Problem of the Week #150: August 26


Cash game, center cube.




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.

B/22* seems mandatory. Then, it would be between 13/11 and 6/4 I guess. 6/4 unstacks the heavy 6-pt and starts an inside point, but leaves a direct shot. 13/11 brings a builder for the 5-pt and also leaves an extra direct shot, although it would require White to break his anchor (unless he rolls the fly shot 6-5).

Race is 134-127 favoring Black, and it will be even more after Black hits White. Despite that, the ugly-looking B/22* 8/6 can't be right, can it? I know the QF switch is on and when in the lead, we should disengage, but I cannot contemplate 8/6. After all, it's far from a pure race with a lot of contact left and we need some flexibility. Well, if we hit and are hit back, we still keep our small race edge, so I go with 6/4 for the deuce.

B/22* 6/4

First of all black should hit, because he has a very poor defence, and will be wacked. i've been playing 13/11 and 8/6 a number of times. My first idea was that 8/6 was the right move, resting on its big racing advantage, and based on a previous problem that was discussed in which extreme passive ugly pay was rewarded. I presume that situation lent itself better, because perhaps white didn't occupy the 7point. White will soon have a strong motherboard, and black needs double 1,2,3 desperately , because in two turns he will not be able to slot in his outerboard. But with his huge racing lead he will need yet more doubles to get off the 13 point, in order not to leave a blot to hit with a 6 of white. A very fragile situation. He should better try to strengthen his homeboard when white's position is still weak, and that is now. No problem when white hits from the 7 point, because he sacrifices that nasty point, giving black an opportunity to start a good defence, so that he can slot checkers from the stacked 6point, and has time to use the dead checker on the 2point.

Bar/22 13/11

The bold vs safe criteria are very important for such a position: Black has a racing lead, a weaker board, and no anchor, all of which argue for a safe play. Additionally, Black's dilly builder and overstacked six point makes the game very difficult if it becomes more complex. Hitting leaves a lot of shots (or a totally ******ed position after 8/6) and risks Black's only advantage (his racing lead). I will play 23 and 5 and hope to roll 55 soon :-)

Using the Safe vs Bold criteria here seems to lead us to Bar/22* 8/6. But it seems soo ugly... I would play Bar/22* 13/11 OTB to try and get something going for me besides this contactful racing lead (give that premature 2 point some utility).

Quiz factor tells me maybe the safe play is right, after all, with all those blots hanging around and lame board we could end up on the wrong side of a complex hitting game.

Only three legal 2's to play after the mandatory hit...over the board I'd probably play b/22* 13/11. The only problem is white has as many or more return shots with 13/11 than 6/4 but some of the hits aren't that pretty after 13/11, with 6/4 white has 13 return shots that are all easy to play. I would almost feel like a heretic playing 8/6 for a 7th checker on the six point but maybe...?

My grunch answer was b/22* 13/11, I'll stick with that...it can only be the 3rd worst play

b/22*, 6/4 because the cube is in the center. b/22*, 8/6 seems to leave you with nothing on your next turn, so if you gotta do something risky to keep a playable position, may as well go all the way, instead of half assing it with 13/11.

big white 6 is a clue: bar/22*, 6/4

Your front position is so stupid already that I'm not sure you can do it much more harm short of playing 55 to the 1-point. It's hard to see the game playing out very often where you win by containing a checker no matter what you do. Since he isn't that close to a prime himself, your lead is enough that you basically have two chances to come around the board and win the race. If he never hits, you virtually always win. If he hits once, you still have a race edge on average if you escape before crunching. All I want to do here is to try to get around to cash in my lead, and any risk of getting a second man sent back works against that plan.

b/22* 8/6

Black has a small racing lead. The pip count is 127 to 134. After this play, he will be up by 12 pips. Add 22 to that when he hits.

There is no rule saying he has to hit. If White would promise not to attack, Black might try the coward’s play, c = cross [bar/23 8/5]. Perhaps he could improve his board next turn, and be ready for a confrontation later. Of course, White is giving no such guarantees, and with 10 checkers in the zone, a fine anchor, and a target in Black’s home board, he won’t need much incentive to launch a blitz.

So Black should hit. He needs the tempo, and the racing pips are nice to have, too.

That leaves him three legal moves for the 2.

• C = Cross [bar/22* 8/6]
• E = Each [bar/22* 6/4]
• S = Split [bar/22* 13/11]

All my instincts rebel against C. It is way to early to worry about complete safety. Black needs to improve his position while White is distracted. Complete safety, candlesticks and all, might be a possibility after Black completes the escape of his back checker, but not now.

