Problem of the Week #135: January 22


Money game, center cube, Black on roll.




(a) Should Black double? If he does, should White take?


(b) Assume Black doubles and White takes. How should Black play the following rolls: 6-6, 4-4, 2-2, 6-4, 6-2, and 4-3?


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.

In Part (a), I would certainly double. Since the cube has not yet been turned, the most I can win right now is 1 point, and many opponents would happily surrender that here.

Although the situation is perilous for White, Black has no immediate threats. In 501 Essential Backgammon Problems, Bill Robertie stresses that a four-board board, with a checker on the roof, is almost always a pass. Without any checkers in the air, however, I think this is a take. Black has no immediate threats, and one nice set of doubles will go a long way towards equalizing things for White. White has a few non-doublets that also give him a new point immediately, and after that, who knows?

Part (b) is all about Black’s game plan. For what rolls should he play this as a blitz? And when should he play to make a prime? Should Black play most rolls down from the midpoint, or should he split his rear checkers in an effort to cover all boards?

• 66 — Go for the blitz. 13/1*(2). Afterwards, Black won’t be threatening very much until he can mobilize the checkers on his midpoint, but I like the five-point board more than I like the five-point prime. Even when White enters, he will often be forced to expose a blot with his second die.
  
  
• 44 — I don’t like switching points. Black does not have enough ammo ready to commit to a full-fledged blitz. I play 24/20 to cover all boards, and then 13/9, 13/5 to unstack the midpoint. Running out with 24/16(2) gives White too many options to slot behind Black’s anchor, and thereby build up his board quickly.
  
  
• 22 — Create as many builders as possible for the 7pt or 8pt by playing 24/22, 13/11, 13/9. Alternatively, this could be played the same as 62.
  
  
• 64 — Slot the bar point, 13/7, and then split to the 20pt. Slotting and splitting!
  
  
• 62 — Slot the bar point, 13/7, and then go all out for the cover by unstacking the midpoint, 13/11.
  
  
• 43 — Play 24/20 to cover all boards, and then come down from the midpoint with the three, 13/10.

My Solutions:

Part (a): Double, take
Part (b):

• 66 — 13/1*(2)
• 44 — 24/20, 13/9, 13/5
• 22 — 24/22, 13/11, 13/9
• 64 — 24/20, 13/7
• 62 — 13/11, 13/7
• 43 — 24/20, 13/10

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 53%.

a) Small cube, I think. Decent advantage, but not all that many market losers.
b)
66 - 13/1*(2)
44 - 5/1*(2) 13/9(2)
22 - 3/1*(2) 24/22 13/11
64 - 24/18 13/9
62 - 24/18 13/11
43 - 24/21 13/9

a) Seems like a pretty easy double because while the race is equal, nothing else is, especially gammon chances. Doubling to turn on gammons is a no-brainer. If we somehow already had the cube, turning it to 4 is far more questionable. White has to take though- his lone back checker isn't even blocked from 6s, and if it gets free, he's quite fine. Or if he just makes progress up front while black doesn't escape. Double/take.

b)
66- Blitzing 13/1*(2) seems mandatory. 13/7(2) 24/18(2) can't be that bad, but we can make a 5-point board with a man in the air and no threats on the other side.

43- This one also seems pretty clean, 24/20 and 13/10 for maximum coverage since there's nothing direct to do with the roll.

62- These are where I'm a lot less sure. 24/18 13/11 and 13/7 13/11 both seem reasonable. I guess with zero progress on his home board, you just risk a shot and expect to hit him later if he rolls a 6. Obviously if you cover next turn, that's A+++. 13/11 13/7.

64- Similar spot, although now 13/7 24/20 is sane because it covers the entire board with direct shots (24/22 is lame in comparison). I'm still going 13/9 13/7 on the same plan.

22- Can be played like 62, but getting a free half turn with his stacked points seems like a good idea. 3/1*(2) 13/11 24/22. So many of his entries leave a shot, and if he fans, that's great for bringing more builders.

44- I have no idea. 13/9(2) 24/20(2)? I don't really want to switch with 5/1 and leave a high hole that I have no realistic prospects of filling anytime soon, so I'll just put pieces in decent spots. Unstacking the 13 can't be bad, and making the 20 can't be bad.

Double/drop. white has nothing here gammons abound
66...13/1*,13/7(2)
44...13/1*,24/20
22...13/11(2), 24/20
64...24/18,13/9
62...24/18,13/11
43...24/20,13/10

Slotting lets white hit and escape at once..not good

Grunch.

(a) Should Black double? If he does, should White take?

