Stuart Katz, MD wrote:
> Usually a player will delay doubling at -2/-2 because they want to
> give their opponent an opportunity to lose their market by waiting too
> long to double. Of course this strategy cuts both ways.
BTW, this is a key part of the "proofs" that it is correct to double
at the first sight of a market loser, because "the opponent will never
make the mistake of losing his/her market". If you throw out this
assumption, then, as Stu points out, you can gain by waiting to double.
There is ANOTHER way to gain by not doubling at first sight of a
market loser, and this second opportunity seems to manifest itself more
often than many would guess. There is an example in the Ortega/Kleinman
book ("Cubes and Gammons at the End of a Match") involving even top level
players. I saw one at the Indiana Open (where Doug Roberts was the 'lucky'
recipient of his opp's error). That other path is:
Your opponent may drop a double that is actually a take.
In simple terms, if you offer an opponent a cube at the -2,-2 score when
the position is close to equal, very few will drop. If you wait until
you are a significant but not overwhelming favorite (maybe between 60 and
65% game winning chances) you will sometimes find that your opp will drop!
Of course, if you wait too long (until you are more than a 70% favorite)
and your opp correctly drops, then you have lost your market, which is
why Stu said waiting "cuts both ways".
On the flip side of the coin, an argument for doubling immediately (even
without sight of market losers) is that you then simplify your thinking to
just trying to win the game (under DMP conditions) with checker play. If
you are constantly trying to decide "should I double yet", it could hurt
your concentration. And there is also the extreme case where you actually
FORGET that the cube is alive. This tactic also enters in post-Crawford
cubes. Doubling immediately may be a SMALL error, but forgetting to double
may be a LARGE one. Making a small error consistently might leave you
better off in the long run than taking the chance that you will make a
large error by forgetting to double.
The bottom line on the -2, -2 doubling debate is that it doesn't make
much difference unless someone REALLY screws up. Some (like Rob Maier) are
of the opinion that it receives WAY too much discussion here, and wastes
time which could be better spent attacking REAL backgammon problems. (But,
should we really be listenting to Rob??)
c_ray on FIBS