E has the big advantage of developing the overstacked 6pt. It leaves 15 ways for White to hit with the checker on the bar (any 4, 53 and 31), plus 2 more shots when White breaks his anchor to hit with 51. A roll of 11 also hits from the anchor, but does not force White to break it. 41 and 44 are double hits. That’s 15 single hits and 3 more double hits. Only 2 rolls force White off his anchor.

Compare this with S. It leaves only outside blots, so there are no double hits. The checker on the bar hits with 44, 53 and 65. By breaking his anchor, White can also hit outside with any entering roll that has a 1 or 4: 11 14 31 34 51 54. That makes 16 shots total, 10 of which force White off his anchor.

You don’t have to do any more counting than that to know that S will give many more return shots in the event that White can hit.

My solution: S (Split) = bar/22*, 13/11

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%.

Excellent point. I had not noticed this.

Could someone explain to me the "big white 6" part? I don't get it.

The hit is clear. It gains a race advantage and send a man back for white, who has nearly solved the opening problem of the back men. The remaining 2 is up for questions.

Marty Storer wrote: “Chapter 16 of Backgammon,“Safe Play vs. Bold Play,” remains one of the finest achievements of backgammon theory.”

Let us look through the principles.

A. Tactical Principles
1. Do you have an advanced anchor? Having an advanced anchor
enables you to play boldly. NO
2. How strong is your opponent’s inner board? The stronger your
opponent’s inner board, the more conservatively you must play. NO
3. How strong is your inner board (especially compared to your
opponent’s?) If you have more inner board points closed than
your opponent, you tend to play boldly; with less points closed,
more conservatively. NO
4. Does your opponent have blots in his inner board? If your
opponent has blots in his inner board, you can afford to take
more chances because of possible return shots. NO
B. Strategic Principles
1. How many men do you have back? The more men you have
back, the more chances you can take. With no men back or
only one man back, you must play conservatively. NO
2. How many men does your opponent have back? When your
opponent has no men back or only one man back, you want
to play provocatively to try to force an exchange of hits. NO



Having the fivepoint, I would call him stronger. Having the twopoint with a dilly builder I would call us weaker. After our hit, there is no more blot to shot on. We have only one man back to scramble home. And after the hit, there a three men who have to come home for white.
According to the criteria, I get a sound “no” for playing bold. And i see only two game plans: scrambling home, winning the race or blitzing out of nowhere with some lucky sets. The man on the eight is a liability, not an asset. And adding liabilities must be even worse.

BAR/23* 8/6.

Epilog. In chapter 16, Magriel wrote “anchor”, in the summery at the end of the chapter, it transformed to “advanced anchor”. Any ideas? I have settled for “anchor” and will not look for it’s quality in the evaluation process.

when you shift the 2 checkers of the 2 point to the 1point, and 1 to the 6 point, making 7 checkers, 13/11 gets very close. I am sure that changing a little bit about reds position, and 13/11 will be best, with above criteria fulfilled. So those criteria are not bullet proof.
And ofcourse, 13/11 is better when a gammon is desired.

To add to my previous post, there are several variations in which 8/6 and 13/11 are about equal ( and more realistic), with the benefit of 13/11 provoking a more complex game.

Post diagrams please from your variants. Then we can better see in which direction you are bracketing.

Btw. i can't imagine a position only slight different, where 8/6 isn't the clear winner.

i have no snipping tool, no paint program, the .sgf is not recognized by http://imageshack.us/, so at the moment i do not know how to put a board image with linux.
if you occupy the 1point instead of 2 point there are several positions which make the two moves about identical in equity.

1. In GnuBg, press Ctrl+C to copy the ID. In XG, press Ctrl+Shift+C.
2. Go the MindGamesCenter.com Diagram Page.
3. Paste the ID, and click Create Diagram.
4. Copy the results into a message here at TwoPlusTwo.

Hope that helps.

^ That position is very different from the original problem. The double hit here seems a must.

I don't see the link between those last 2 positions and the original one. The White blot on the 11-pt change things greatly. In fact, it is such a major change that I wouldn't consider those positions "variations". Visually, the position might look similar, but conceptually, it's now a totally different game.

But he got the big white 6 right.

I still don't know what that means.

you are right. i trusted my memory, suppose i was tired. the difference is negligible: