan -- Tal
E55 35 (R13) = Tal -- Gipslis
B41 56 (R14) = Skuja,R. -- Tal
Can anyone help me identify the sources for these games?
I suspect Hilary Thomas' <Complete Games - 1960-1966> have them:
and having those refs would be great. Though even better would be the primary sources - does Thomas give those in his books?
|Feb-11-18|| ||zanzibar: Oh, and a less important request - what about the actual dates of the games?|
If somebody provides the RD's I'll update the PGN.
|Feb-11-18|| ||zanzibar: OK, this sort of sordid material is generally better handled by <MissScarlett>, so I'm disappointed that <CG> has no games of either <Hilary Clifford Thomas> or <Richard Pentreath> (that I could find anyways).|
We know HC wrote the Tal books, but we might not know he later became <Richard Pentreath>, who..., well, let Spraggett tell it (nice book cover photos of Tal series are there as well):
<Several weeks ago in England the story of Richard Pentreath (age 64) , also known as Hilary Clifford Thomas, and his being sentenced to 35 years in prison began appearing in almost all of Englandís MSM, including the BBC.>
His wrong-doings, and ultimate punishment can be found here:
This looks like it's fairly recent news - so <Missy>, get on it!
|Feb-11-18|| ||zanzibar: H.C. Thomas was also the driving force behind the short-lived <International Chess> magazine (~1979):|
|Feb-11-18|| ||Retireborn: <z> Wish I could help, but the Convekta database, although it has all these Tal games from the Latvia ch, doesn't give any dates, round numbers, or sources. They also misidentify Grave(s) as "Tauve", as I mentioned before.|
Also I'm afraid there could still be a query about the round numbers. When we first discussed it <hemy> told us (from a Latvian newspaper) that Tal vs H.Skuja was a 12th round game, which implies that Tal had a bye in one of the earlier rounds. You also quote Tal himself as "playing in the last round until there were only the kings left, which suggests that Black v R.Skuja was the 15th round, not the 14th.
Tal vs H Skuja, 1965
I do think the order of the Tal games is correct though.
|Feb-11-18|| ||hemy: <chessgames.com> Still waiting for publishing games of Israel Rabinovich-Barav|
<Dec-19-17> <It will become available on our site one day soon>
The zip file with 34 games is in the folder https://www.dropbox.com/sh/eia0mi3p...
It the same folder placed the picture of Israel Rabinovich-Barav and also the photo album (Barav.pdf) including translation to English and Tournament table of Eretz Israel 1945 Championship.
Dr. Ami Barav would like to replace the photo album at http://www.chessgames.com/documents... with this one.
I already posted in the comments (Israel Rabinovich-Barav (kibitz #12)) a brilliant game that was published in many chess magazines, periodicals and in the book in Hebrew (M. Marmorush, "Chess, Tutorial Book to Study Game", Tel Aviv, 1945)
|Feb-11-18|| ||zanzibar: <RB> I think the caption says they were the last to play on the last day - which could be just the resumption of R14's game (which was 56 moves, and almost certainly was adjourned).|
It would come after R15's normal games, or after R.Skuja's R15 game. This would explain why it was the last game to finish.
|Feb-11-18|| ||Retireborn: <z> Certainly when Tal talks about the last round he could just mean *his* last round, regardless of when it finished anyway.|
I think I'd like to ask <chessical> how confident he is that Tal had the bye in the 15th round, or if (like me!) he got the round order from Convekta. Not sure if that's acceptable etiquette on here though.
|Feb-11-18|| ||zanzibar: <or if (like me!) he got the round order from Convekta.>|
<but the Convekta database, although it has all these Tal games from the Latvia ch, doesn't give any dates, round numbers, or sources.>
Which one is it, <RB>?
|Feb-11-18|| ||Retireborn: <z> Convekta present the games in the same order as the <jfq> Tal collection. I've been assuming that this is a correct chronology at least. They don't offer any round numbers or dates, unfortunately.|
|Feb-11-18|| ||zanzibar: <RB> thanks, I see.|
Maybe <chessical> would be kind enough to dig up what source he used originally then - let's hope.
|Feb-11-18|| ||MissScarlett: < so <Missy>, get on it!>|
I'm on the job, so to speak. #crookedhilary
<We know HC wrote the Tal books, but we might not know he later became <Richard Pentreath>>
It occurred that <Hilary Clifford Thomas> may only have been a pen name, but apparently not.
|Feb-11-18|| ||Chessical: CSs Stars "Mikhail Tal Games 1963-1972", pages 94 -100 has:|
"22nd Latvian Chamiponship, Riga 1965".
R.1, Tal v Zhuravliov
R.2, Petkevich v Tal
R.3, Tal v Tauve
R.4, Mileika v Tal
R.5, Tal v Peterson
R.6, Grant v Tal
R.7, Tal v Kiriakov
R.8, Kapengut v Tal
R.9, Tal v Shmit
R.10, Kirilov v Tal
R.11, Tal v H.Skuja
R.12, Klovan v Tal
R.13, Tal v Gipslis
R.14, R.Skuja v Tal
|Feb-11-18|| ||Retireborn: <Chessical> Many thanks for that. That's the book form of the Convekta electonic database, I think. I would not be inclined to trust their round *numbers* though.|
I have the Chess Stars books for Capablanca and they make this sort of mistake for odd-numbered participant tournaments, eg for San Sebastian 1911 they put Capablanca-Spielmann in round 11 (instead of round 12) and all their subsequent round numbers for that tournament are wrong, of course.
|Feb-11-18|| ||zanzibar: Yes indeed <chessical> thanks for that info - it was extremely useful to me because it yielded the Berger ordering аor the players = full tournament schedule.|
I assume I can use the book as a source for the games - though I can't validate correctness of the movelists without the book.
I'm still quite curious where Khalifman (or Solokiov) found the games which I couldn't source. Perhaps they got dolled out in subsequent issues of the Bulletin, or Saks (Riga)? Or maybe they were published in a Latvian newspaper. Or may Chess Stars had access to unpublished sources?
It's not really such a big deal, but having done this much work it really would be nice to dot every И and cross every т!
|Feb-11-18|| ||Retireborn: <z> As far as movelists go, the book is likely to be identical to my Convekta - and I can tell you now that there are no discrepancies with your pgn, even the Mileika game is the same as yours.|
|Feb-11-18|| ||zanzibar: <RB> good news then, thanks. |
Of course I always invite double-checking, which can be a demanding challenge, depending on how deep one wants to go.
|Feb-11-18|| ||zanzibar: <Chessical> oh yeah, I forgot to mention that R3 is Tal--Grave, not Tal--Tauve or Tarve.|
I've commented on that extensively on the game page:
Tal vs U Tarve, 1965 (kibitz #38)
< Earlier Kibitzing · PAGE 201 OF 203 ·
A free online guide presented by Chessgames.com