Take at 3 away, 5 away in 5 pt. match here?

[mont...@lycos.com]

I've mostly played 3 pt. matches, as far as match play is concerned,
but I got into this position in a 5 pt. match, while up 2-0:

GNU Backgammon Position ID: 3XYDAQK7ezgAAA
Match ID : MAmgAAAAEAAA
+12-11-10--9--8--7-------6--5--4--3--2--1-+ O: opponent
| O | | X O O O O | 0 points
| O | | O O O O | On roll
| O | | O O O |
| | | O |
| | | |
^| |BAR| | 5 point match (Cube:
1)
| | | |
| | | |
| | | X X |
| | | X X X X X |
| X | | X X X X X X | 2 points
+13-14-15-16-17-18------19-20-21-22-23-24-+ X: me

GNU has X winning about 200 games here, but advises O double and X
take. If the score was 0-0, GNU has it as a clear drop. What is the
theoretical reason for this difference?

TIA.

[Michael Petch]

I'd say the difference at 0-0 VS 2-0 (and the cube decision) is because if X
is up 2-0 and takes and wins he is at Crawford (And I believe this is the
key) and that much closer to winning. If X takes at 2-0 and loses its likely
going to be a tie match in worst case. If he drops its going to be 2-1. I
think the risks of losing are outweighed by the gains in taking the match to
Crawford.

At 0-0 since Crawford isn't a factor its a drop since there is no value in
taking and letting your opponent go up 2-0.

Mike

On 9/4/06 11:26 AM, in article
1157390773.3...@m79g2000cwm.googlegroups.com, "mont...@lycos.com"

[Grunty]

mont...@lycos.com wrote:

Maybe I´m not familiar with gnu defaults - 200 games out of how many?

[mont...@lycos.com]

It was 199 out of 1000, if I remember correctly.

[gammo...@yahoo.com]

It is a simple matter of match equity.

Without going into all the calculations, there is not that much
difference between up 2-1 and up 2-2, but there is more benefit to
being up 4-0 Crawford vs. 2-1. Getting to Crawford is worth the risk.

Being up 2-1 has the downside of making opponent's gammons very
powerful. So there's more reason to avoid that score.

At 0-0 to 3 the loss from losing the game is greater. You lose the
chance to play the game at score 1-0 to 3, where you have a lot of cube
leverage.

Snowie has the effective takepoints as:

2-0 to 5: 19.1%
0-0 to 3: 25.3%

You should read Kit Woolsey's article on all the scores in 5-point
matches.

[Michael Petch]

I must be getting slow in my old age. Your post was good, however I have to
admit when I read it I had to think to myself "That's exactly how someone
else I know speaks". Happened to notice where you are posting from. Nice to
meet you :)

On 9/4/06 4:26 PM, in article
1157408807.4...@m73g2000cwd.googlegroups.com,