Friday, November 10, 2006

Choosing what cards to play

Whatever the sport, it seems that achieving Boston's aspirations always requires going through New York. With the US Chess League first round playoff matches decided, the discussion of lineups can now move from general to specific.

Poker offers a useful framework: Some people play their own cards; others play based on what they believe their opponents are holding; and the best players make their decisions based on what they think their opponents believe they are holding.Playing your own cards - Especially given that they hold draw odds in the match, this is fairly straightforward for the Blitz. The two GM lineup of Christiansen, Perelshteyn, Martirosov and Krasik rates out right at 2400. If ratings mean anything, then LarryC is overdue for a win against Charbonneau. That, combined with the fact that the Knights don't really have anyone who matches up well with Perelshteyn on Board 2, means that Boston should be favorites to advance to the Championship just based on the top boards. Should the GMs surrender a half point, they can still fall back on the lower boards where both Martirosov and Krasik have been able to score points against higher rated competition.

Playing based on your opponent's cards - In setting their lineup, New York has to assume that they will be facing the two GM lineup. Their most likely winning strategy against it is securing a half point on Board 1 (surely they can't expect Charbonneau to defeat Christiansen three times in a row) and sweeping the lower boards 2-0. Given this, one has to expect them to choose the two 2300 players who they think are most likely to win on Boards 3 & 4 from among Hess, Molner, Bonin, Privman, and Shahade (I know they've been using Herman on Board 4 from time to time and that he did beat Krasik earlier in the season, but I think it would be very risky for New York to give up the 200 rating points on a board they must win). Hess and Molner are not a feasible combination (on Boards 3 & 4) based on ratings and a few combinations work only if Krush doesn't play Board 2. Given this, here are some likely New York lineups:
  • Charbonneau, Krush, Bonin, Privman - 2409
  • Charbonneau, Krush, Hess, Shahade - 2408
  • Charbonneau, Krush, Molner, Shahade - 2408
  • Charbonneau, Hess, Molner, Bonin, Molner - 2398
  • Charbonneau, Hess, Molner, Privman - 2394
If the Knights are not committed to the Pascal-Irina "coupling", I wouldn't be surprised if they chose to sacrifice Hess on Board 2 to make Bonin Molner a force on Board 4.

Playing based on your opponents beliefs about your cards - Here's where things might get interesting for the Blitz. If they think there is a good chance that New York might be weak on Board 2, they could consider a lineup of their own which trades off some strength on Board 2 for more power on the lower boards. One possibility would be Christiansen, Foygel, Riordan and Krasik (avg. rating 2389). IM Foygel ought to be able to handle Hess as well as Perelshteyn would. And just in case the Knights do keep Krush on Board 2, keep in mind that Igor did beat Irina convincingly when they met in Week 8. In return, Boston's hottest player would get a shot on Board 3.

Would I go with this lineup? I'd think about it! Yea, but would I actually submit it to league headquarters on Sunday night? Fortunately, that's Globular's problem, not mine.

Of course, the poker game doesn't really end there. If New York thinks that the Blitz might actually put Foygel on Board 2 then surely they will keep Krush there. This is turn will influence Boston to stay put with the two GMs, at which point the Knights need to think about Hess on Board 2 again ... and on and on it goes.

No comments: