Problem of the Week #98: February 27


Cash game, White owns cube. Black on roll.





Black to play 6-4.


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.

21/17* and 13/7 - get another in the sky and threaten the other blot, plus some ammo to help close the board more. Seems way too obvious so I'm probably missing something but certainly what I would do otb.

Hmmm might be 13/7 and 13/9 and just try and close the board out. Generally like plays that work in as many directions as possible but this might just be one where you are better off going for broke.

24/14 isn't obviously wrong to me either. Hell, even 21/17* 24/18 to maximize the chance of hitting the last blot on the next turn isn't obviously wrong. I'm clearly missing the point here.

I don't imagine that 13/7 13/9 is all that bad, since two on the bar is already quite good. But how can I not put a 3rd checker on the bar with a 4 point board and the aroma of backgammons hovering in the air? This makes 21/17* automatic. Then I look at the 11 and 55 rolls and realize that I need to play 17/11 to guard against the 55 game-changer.

And then in retrospect, 13/7 13/9 looks very loose and dangerous if white rolls 55, so I'm doubly glad that I didn't do that.

21-17 13-7 was my answer. Then I read Aaron's answer and realized that I forgot to consider double 5's. Agree that 17-11 instead of 13-7 looks best.

The problem states center cube which made no sense, I assume that the picture showing cube at 2 is correct.

Yes, White owns the cube. I've fixed the original diagram.

I'm settling on my first instinct 21/17* 13/7. 55 is a swingy roll anyway, and it's not that much worse getting hit vs. not getting hit, and it's a 1/36 shot anyway. Against 11, or on a followup with nothing else to do (or something like 63 with 2 things to do), being able to make the bar point in a lot more ways seems significantly nice. And while I don't know how much it matters, having a builder in range of the 1 has to increase backgammons. I think worrying about making 55 play better doesn't compensate for taking the immediate pressure off the other blot and for all of the other times when having the guy on the 7 is better.

21/17* 13/7

My guess is that in this position the key is for black to go for a backgammon. So is the best play to concentrate on hitting the open blots (24/18 21/17*), or getting builders down to close out the board (13/9 13/7), or some combination of the two (21/17* 13/7 or maybe 21/17*/11)?

I believe the Robertie Rule is -- Do What's Harder First.

So I'd lean towards 13/9 13/7 because closing out the board looks harder than hitting the open blots, and with two blots in the air it seems like getting builders down is the highest priority.

Way back when, Barclay told me to bring two builders down whenever there were two checkers in the air. But now I read Robertie, and he says hit!

My solution: 21/17*, 13/7.

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. Grunch: 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 54%.

I like it. This would be my second choice.

We already have two on the bar, so I think hitting can wait. I want that 5-point so I'll bring two checkers down.

Would hitting a 3rd man increase our gammon chances (already quite high!) by that much? I don't think so. Plus, there's no hurry, White has still 2 men on the bar so we'll probably have the opportunity to scoop those blots later. We'd like very much to make our 5-pt asap, so bringing as many builders is possibly the idea here. If White then enters on the 5-pt, we're going to hit him loose and repeat until we secured the 5-pt.

My pick: 13/7 13/9

Interesting situation. Black is trying for the gammon here. The best way to get it is to try and prevent white from getting an anchor. I really don't think playing 24/14* does a whole lot for us. White already has 2 on the bar.

I'm going all out with 13/7, 13/9.
This gives me a lot of chances to make the 5 point and I also have opportunities to hit on the 1 point and keep white from making the anchor.

13/7 13/9... same reason a couple others have said, we can hit the blots later, we're only gonna backgammon if white runs bad at getting off the bar anyway (both in the next couple rolls and after (if) we get the other two on). Also with builders on 7/8/9 we have an easy job of putting him back on the bar if he rolls a 5 and of attacking the blot on the 1 before he makes the point. I don't think we should fear a 55, it sucks of course but I think we can take the risk of a one-time 1/36.

In general it seems for our gammoning and backgammoning chances getting a third one on the bar now is a lot less important than increasing our chance of closing out before he makes an anchor

I'm gonna do 13 - 7 and 13 - 9
I really don't want white to get the 1 point.
Slotting where i have gives me the most common rolls/numbers for hitting white in BG if white fails to cover on the next roll.

Not sure if i made this clear, so after reading what everyone said after posting my answer, and reading what i said above. I have no intrest in the 5 point, i slotted to get white off the 1 point. If white makes the 1 point it/they/he/she white is in it for the whole/rest of the game

One reason not to go all out for the shutout by playing 13/7, 13/9 is the immediate 14/36 chance that White will anchor on the one or five points. When he does, the blitz will have failed, and Black will be scrambling to saftey five blots. The more balanced play, 21/17*, 13/7, gets a rear checker moving while putting a third White checker on the roof. Should White be successful in making his anchor, Black will be much better off with this play. And if White doesn't make an anchor, well, then the blitz is still on.

Even if he makes the anchor on the 1, which is 13 of those 36 (obv 55 is awful), he still has a checker to get out, and it seems like the majority of rolls either send a second checker back up to give us time to make the 5-point/a 6-prime, make the 5 point to help keep him on the bar, or make at least the seven point and usually only give him a 3/36 to hit us ((5-3 or 5-4) and 5-5)

We've got 4-3/4-2/3-2/2-2 to make the 5 point (or 4-3 can doublehit which I think is better now that we aren't trying to blitz away his chance at an anchor) and 4-4 to both hit the 17 and make the 5 point

5-2/6-1 to hit the 14 or the 17 or to make the 7 and 8 together (not sure which is best),

4-1/4-6 to hit and make the 7, 4-5 to hit and I guess make the 8

So that's 17 rolls that seem very good even if he makes the anchor and the other rolls aren't much worse - it appears we can basically always do two of - make the 7, slot the 5, move the 24 checker up into one-die hitting range/make an anchor.

And then two-thirds of the time he doesn't make the anchor and we're in real good blitz shape.

Disclaimer - I'm sure I'm worse than most of you - my snowie error rate on gnu is ~10-12, almost hope there's some bad logic in this post so I can get corrected

Correction: 12/36 (any 1, plus 5-5)

After White anchors, Black's game plan is to leverage his superior board and run. By hitting a third checker now, Black begins this process of running, without necessarily foregoing the shutout. In Robertie's MODERN BACKGAMMON, he discusses this kind of blitz position. His principle of "non-commitment" is invoked to explain why the balanced play wins more games overall. Fully committing to the blitz is fine when it works, but wins fewer games overall because it so often fails.

BTW: With my modest 50% success rate in solving these problems, I'm afraid we're both in the same boat! Unfortunately, I'm as likely to be right here as not.

Not hitting another checker (21/17*) is an error as it reduces both our gammon and backgammon chances. After that 13/7 aims another builder at both the open 5 point and the blot on the ace. We can accept the payoff of a return 55.

IB

this is a situation I feel I make mistakes in and don't capitalize on as well as I could. My instinct is to hit, but after thinking about it I really like 13-9 13-7. We want to prevent white from making our five point with all we've got so we should slot it as soon as possible, no biggie if we get hit there because white has a weak homeboard and we have extra builders. we can still hit when we've made the 5 point

At first I liked the hit at 21-17, 13-7. Reading answers and reconsidering. 13-7 & 13-9 now looks stronger to me. Double fives changes the complexion of the game no matter what, but if white rolls one five, black should start a war for the five point. He will- eventually- be a favorite to gain the point with 13-7 13-9, and will also have a great chance to hit white's other open men later on.

Some interesting points come out of that thread:

- Gammon vs Backgammon: at this stage of the game, should we factor in backgammon possibilities in our selective process? If the answer is yes, I guess it favors hitting a 3rd checker.

- Protecting against the 5-5 joker. Should it weigh heavily in our decision? Even after we play 13/7 13/9, a 5-5 from White would alter the game considerably, but White wouldn't become an instant favorite I think. I guess he would play something like B/20(2) 6/1*(2). But the game would still be up for grab (Black may even remain a favorite), as Black should re-enter quickly. So it's not like the non hitting move is that risky.

I guess this is where I differ with most of the others who think 55 isn't that much of a worry. You'll have one checker in the air and 4 blots. It looks like it goes downhill very, very quickly. It's important not to over-weight the disaster scenario, but it's also important not to under-weight it. Going from winning backgammons to losing gammons is a gigantic swing.

I see your point, makes a lot of sense. But if 13/7 13/9 wins substantially more games/gammons than a safer play, it might offset the risk and then more, but it's a big IF.

Really interesting problem.

I don't get it sorry. just learning.
lets say we play 13-7 13-9
white rolls and dances
we roll 4-3 is there anyone who would make the 5 point?
isn't hitting on the one point mandatory?