Problem of the Week #113: July 22


Cash game. Center cube. Black on roll.




Should Black double? Should White take if doubled?


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.

I think black should double - he is about to build a 3-prime in white's home if he rolls a 4 or a 2, and has little pressure on his back men. Plus, plenty of builders on the 13 might help him extend the prime.

White should accept. His back men are safe and at first glance, it doesn't look like he's anywhere near a 75-80% dog

Answer: Double, accept

Just seems like a routine double/take. Black's a little ahead in the race and structure, he might pick up another blot if he rolls well, and gammons are clearly asymmetric, but white can't possibly be dropping here with a made anchor, a sane position, and black with only a presumptive split 4-prime and still having to get 2 men around. Maybe there's some reason you wait one turn in case you roll useless and get hit, but you're probably losing your market after 11 or 44 (maybe even 33/66) and a bad return, so I just do it now.

Grunch.

Seems a pretty clear double. Threats, assests, man in the air, nice proto-prime.

Pretty simple take imo. Couple of points made, anchor on the ace. Lots of possibilities left.

Double/Take

The problems are always very difficult, so there is no way the solution is a simple double/take.

How about something more complicated like this:
Black should roll. If he makes his 4 point and white fails to enter black should double next turn and white should take.

With only two points made and a blot in a home board, black's forward position isn't too strong.
Also white has her 4pt made. Not as strong as 5pt, but still strong.

White's blot on 15pt isn't in a big threat as black will use 63 and 54 to cover the 4pt blot.


I can't imagine anyone pass this as white, so I would wait with doubling to give her harder take/pass decision later.



No double / easy take

I think you could spin or shake here, doesn't seem like it would be a huge error either way. I would probably spin OTB. Take looks easy.

D/T

Enough threats- hit, miss... make 4pt,dance- to double now.

Double/take

Looks like a standard double/take for me.

I would not double here as the edge is not significant and the take is easy. Re-eval, double if position improves more in next 2 rolls.

No double/take. The take is super easy. One roll from now, if he cubes, black is more likely to regret that he doubled than he is to lose his market if he doesn't double.

When White fans in this position, Black should think, “Double!” And, a moment later, “Am I good enough to double?” The famous “PRaT” rule seems to indicate, “yes!”

• Position
  The structure is about even. Both players have made a second inside point. Black has a third point slotted. The biggest difference is that White has three checkers back, and one of them is on the bar. Edge to Black.
  
  
• Race
  Black leads, 153 to 170. Edge to Black.
  
  
• Threats
  Black has two threats. Eights and nines allow him put a second White checker on the roof, while twos, fours and nines will let him cover on the four point. Black’s blitzing opportunities are limited, however, because White is anchored on his one point. Edge to Black.

Conventional wisdom says that when you have an advantage in two out three of these categories, you should double. When your opponent leads in all of them, you should pass.

Let’s take a look at Black’s upcoming roll:

11: make 4 point and bar point
22: make 4 point, slot 11 and 20 points
33: make 4 point, slot 20 point
44: make 4 point and 2 point, slot 20 point
55: make 3 point, slot 18 point, play down to 8 point
66: make 2 point and bar point
12: make 4 point, slot 22 point
13: make 4 point
14: make 4 point, slot 22 point
15: make 3 point (saving the blot on the 4 point)
16: make bar point
23: make 4 point, slot 20 point
24: make 4 point, slot 20 point
25: make 4 point, slot 18 point
26: make 4 point, slot 18 point
34: make 4 point, slot 20 point
35: hit on 15 point
36: make 4 point
45: make 4 point
46: make 4 point, slot 18 point
56: make 18 point

Wow! Black does not have any poor rolls! There are 26 rolls that make the 4 point. Four of the remaining rolls miss making the 4 point, but allow Black to make another inside point. That’s 30 rolls in all that leave Black with (at least) a three-point board. With two others (that is, when Black rolls 61), he can make his bar point, a powerful block against White’s anchor. So what are the four rolls that won’t make a point? Well, Black’s worst is 56, and it gets him an advanced anchor on the 18 point. And then there is 53. All Black can do with it is hit White’s outside blot!

The fact that Black’s back men are split adds real power to this position. Fully half of Black’s rolls allow him to make an inside point, and simultaneously to bring a checker within range of a direct shot on White’s outside blot. This means that White will be scrambling to button up, or else will have to face having that checker sent back.

If we assume that Black makes his 4 point on his turn, then White will have 19 rolls (11, 22, 33, 12, 13, 14, 16, 24, 26, 34, 36) that let him both enter and pick up his blot. (When he rolls 23, White will probably be forced to advance his anchor, and leave the blot.) Of course, White will be dancing on 9 other rolls. For Black, this means that his next roll may be as much fun as this one. He will have a direct shot at a White blot about half the time.

