hitting is obvious for me. I also make the 8 point, because if white gets in our prime is stronger and we have less blots. I guess it even increases gammon chances (not sure about that??)

otb I play 14/8* and 13/8 without much thought. Could be wrong but generally think our priority in such a spot should be to get as many in the air as possible, build a nice prime in case they do get back in quick, and hope to maximise our gammon wins over playing to try and get a crunch or a break in the back game.

The priority after hitting is coming out 21/15, not making the eight point.

+1

I'll hit and jump

Question for some of the more advanced players.

In a position like this, does hitting the blot on 17 play a factor in your decision? My own thinking OTB is "I don't like having to take a chance at losing to a backgame, but as long as I have to play it then I want to win a gammon. The best way to win a gammon is to hit as many of his checkers as I can (just don't let him get 3 points in my homeboard). I'm really not too concerned about getting stuck behind his prime. I think he's at least as likely to give me shots at his men as he is to trap me. Plus, making the 8 point helps keep his men stuck behind me." Anyway, that's my thinking.

If you are an advanced player and chose to make the 8 point, was your thinking similar to mine?

For those of you that prefer running with the back checker, are you more concerned about getting stuck back there? Do you not want to give him more ammunition to hit you with by hitting extra men? Is there something else I'm missing?

BTW, I'm just an intermediate player, so I'm not being a wisea$$ here. I really do want to know the thinking that goes into a position like this.

I like the questionning. Being an intermediate player myself, I would probably have made the 8-pt over the board. However, after having analyzed the position, I realize that leaving 4 rolls for White to hit on the 8-pt looks like a small price to pay to avoid some bumps along the road.

If White rolls 3-3 and points on us, for instance, that'll be a little annoying. But the real disaster after 14/8* 13/8 would be the following scenario: White dances followed by 5-5 for Black. It's only one roll, but it changes the whole game, so why not prevent against that unlikely but game-changing scenario, especially if the risk involved is not that great?

If White doesn't hit back on the 8-pt (which is 32/36 rolls), Black will either make it next turn (32/36 rolls) or safety the blot (3 more rolls). Only the infamous 5-5 would leave the blot on the 8 pt, but it's still a roll that plays a lot smoother when we have jumped the broken prime before.

Does it make any sense?

I think I should make it a requirement that all posters consider at least 3 plays
Where is Bill? He's good at back games rumor has it...

Just hanging around waiting for everyone to get a word in.

I think 21/15 13/8* is pretty clear. You have to hit because all your blots are vulnerable, so 13/8*. Now your blots on the 8 and 9-points aren't in much danger, since White only wants to hit with 36 and 35. But you really want to get your back checker out, both because it could get stuck with some bad rolls, and because White might enter and hit you inside, forcing you to hit more checkers on the way out and improve your timing. So 13/8* and 21/15.

Yeah, to be honest, I probably wouldn't be that worried about rolling a 55. Even if I did roll 55, I still like my chances (clearing the 7 point and safetying the other blots isn't horrible). However, I think InsideBackgammon is giving us a hint that there may be a play we're missing.

EDIT: Nevermind uberkuber. It looks like you got the play right.

Bill,

I'm one of people who made the wrong play of 14/8 13/8*. Would you mind explaining what you mean by "because White might enter and hit you inside, forcing you to hit more checkers on the way out and improve your timing".

I'd had also mentioned in my analysis of the position that I'd want the opportunity to hit his other exposed blot to increase my gammon chances. Was this thinking also wrong? (I'm guessing it probably was based on your analysis.)

Is there a general principle I should be looking for when I'm facing a backgame?

I think he meant White's timing. (We don't want to give him a better timing for his backgame.)

I agree with the hit and escape theme here. 13/8* 21/15. leaves only 35 and 36 to whack (vs leaving the blot in the inner board) and gives us the best chance to extend our prime.

Yes I did. Thanks.

If you make the 8-point and thus stay in White's board, he can enter with something like 3-2 and play Bar/22 6/4*. Now you'll probably need to hit one or more blots on the way out, which might be enough to prevent his board from collapsing later. You can avoid all these variations by leaving now with 21/15. Now there will be plenty of variations where White enters quickly and rolls medium-sized numbers and collapses by the time he gets a shot. You'll win plenty of gammons in these variations.

