> 24 23 22 21 20 19 18 17 16 15 14 13
> | O X X | | X X X O |
> | O X X | | X O |
> | X X | | X |
> | X | | X |
> | | | |
> | | | |
> | O | | |
> | X O O O | | O |
> | X O O O | | O O O |
> 1 2 3 4 5 6 7 8 9 10 11 12
> score -3:-3, cube=1, pip count: O-145, X-139
> O to play 2-1
> [ USRobots played 13/11 24/23 ]
> 1.) Was my actual move the best?
There are so many possibilities here that look plausible
that I would not expect to find the best move over the board.
In fact, sometimes in positions like this I can't find the
right move using rollouts until I've looked at if for a
couple of months, since the best move may not occur to me at first.
My first reaction is to play 24/23 10/8. I think this is
better than 24/23 13/11 because:
1) The 11 point is not such a valuable point, being 6 away
from the 5 point. The midpoint isn't vital either, but
I think since it provides better coverage of the outfield
and blocks 6-6, it is a slightly better point to own.
2) 10/8 leaves only 1 outfield blot, subject to 2 hits (6-4),
while 13/11 leaves 2 outfield blots, subject to 4 hits
(6-3, 6-6, 3-3) of which two are tremendous rolls for X.
3) The fact that 10/8 removes a builder 6 away from the 4
point is not that significant because O has a lot of uses
for sixes already, either running or making the bar point
or even making the two point.
> If not [best], was it at least close, or was it clearly inferior?
In my opinion, it is close. It looks a little worse to me, but
not by much.
> 2.) What features of a priming game are most important when one is
> deciding whether to split one's runners? Magriel's "Backgammon" merely
> states that such decisions are very difficult, while Robertie's "Advanced
> Backgammon" doesn't seem to cover the issue at all.
Whenever you split your back checkers, a principle consideration
is how strong your opponent's attack is likely to be. This is
primarily determined by
1) your ability to defend yourself, which is determined by
A) whether you have an anchor
B) how many blots you have about the board and how vulnerable they are
C) how loose your opponent's position is (i.e., whether you are
likely to be able to return hit ).
D) how strong your offense is, since a strong offense may
deter your opponent from a full-fledged attack.
2) your opponent's ability to build a strong board, which is determined by
A) how strong the opponent's board is all ready
B) how much and how readily material is available for building
3) your opponent's ability to come home after a successful attack.
For example, if you have your opponent primed, then your opponent
will never be able to come home after attacking you. Even if
you have a 4 or 5 prime, the opponent's attack may temporarily
succeed, but then fall apart because of an inability to free
the back checkers.
The above I think applies quite generally. In priming games,
the additional consideration of note is timing. If you will have
great difficulty extracting your rear checkers, and you have significantly
less timing than your opponent, then splitting is very often called for
even in the face of a potentially strong attack by the opponent. By
splitting you hope to gain timing in three ways:
A) the actual movement of the back checkers are pips that you
avoid moving on the offensive side.
B) if you can escape a back checker, you usually pick up about
two rolls worth of timing.
C) if you are hit, you move backwards, and you may gain even
more timing due to fanning.
On the other hand, you may just get closed out and backgammoned :-).
But if you have a strong prime you hope that category 3) above will
give you good winnning chances even when attacked.
On the other hand, if you will probably not have such great difficulty
escaping the back checkers (and your opponent will), and you have more
timing than your opponent, then there is no need to split. Just contain
the opponent, wait for his or her prime and board to bust to splinters,
come around (maybe trapping the opponent off his or her anchor while
you're at it) and win a backgammon :-).
In between is a huge gray area.
In the given position, I don't think priming considerations give
any great motivation for splitting. O looks to have the better timing,
by just a bit. Since the opponent's bar point is open, O should have
escaping chances for some time. Even if X closes the bar, X will only
have a four prime, and O will probably be able to split to the 22 or
21 points if necessary. For all these reasons, I'm not 100% convinced
that splitting is correct, although I would do it.
For me the split is motivated by the fact that O's prime is not so
strong either. With the bar point open, X may be able run to a checker
out and gain both a timing advantage and the advantage of having one
less checker back. It may turn out that neither side ever fully
primes the opponent, and the advantage in this case is likely to be
with the player that can escape a checker. By splitting, O increases
the likelihood of being this player, as well as getting more fly
shots at X's outer board blots. X has a lot of material ready for
an attack, but currently has only a two point board, so I think that
the blitz is only a small danger.
> 3.) (Bonus question) What are the backgammon chances with four checkers
> closed out?
I don't know. I'm sure I've seen this rolled out somewhere though --
I think in _Fascinating_Backgammon_. If no one else posts this I'll
look it up.
> Thanks for any input,
monty on FIBS