“PRaT" notwithstanding, I find it difficult to pass in this position. White’s anchor will give him chances right up to the end. In addition, Black is a little bit short on attacking material. His 8 point is stripped, and may have to be sacrificed on this turn. White should have a decent chance to regain some balance.

My solution: Double/take

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 48%.

Grunch.

Working with "What is Your GAME PLAN?", i almost forgot to post here. Btw, anyone a suggestion for an average equity to get working through this problems? Woolsey/Heinrich suggested .03 on average for the "New Ideas in Backgammon" Problems.

Here my evaluation:

Black has 2 men back, White 3 men, with one on the roof. White has 2 points in his board, Black 2,5. Both have 9 men in the zone. It isn’t clear, if the white builder on the 10 point is an asset or a liability. The race is 153:170, a solid lead of 17 pips for black.

Position: Black has the better point in the house. The 4 point from White is a potential defekt. If Black can anchor on the golden point, the 4 one is then out of play for blocking purposes. The third man behind from White is a violation of efficiency, it is difficult to get this man to work again.
Race: If the race would be the only issue, then Black would have a very clear double and White only a take, but it would be no bargain.
Threats: The main threat from black is just to make the 4 point. White can then only form an anchor on the 22 point, and, citating a leading person in backgammon, “the 22 point is a piece of ****”.

In spite of the fact that every criteria points out for a double, there is not a faintest doubt, that this is a take. Making the 4 point isn’t the end of the world. Only if Black makes the 4 and White then dances, White would regret the take. Black has a position, which is almost getting better, without backswings in the next move. White’s position is sound enough with no major defekts to justify a take.

Good initial double, and a not an easy, but clearcut take.

Good post IMO. I'm impressed by people who take the time to write such lengthy explanations of their thought process.

Giving this position another look, I'm now guessing that it is a double/take though I'm far from sure. The position looks really simple at a glance but there's a lot of subtle factors here. The fact that white has 3 checkers back is a big factor. Also, while there may be few market losing sequences there are some and at worst, one roll from now black will be equal.

Seemed simple at first, but as usual, things are far from being that simple.

Could this be none other than Stick Rice?

You bet!

Source: http://www.bgonline.org/forums/webbb...mes;read=98881

The 22-point is a transitional point between the high anchors (20-point, 21-point) and the low anchors (24-point, 23-point). It has its own value and its own set of strategies and take points.

As you play backgammon, you'll run into lots of folks who say silly things in an attempt to be colorful or edgy.

There are no bad points, only bad players who can't handle them well.

Ya. Stick is a real fish.

As for the problem at hand, looks like a solid double. Ahead in the race, and structure with a checker on the roof. The take I am less certain about, but I am tempted to take it. We are still unsplit off the 24, but our front structure looks good. We have the four point, a spare on the midpoint and a builder in place for the five point. I think this will turn out to be close. Either a small take or a small drop.

Double/Close take

it's way closer to a double/no double decision then a take imo

over the board I'll take this so fast I spill my beverage

All 2's 4's and 9's cover the blot, thats 26 covers, above that theres 53 which hits.
After the most plausible cover, white will probably enter.
The defender has an anchor so lets use the Robertie rule for this most plausible situation.

contact + 5-prime = pass

I enlarged the Robertie-rule with my own experience:

little contact + 4-prime = pass

(I speak of little contact in situations where both black back checkers would be for example on the 15-point,
or when white would not have made an extra point, here the 4-point)

little contact + 3-prime + extra checker back = pass

So it only seems logical to conclude:

contact + 4-prime + extra checker back = pass

So double pass.

It seems I was wrong:

But I still think this is an easy take and you guys are overestimating black's chances with only 9 checkers in the zone, blot in a home board (opposed to made point) and made 4pt for the defender.

If I knew my opponent overestimates these positions, I would double. Otherwise, I would wait as I already said.

I agree, though I wouldn't spill my beverage. Beer is a great gift from the gods and should be respected and honored at all times.

I think you are overestimating Woolseys Law. Even if there is no doubt, that a position is a take, it can be nevertheless a double. First. It isn't a double. But doubling the stakes in a position where you are better, how bad can this be? Second. You have to look at the volatility of the position. Mostly Black gets better, with no backfire. Third. Your opponent has to make a cube decision. Since there is also a vote for pass, you see, it is possible, that your opponent can blunder. At least, words are qualifiying. A quantitative judgement you will get through statistics.

[Ya. Stick is a real fish.]

Which PR qualifies in your judgement for the distinction "real fish"?

I agree, IF there are many market losers.



While this could be theoretical double, look what Alpha Fish said - most of my opponents would take this without thinking, so by doubling now - I'm giving my opponent very easy decision.
That's why I would wait with doubling - to give my opponent harder decision next time, and bigger possibility to make a mistake.


----

By the way, shouldn't the solution have already been posted?

I was joking. Just making light if the fact that bill said, ther are no bad points just bad players. It was, in fact, stick that called the 22 a piece of....