Problem of the Week #52: February 28


Cash game, center cube.




(a) Black to play 3-1.

(b) Black to play 3-2.


Note: All ‘cash game’ problems assume the Jacoby Rule is in effect. That is, you can’t win a gammon unless the cube has been turned.

8/5,6/5
13/10*, 23/21 Structure looks best with these

(a) 2 approaches, making the 5-pt or hitting. Making the 5-pt probably can't be that wrong, although it leaves White's blot free while we could have sent it back. If we hit with 13/10*, what about the ace? I eliminate 10/9 because it doesn't do anything constructive. We could slot the 5-pt, exposing us to a double shot, or slot the bar point, giving return 6's and 7's. Or we could anchor on the 23-pt, but we would be 6 pips away from White's 8-pt, which isn't that great. Since I'm not really satisfied with any ace, I think I would make the 5-pt. White still has to safety his blot anyway.

8/5 6/5



(b) Again 2 approaches, anchoring with 24/21 23/21 or hitting with 13/10*. Although anchoring would be nice, I prefer hitting because White has already a back man escaping so sending him back is pretty good, as it also brings a builder at the same time. What about the deuce? That could be 23/21, preparing to anchor on the defensive side, which I find too passive. That could be 6/4, trying to build the home board, but that leaves a double shot, so it's too aggressive. I would bring another builder from the midpoint with 13/11. Sure, it strips the midpoint, but I want to make an offensive point (ideally the 5-pt, maybe the bar point) as soon as possible, so I'm moving my men in the building zone.

13/10* 13/11

No grunch as a page opening issue prevented me from seeing the....problem.

(a) 13-10*-9.

(b) 13-10*, 23-21.

a) The 5 point is nice but I think I go with 13/10*, 24/23

b) 13/10*, 23/21.
I think hitting is better than making an anchor.

Not confident, but I take the five point and the anchor.

I really want to develop those stacks in a) and making an anchor in front of whites already made three point in b) makes things a bit awkward for him.

a) Maybe I'm terrible, but I need a compelling reason not to make my 5, and hitting a checker that isn't his last, and making the 23 when he has the 17, aren't compelling to me. The race and board aren't skewed enough to do something strange.

8/5 6/5.

b) I'm hitting for sure. Making the 21 is ok, but we're not in huge danger of getting blitzed out here. So it's a question of what to do with the 2. 10/8 or 8/6 just seem lame. I'm sure they aren't godawful, but they're not advancing any goals. So it's 23/21 or 13/11. For the longest time, I couldn't decide what I wanted to do, since both advance legitimate goals and don't seem to have a big difference as far as risk goes. Then I realized that 13/11 doesn't gain me a hell of a lot on the next roll if I don't get hit. The rolls it enables are mostly duplicated after 23/21

61 making the 5 can be played to make the 7. 41 making the 7 can be played to make the opponent's 5. 21 making the 9 can be played to make the opponent's 4. 64 making the 7 can be played into the home board. 43 is a bit worse, but it does make the opp 4. And any 3 now makes the anchor. It just looks like 23/21 can play all of 13/11s good rolls about as well and turns some of its random rolls into something more constructive.

13/10* 23/21.

Hitting seems mandatory -- even at the expense of passing up the five point or a high anchor. In both Parts (a) and (b), therefore, Black should hit loose on the ten point. In part (a), we would finish by making the twenty-three point, taking away White's sixes. In part (b), Black has three blots already, so rather than create a fourth by playing down from the midpoint, we would bid for the twenty-one point. This, too, leaves White without a six.

My solutions:
Part (a) - 13/10*, 24/23
Part (b) - 13/10*, 23/21

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. Including the 39(a) tossup, my record at this writing is 55% correct.
17 Correct: 28a, 29, 30, 32, 35, 36, 38, 39a, 39b, 42b, 43, 44, 45, 47, 48a, 48b, 50. 14 Incorrect: 28b, 31, 33a, 33b, 34, 37, 40a, 40b, 41, 42a, 46, 48c, 49, 51.

I side with the contingent that the blot on the 10-pt must be hit in both cases. I am not in favor of making the 23-pt -- I want to make a higher anchor, and that is much more likely with the checkers split. The 9-pt is a valuable location for a builder, so 13/10*/9 is my choice for (a). For (b), I make the bid for the higher anchor while white is on the bar, moving 13/10*, 23/21.

I think this is to White's advantage, since he's the one blocking us 6 pips away.

For part b), having read the other comments, I now prefer 13/10* 23/21. Looks more balanced and you don't gain that much with the extra builder, as explained by TomCowley. I think my play pas too aggressive, as I was impatient to attack/block White. Plus, the balanced play don't strip the midpoint, leaving a more flexible position.

I always feel confident when I find someone who comes up with the same answer as I.

Could we be wrong? Good God, nooooooo.

question a:I doubt between 8/5 6/5 and 13x10-9, Of course having two builders one pip apart is useful, the problem here is that the splitted checkers are also one pip apart, this gives some duplication with a following 34 or 56. Because of this extra consideration i tend to 8/5 6/5.

question b:I think the defensive four is less important than the offensive 5, so thats why i tend to 13x10 here and 23-21 to get a chance making the 4 point next move.

I'm surprised at the wide variety of responses Part (a) has generated. In reply to an opening move, most would automatically hit with a roll of 31 rather than make the five point. Yet in this position, where Black is making only his second move, many now prefer to make the five point. For all the usual reasons (race, tempo, send a man back), I like the hit.

More difficult, I think, is the choice of ones. Making the twenty-three point walks right into a prime, as Uberkuber points out.

There are two reasons, however, why I prefer it over moving forward to the nine point.

First, the ten point is a better place for a blot than the nine. It can be hit with fewer combinations. Consider also what happens when White misses. With a Black blot on the nine point, White can pass with any roll of ten. When the blot is on the ten point instead, White must face a double shot if he elects to run with a roll of nine.

Second, making the twenty-three point takes away all of White's sixes. And, he'll roll one 30% of the time. Here's what I said about this in Problem 48:

This is an often overlooked advantage in the early game. At the start of a game, a player holds only the six point in his home board, and therefore has little expectation that a hit will keep his opponent on the bar. Only 66 is blocked. Nevertheless, an early hit can be a strong forcing play when your opponent has no convenient six to play. When he rolls a six from the bar (10 times in 36), the other die must be used to enter, and the six then becomes a problem for him. For this reason, it often pays to avoid leaving your opponent a six after an early hit.

All of the foregoing notwithstanding, consider how often the consensus/majority solution on this board is right. I should not be surprised, therefore, to learn I am wrong again! Kudos to Bill Robertie for posing yet another good problem.

It is sometimes true, but in this case, White could play his 6's with 24/18 without any major problem I would think.

Other things being equal, the 9 pt is a better place for the blot than the 10.

Thanks, Atrifix. You're right on this.

After posting, I went back just a bit too late to change the word "better" to the word "safer." Offensively, the nine point combines with other builders to make the four, five, bar or nine points more often than the ten point does. Defensively, a blot on the nine point is easier to hit, and does a worse job containing an opponent's backmen.

BTW, I notice your qualification, "Other things being equal," seems to keep your comment neutral regarding Problem 52.

The bigger issue is whether to hit at all. I'd like to hear more discussion of this. If hitting is correct, then let the debate begin as to the best one.

a) 13/10*, 24/23

hitting is awesome, of course, but people often pass up the 23 point when they shouldn't. Its not the greatest feeling anchor, but its better than the 1 point. and it provides some security for the aggressive building you are about to do.

b) 13/10 , 23/21

a) I make the 5-point because I'm almost always wrong when I don't. Here, I think it's a matter of unstacking. Hitting is an interesting option, but it seems like gaining back 10 pips doesn't really offset the cost of not making the 5-point.

b) I hit instead of taking the anchor. Now we're in a position where the 10 pips is more valuable than the point being made. The two checkers in back aren't being threatened, so you're not in a hurry to anchor up. After hitting, I move to the 21 point to make it easier to run or make a high anchor on a future roll, depending on the dice.

Aaron, I don't think the pip count is the major gain when you hit in position a. White is making a bid to escape a checker, and hitting foils that bid. Making the 5-pt is a substantial structural advantage for the entire game, but the immediate need is to keep that checker from escaping! Containing two checkers is much easier than containing one.

Hopefully we can make the 5-pt later, and putting a builder on the 9-pt with 13/10*/9 gives us more opportunities to make this or other valuable points, but we may never get the opportunity to send that checker back again. White will have to leave a blot, and then we have to roll the right number to hit it. We've got that opportunity now, and we must take it -- anyway, that's my view.

That atrifix is one sly fellow.

I would make the 5 point. There are significant differences between this position and a second-roll position. The race is approximately the same but White has made the 3pt which is more of an attacking and less of a priming structure (argues for a counterpriming approach and for not leaving more return shots & blots). Additionally Black has already brought a checker down from the mid already and split. All that seems to point towards making the 5 point.

Of course hitting with a 32 is totally clear.

Why do we need to keep the checker from escaping? And supposing that you hit, what do you think your chances of containing both checkers will be with your undeveloped structure? It's a nebulous gain because white's just going to come back in right away, and maybe making an advanced anchor, rendering your two big stacks rather useless. And if he ends up in the outfield again, you're going to have to strip your midpoint to hit. I just don't like it.

You have plenty of ways to win if that checker escapes. We're only two rolls into this thing, and there's a lot of time. I guess I'm just not worried about it. Besides, your stacks are better suited for blitzing than priming.

I once had a friend tell me, "It's better to make a point than to prepare to make a point."