| |
BGonline.org Forums
Question about the 1973 Cooke/Dwek/Martyn match
Posted By: Timothy Chow
Date: Monday, 4 April 2011, at 4:06 p.m.
I just discovered that the local university library has a couple of old books on backgammon. The first one that I've checked out is Championship Backgammon by Barclay Cooke and René Orléan, published in 1980. This book analyzes the first few games of a match that took place in London in 1973, between Barclay Cooke and Walter Cooke on the American side, and Philip Martyn and Joe Dwek on the British side. As one might expect, the play and the analysis are way off the mark by modern bot standards, so I'm not going to say much about that here.
What is most interesting to me is the format of the match. Curiously, the book does not clearly lay out the rules for the match anywhere, but I was able to infer most of them from Cooke's commentary. Barclay played Joe and Walter played Philip. The most striking thing, in my view, is that they did not play a "match to n games"; instead, they played a fixed number of games (40 games each) and the score of each team was simply the sum of scores of the two team members.
Another somewhat odd feature of the setup was that the players were kept in the dark about the scoreboard of the other pair, except during breaks. That is, for the first 15 games, Barclay and Joe had no idea what Walter and Philip's scoreboard looked like, and vice versa. At the end of 15 games, they took a break and were able to compare notes. There was a second break after the end of the 30th game. Furthermore, the breaks did not occur at the originally scheduled times; originally there was supposed to be just one break at the halfway point, but they decided to try using duplicate dice and this made the games take longer than they anticipated.
Match play strategy under these conditions is an interesting question. A couple of years ago I toyed with the idea of having backgammon matches with a fixed number of games, but someone (on rec.games.backgammon I think) convinced me that it was a weak idea, because if one player took a runaway lead then you would likely conclude with a series of boring games where the trailer tries desperately to win an undoubled gammon from the outset (since any double would be dropped to preserve the lead).
In the London match, the Americans did have a strong lead after the second break. However, during the final 10 games, apparently the Brits did not try the undoubled gammon strategy; instead, they just cashed their games quickly. The book does not give the transcripts of the later matches or even the scoreboard (it concentrates on the first 8 games of each player), so it's hard to tell whether the cube strategy was correct or not. Moreover, it's not completely clear what the correct cube action is when you have imperfect information about the score (since you don't know how well your partner is doing in the other room).
Anyway, I'm surprised that for such a high-profile match, these sorts of details were apparently not thought through carefully. Was this sort of setup for a match ever repeated? Was there ever a time when it was common practice to play a fixed number of games in a match, rather than playing a "match to n games" as is standard today? Finally, were the remaining games of the London match ever published anywhere?
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.