Interesting play

[Simborg]

I have to admit this one was a toughie for me. Match to 7, black to play 5-4






The right play is to move both checkers off your opponent's bar point!

[LuvChild]

I know I usually post here with my "feelings" about my side of the board, so this time I'm going to try and be slightly more analytical and take it from both perspectives... I apologize in advance for the length...

The #1 play according to Snowie (from Black's perspective):
Playing this leaves white with a double shot (20 of 36), but it provides better odds for me not getting hit than if I play both off the 22 point (22 of 36) or run one off the 18-point (23 of 36); with the added bonus of white having only 4 rolls that will hit and cover their blot on their 2-point (1-4, 2-4) whereas the number of hit and run/cover options on their loose 2-point increases significantly with either of the other plays (because almost any ace or six could then be used to bring up or cover, respectively, the blot on the 2-pt.) and looking for a way to duplicate white's good rolls has proved otherwise fruitless...

After the #1 play according to Snowie (from White's perspective):
I'd love to hit here and get a gammon on this to be 1-away / 3-away (additional 5% equity), but with black holding a 5-point board/prime unless my roll is perfect do I want to risk getting gammoned here myself in a game I'm winning right now by exposing my blot on the 2-point? Do I have to let Black basically have a pass and try to just race with him... with his anchor on my 3-point is it even really a race- if I don't hit now he probably gets a shot at me down the line here anyway since I'd have to play most rolls off the 9-point to avoid leaving a triple shot for black next roll. ***Knowing (and hoping to avoid) this, I would definitely consider offering the cube after black plays this roll.***

*** After reading your article on gammoned.com this past month I wonder what the proper cube decision is here though... I'd like to think that if I were White it would twist my opponent in knots seeing the cube slide over to him after making the best move (thinking that taking and getting gammoned would end his night); but fear at this match score that since he's 35% favorite to win the game and only 30% to win the match if he drops, that I'm more likely to have him turn it right back at me... I will have to check this out (if you don't have it run already, that is).

My *original* thoughts (as Black)...

"Odds aren't great to win this game, so let's just try and keep it at 3-away / 2-away by keeping from giving up shots for as long as possible and just play 8/4, 6/1; and maybe we'll even get one down the line here too since White has no safety net for their outer board checkers and I am gumming things up by holding his 3-point."

Usually when I make these kinds of decisions over the board and take a look later on my assumptions as to my opponents win /gammon percentages based on an individual move are usually pretty good (i.e.- this move reduces my winning chances but at least it reduces their gammon chances too, etc) even if ultimately it isn't the best possiblke move... here that wouldn't have even been the case (and not even close at that).

I think that's the thing that you say most often (though sometimes indirectly)... it shouldn't "feel" like the best play... it should simply BE the best play.

As an aside, I just got a copy of Magriel's "Backgammon" as an early B-day gift to myself... as soon as I take my PE exam at the end of the month it becomes my new "study material"...
_________________
I judge myself a success not by my accomplishments, but rather by the number of times I fail, but continue to try.

[Simborg]

Excellent comments and thoughts. First, I think it might well be a double/drop with any play. Second, I sent this position to some of the best players in the world, and many of them found it very tough to play. Generally, I only post very tough positions.

Robertie's and Magriel's basic backgammon books are still terrific teaching aids, even though they are over 20 years old. A few of the plays are no longer right, since the computer programs have proven that some of the opening rolls are different from 20 years ago, and it no longer is a "sin" to make your 1 or 2 point early in the game, but the basic theories are excellent.

I have taught over 100 students, and I still recommend those two books.

[LuvChild]

Wish I knew how to attach screen captures using the site's format page, but I just ran the position through GNU-BG and it says that after black's move White has a double; but as I "feared" black has a take. Since I'm not sure how to input the screen capture, here's the anaylsis in print form...


Cube analysis
Rollout cubeless MWC 54.87% (Money: +0.434)

Cubeful equities:
1. Double, take 58.41%
2. Double, pass 60.30% ( 1.89%)
3. No double 57.23% ( -1.19%)
Proper cube action: Double, take
Rollout details:
Centered 1-cube:
0.675 0.136 0.004 - 0.325 0.054 0.002 CL 54.87% CF 57.23%
[0.001 0.001 0.000 - 0.001 0.001 0.000 CL 0.03% CF 0.07%]
Player Jonathan Moffitt owns 2-cube:
0.679 0.138 0.004 - 0.321 0.055 0.002 CL 61.27% CF 58.41%
[0.001 0.001 0.000 - 0.001 0.001 0.000 CL 0.08% CF 0.09%]
Truncated cubeful rollout (depth 11) with var.redn.
1296 games, Mersenne Twister dice gen. with seed 1043203749 and quasi-random dice
Play: 0-ply cubeful prune [expert]
Cube: 0-ply cubeful prune [expert]
_________________
I judge myself a success not by my accomplishments, but rather by the number of times I fail, but continue to try.