Problem of the Week #38: November 22


Cash game. Black owns the cube. Black on move.




Black to play 2-2.

9/7*(2) 5/3* 4/2

I think that Black's goal is to make White crash his prime (by keeping some back men trapped behind his own blockade) while, at the same time, he wants to keep his blockade (a broken prime for now) as long as possible.

With that in mind, I don't like 9/3* 5/3*, because with 2 men on the bar for White against a 4-point board, he will be happy to dance and keep his prime while Black has to roll his blockade forward too soon. On the other hand, I tend to like 9/7* (3) 5/3*. Black now makes a 5 prime and slots the front of the prime. White would still have 2 men on the bar, but now it's against a 3-pt board and a Black's blot. If that blot is hit, Black hopes he dances a few times and pray that White doesn't roll a 6 for a while. If White enters on the 1 or 2 points, he can't jump the prime with one dice.

The thing is, Black has the timing on his side in a prime vs prime battle because he can roll his prime foreward, while White can't roll it farther.

My pick: 9/7* (3) 5/3*

These two statements seem contradictory. Black can't keep his prime rolling AND allow white to crash. If black's prime is rolling, then he's putting white on the bar, and if white is on the bar, white cannot crash.

Black wants to freeze his prime by putting his own checkers on the bar. I would make manticore's play with very little hesitation. It forces white to hit most of the time that he enters, and it will give white a chance to roll some 4s, 5s, and 6s (well... hopefully not 6s if he enters on the 3 point) and eat up what little time white has left. It's really only going to take about 2 rolls before white will be forced to break the bar point, so white is not very likely to escape over the 5 prime coming off the bar before he crunches.

Yeah manticore's play looks like the business.

When Aaron W. and ChrisV agree on a play, it's the best like 99,99% of the time, so I give up right now. Sigh!

About my statements, I know it can seem contradictory, but what I meant is that Black could still roll the prime forward without breaking it, while White is already side by side with Black's ace point anchor and, therefore, can't move it forward and will have to crush it unless it can escape some back checkers. But I agree with Aaron that moving Black's prime forward is not desired because of the risk of putting too many White checkers on the bar, thus stopping White from probably crushing in a few rolls.

9/7* (2) 5/3* 4/2 makes a lot of sense.

Question: If White fails to enter on the 3-pt and Black has a chance to make the 3-pt, should he complete his 6-prime? Would he then have the edge in a prime vs prime battle? Or is it too dangerous that White fans a few times and regain the edge in the aforementioned prime battle?

Depends on the exact circumstances - has White entered both men, one, or none? - but basically I think Black has to keep a 5-prime and hope White can enter and crash quickly enough. So as soon as Black will have to break the 8 point, he should be looking to make the 3. The only way Black can win the game is have White enter and then be trapped behind the 5-prime and have to crunch. Not allowing White to enter will be fatal and so will not maintaining a prime, so his decision on making the three point is governed by trying to achieve those two goals. Leaving the 1 and 2 points open to enter on but maintaining a 5-prime is likely to be a good compromise usually.

It really depends on the timing. If white rolls an immediate 22, his timing is immediately all gone and I probably make the 3 point at the earliest opportunity. (But that might not even be right... white will be "forced" to use any 1 to move up to the edge of the prime, and I *still* don't think I mind being on the bar.)

If white enters only one checker with a 1x, I leave the blots open if at all possible. Basically, I really, really, really want to be on the bar after this roll. With white having two checkers to escape and very few spare pips on his side of the board, I'm willing to risk disaster combinations like white rolling 23 to anchor on the front of the prime, followed by 6x/6x to escape both checkers and leaving me with 2 behind his 6-prime and one on the bar.

I seem to recall an example somewhere where you had a priming battle in which both players were out of time and about to crunch. The question was something like "what would be your best roll?" And the author concluded that the best roll would be not to roll at all, but it's too bad you can't do that.

Thanks for both comments.

The only way to not roll (or not use a roll) is to be on the bar, basically, like you said Aaron. I mean, be on the bar and dance, ideally!