Thank you for the explanation Mr. Robertie, and that does make sense. I've just started the backgame section of Trice's Backgammon Bootcamp, so hopefully I'll soon be getting better at handling these type of positions.

The more I learn about backgammon the more I realize that there is a lot to backgammon that I never really appreciated before. In order to play the game at an advanced level there is just so much to learn. I'm also finding there is a big difference between understanding a concept and really learning something so that it is second nature.

Once again, thank you for your help.

OK I tried to get members to consider a third play but no-one took my hint..
Here's the XG rollout:


It's correct to spring the back checker for reasons Bill has already explained but superior to play 14/9 to block white and eschew the hit which would improve white's timing.
A gnu BG 2-ply rollout had very similar results.
0.319 21/15 14/9
0.260 21/15 13/8*
0.182 14/8* 13/8

Snowie gnu BG and XG all had problems finding this play on double lookahead.
XG2 which is due out soon has a specific back game net so it will be very interesting to see how it fares in this very instructive position.

IB

Here are XG2 evaluations in 1, 2 and 3 ply and a short RO. 1-ply does not find the move, but all subsequent level do. (note also that confidence interval are very similar to the version 1's even though the RO is 4 times shorter)

XGID=-bBbBBBBaA-A-AA-babcbA----:1:-1:1:65:0:0:3:0:10

1. 1-ply 21/15 13/8* eq:+0.317
Player: 58.89% (G:30.87% B:5.13%)
Opponent: 41.11% (G:4.34% B:0.28%)

2. 1-ply 21/15 14/9 eq:+0.256 (-0.060)
Player: 59.21% (G:26.84% B:3.51%)
Opponent: 40.79% (G:5.47% B:0.32%)

3. 1-ply 14/8* 13/8 eq:+0.249 (-0.067)
Player: 55.52% (G:31.32% B:6.02%)
Opponent: 44.48% (G:4.26% B:0.28%)
--------------------------------------------------------------------------
1. 2-ply 21/15 14/9 eq:+0.312
Player: 60.97% (G:27.75% B:3.64%)
Opponent: 39.03% (G:5.50% B:0.23%)

2. 2-ply 21/15 13/8* eq:+0.296 (-0.017)
Player: 58.40% (G:29.95% B:4.44%)
Opponent: 41.60% (G:3.81% B:0.15%)

3. 2-ply 14/8* 13/8 eq:+0.229 (-0.083)
Player: 54.83% (G:31.07% B:5.83%)
Opponent: 45.17% (G:4.31% B:0.14%)
-------------------------------------------------------------------------
1. 3-ply 21/15 14/9 eq:+0.317
Player: 61.06% (G:27.75% B:3.71%)
Opponent: 38.94% (G:5.12% B:0.33%)

2. 3-ply 21/15 13/8* eq:+0.312 (-0.005)
Player: 58.71% (G:30.39% B:4.76%)
Opponent: 41.29% (G:3.61% B:0.21%)

3. 3-ply 14/8* 13/8 eq:+0.233 (-0.084)
Player: 54.81% (G:31.29% B:5.97%)
Opponent: 45.19% (G:4.15% B:0.24%)
--------------------------------------------------------------------------
1. Rollout¹ 21/15 14/9 eq:+0.317
Player: 60.96% (G:28.06% B:3.84%)
Opponent: 39.04% (G:5.11% B:0.18%)
Confidence: ±0.021 (+0.296..+0.338) - [99.1%]
Duration: 1 minute 58 seconds

2. Rollout¹ 21/15 13/8* eq:+0.283 (-0.034)
Player: 57.93% (G:30.28% B:4.57%)
Opponent: 42.07% (G:3.53% B:0.17%)
Confidence: ±0.019 (+0.264..+0.301) - [0.9%]
Duration: 1 minute 59 seconds

3. Rollout¹ 14/8* 13/8 eq:+0.204 (-0.113)
Player: 54.03% (G:30.86% B:6.01%)
Opponent: 45.97% (G:4.27% B:0.18%)
Confidence: ±0.023 (+0.181..+0.227) - [0.0%]
Duration: 2 minutes 27 seconds

¹ 324 Games rolled with Variance Reduction.
Moves and cube decisions: 3-ply


eXtreme Gammon Version: 2.00.157.alpha