| |
BGonline.org Forums
Calculating When To Double Or Lose Your Market
Posted By: Timothy Chow In Response To: Calculating When To Double Or Lose Your Market (trichard)
Date: Friday, 26 March 2010, at 12:29 a.m.
For money games, I've found that there are not many times when I feel it would help to do a calculation, even if I had plenty of time to do so. Much more helpful in practice has been the study of reference positions, such as in Woolsey's Backgammon Encyclopedia. If you know these reference positions thoroughly, then you can often compare your current position with a reference position that you've memorized. Call this "experience" and "gut feeling" if you like, but the more reference positions you know, the more accurate your cube decisions will be. If you find yourself making a huge cube error, then roll it out and create a new reference position for yourself.
The two occasions where I've found it useful to calculate are (1) non-contact races and (2) highly volatile positions with potential game-winning hits. In each case, though, what I'm calculating is not equity per se. In races, the thing to calculate is usually an adjusted pip count, which you can plug into a formula. For the last couple of rolls of the game, you might be counting things like rolls that miss or doubles that don't bear 4. Volatile positions are the positions where I find myself closest to calculating equity directly, but often this can be simplified to counting wins and losses.
For match play, I understand the theory but am still a novice in practice, so there are other people here who may be able to give you sounder advice, but I'll give it a shot. It sounds from what you wrote that you actually understand quite well what needs to be done in principle to assess a double. Namely, first you assess the take. I assume you know what's involved here (raw take points and gammon prices and recube vig). Then you need to assess your market losers and market freezers (my term for sequences that make you wish you had not doubled). Deciding whether some sequence is a market loser amounts to deciding whether it will be a take next turn, so again you're reduced to assessing a take.
Your real question, I think, is whether you should be aiming to do these calculations over the board. Most of the time the answer is no; you just don't have time to calculate all this in detail even if you could correctly judge all the gammon chances and whatnot. Does that make the theory useless? I don't think so. What you can do is to study selected reference positions ahead of time. Quiz yourself as to whether certain sequences lose or freeze your market, and check against the bot. Look at the win/lose/gammon numbers carefully if you misevaluate the take. Studying thoroughly just a few reference positions can go a long way to improving your ability to make the right decision over the board without detailed calculations.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.