The thing is, as soon as White escapes only ONE back men (without having crushed yet), he gains a lot of timing compared to Black and then Black is in some kind of danger, isn't he?

There's also the (slight) possibility that both sides will have to crush, which would lead to an interesting end game with perhaps some tricky checker play.

9/7 (3)*, 5/3
It looks like we have to make the bar point to have any shot at stopping white. So that takes care of 9/7 (2)
We want to keep our checkers in front of white's blots so we might as well move the last checker off the 9 point and hit on the 3 point. Hitting on the 3 point will also get white off the edge of our prime. Ideally, we want to make the 3 point and maybe the 2 point depending on where white ends up. If we succeed in building a 6 prime, white may have to crunch on the next roll.

Having read the other solutions and analysis, I like manticore's play better.

Yes, he does get a lot of timing, but even then it's not over for black. It only takes one stuck checker to cause the board to crunch.

If white gets both out, that's usually the game (Edit: but not always, since black has a decent ace point game). But you can never really stop that from happening.

Just for fun, take manticore's play and suppose that white rolls 53 to put one checker at the edge of the prime and black on the bar. Then just pretend black never gets off the bar and see how often white is able to escape both checkers before he has to break his 6 point. Then do it again, but this time giving white an opening 52 so that he's not on the edge of the prime. It's not nearly as easy to get out as it might initially seem.

(Black's winning chances are even better because he has the cube. If white crunches, black probably has a double and white's take/drop decision isn't going to be very pleasant.)

+1

This is my move too and I fully expect white to come in immediately, hit all three of my blots, escape both checkers, and backgammon me.

....and some smart **** will ask why you have not played vs gammon.

That's a little bit pretentious!

Black's first priority in Problem 38 is to prevent White from escaping his backmen. One they're gone, no other priority matters; Black will be trapped in a hopeless ace-point game.

Black's second priority is to keep White moving, so that White's prime will collapse. Unfortunately for Black, his first goal, containing White's backmen, means hitting one or both of them here, the very antithesis of his second goal.

With these objectives clear, it is obvious that the "natural" play, making the three point with 9/7*/3, 5/3, is wrong on both counts. The bar point is far better than the three point as a block, and, if Black makes a four-point board, he's increasing the likelihood that White will get stuck on the bar, unable to move.

So, Black should use two of his twos to make the bar, 9/7*(2).

For the third three Black must decide whether to hit loose on the three point. If he doesn't hit, White will have a direct six to escape. If he hits, White may dance. It's an easy choice. Black cannot give White repeated opportunities to run. He should hit now, 5/3*.

Now comes the hard part: what to do with the last three. Two plays seem reasonable: playing safe, 9/7, and slotting the two point, 4/2. In either case, Black does not mind being hit. Unlike White, he is not favored to enter, and so long as he fans, his prime will stand.

A strong argument can be made for playing safe. The alternative leaves three blots scattered about, and although Black doesn't care if he stays out for a while, he is risking a gammon. Especially damaging is for White to hit the third blot, simultaneously escaping a runner. The problem with the safe play comes when White doesn't hit at all. How will Black stop his own forward motion? With any five or six, Black may be forced to abandon the eight point on his very next roll, perhaps hitting White in the process.

It's a close call, but Black will do slightly better to slot the two point. If White hits, then he may stay out, and if he misses, then Black can roll his prime forward without re-hitting White.

My solution: 9/7*(2), 5/3*, 4/2.

For the Record
I am so often wrong that I like to post my record in these messages. It's kind of a truth-in-advertising thing. I have been answering these problems without the use of a bot, and before checking the excellent solutions of others, since Problem 28. My record at this writing is 50% correct.
Correct: 28a, 29, 30, 32, 35, 36
Incorrect: 28b, 31, 33a, 33b, 34, 37

priority as usual is tempo ... said that, since we're primed, the best option for us should be , dance on the bar and hope white crushes;the only move for doing that is hit, close the 5 prime and let 2 blots inside our prime.
9/7*(2), 5/3*, 4/2.