No, yes. Black has a slight racing lead and a four point homeboard, but he has not escaped either back man whilst White has escaped one of his. White has an easy take if Black incorrectly doubles.

(b) Assume Black doubles and White takes. How should Black play the following rolls: 6-6, 4-4, 2-2, 6-4, 6-2, and 4-3?

6-6: 24/18 (2), 13/7 (2)
4-4: 24/20 (2), 13/9 (2)
2-2: 24/20 (2)
6-4: 24/18, 13-9
6-2: 24/18, 13-11
4-3: 24/20, 13/10.

Black needs a forward anchor and the doubles allow that. The non-doubles ask Black to slot for such an anchor and create a builder in his outfield.

Post grunch view: everybody but me seems as aggressive as hell!

I think this sort of thing would probably have been put down to style pre the computer age but I think the rollouts will probably show me to have been too conservative...

Cube/Easy Take

66 - 13/1*(2)

44 - 24/20 13/9 13/5

22 - 24/22 13/11 13/9

64 - 24/18 13/9

62 - 24/18 13/11

43 - 24/20 13/10

a) Double/Take

(b)
6-6: 13/1(2)*
4-4: OTB I would play 24/20 (2), 13/9(2), but something tells me that 13/1*, 13/9 is the correct play, although I don't like it.
2-2: Lets try something violent and sexy like 13/7, 13/11. Getting hit doesn't seem to be all that bad, but white not hitting on the next turn would be full of win.
6-4: For similar reasons, 13/7, 13/9
6-2: Similar reasons, 13/7, 13/11
4-3: 13/10, 13/9 or 24/20, 13/10. I would probably play the latter.

This is another Escher Position. We see an overwhelming board, on the other side three stacks. Must be a double, and a questionable take. But, is there anything week, difficult to repair for white? How about the race? What about one man vs. two men behind? That’s a good position, where a chess player, who had just learned backgammon, would have problems not going on tilt, loosing this after a double against the levanter. The black advantage is obvious. The white resources are hidden, but visible after a second view. Double, perhaps, but a trivial easy take. I would wait a roll with the double, cashing more with the passes from colonel whiteflag’s pigeon pass next turn.

No double (from a practical Gaby Horowitz PoV).
Trivial easy take.

66. I am not a bg informator, but i remember, that it is always right, to make both bar points with boxes. But here, I think, with activated gammons, it is right to slash away.

44. Connecting our position a little bit and developing four men must be the winner. Making the golden point and a good outfield point.

22 Against the chessplayer, I would play 24/22 13/11 3/1*(2), but against the pigeon 13/11(2) 3/1*(2). The pigeon knows, when he is overmatched, and would quit perhaps. Against the chessplayer, it is another opportunity, to put him on tilt.

64 Once again, for the same reasons as above, I would play against the chessplayer 24/18 13/9 and against the pigeon 13/7 13/9.

62 Oaftsraa, cp 24/18 13/11, p 13/7 13/11.

43 cp 24/20 13/10, p 13/9 13/10.

For these problems, I believe Robertie wants us to assume that we are strong players opposing a strong opponent. Would you double in that case?

And what about your checker play? Against a strong opponent, I assume you like the "chess-player" moves, but I am uncertain what you mean by "Oaftsraa" and "p."

As always, thanks for a considered post.

— Mike

A chessplayer, with intermediate knowledge of backgammon is corrupted for the variance of backgammon. He might overestimate the black position, because it is similiar to a strong center position in chess. But the pigeon, who has no clue about chess, will see probably the strong and best moves as what they are, the sign of true expertise. He might quit. I assumed, that 66 44 had only one solution, giving a lot of equity away with the second best. But for 22 and the others, the difference isn't big. Perhaps playing your opponent, and not the best bot move was also a nice deliberation, i thought.

"Ladyfingers: You raised tens on a lousy three-flush?

The man: That's what it's all about, isn't it? Making the wrong move at the right time."

Nack Ballard: "I will often redouble what I am pretty sure is not a redouble on the chance that my opponent will pass. This is big key to winning matches about which the bots are clueless.”

(a) Black leads 149-151. He has a (much) better structure. Nobody really has threats for now, unless Black rolls 6-6. Sounds like a good double/take for me.

Double / Take


(b)
6-6: 13/1* (2) --> Hope White dances a lot and go for an attacking plan
4-4: 24/20 (2) 13/9 (2) --> Advance + block
2-2: 24/22 (2) 13/11 (2) --> Advance + guard
6-4: 13/7 13/9 --> Prepare to extend the prime, return shots if hit
6-2: 13/7 13/11 --> Same plan as above
4-3: 13/9 13/10 --> Preparing to make the bar point