the big thread about cube decisions - Page 9 - Backgammon Forum - Discuss Backgammon Strategy, Clubs and Books

Thanks for your linguistic support. Seems I have given you enough reason to doubt me at times. It also suggests that for the rest of the post I am doing pretty okay.

By the standards of this monolingual fool, you are doing great.

Mike

Quote

05-26-2015 , 04:15 AM

#202

yogiman

adept

Join Date: Jun 2012 Posts: 784

Position ID: bjwzIEBv2wAgIQ Match ID: UgkAAAAAAAAE

White - Pips 114

Black - Pips 94
Black on roll. Cube action?
Created with www.BGdiagram.com

Quote

05-26-2015 , 05:49 AM

#203

bleep69

grinder

Join Date: Sep 2008 Posts: 483

Wow, I have no grounding here on which to base a decision, meaning this is really just a guess. I'll go with double/take. Black has 55, 66, 64 that hit immediately, with more ammo coming down to the white blot on 8. If white doesn't get out pretty soon, the future looks grim for him. However, should he get out sharpish, he could well put black on the back foot.

Quote

05-26-2015 , 06:59 AM

#204

RolldUpTrips

old hand

Join Date: Oct 2006 Posts: 1,334

At first I was thinking double/take, but the more I think about it the more I'm guessing double/pass. Black is 20 pips ahead and White's front position is kinda messed up.

Quote

05-26-2015 , 01:18 PM

#205

uberkuber

veteran

Join Date: Jul 2009 Posts: 3,032

Quote:

Originally Posted by bleep69

44 also hit immediately.

I will venture a double/slight drop.

Quote

05-29-2015 , 06:54 PM

#206

TheRealFatboy

centurion

Join Date: Sep 2013 Posts: 165

Double looks clear

Take, hmmmm I'm not sure, small drop.

Quote

05-30-2015 , 10:53 AM

#207

yogiman

adept

Join Date: Jun 2012 Posts: 784

No worry, I am just ill. Maybe tomorrow I come with the results.

Quote

06-03-2015 , 08:57 AM

#208

yogiman

adept

Join Date: Jun 2012 Posts: 784

Code:

Cube analysis
Rollout cubeless equity  +0,656

Cubeful equities:
1. Double, pass         +1,000
2. Double, take         +1,117  ( +0,117)
3. No double            +0,780  ( -0,220)
Proper cube action: Redouble, pass

If white would have a 4-point board, black would be very far from a double. Getting a checker in direct range of the white blot next turn, followed by white not entering, will make the situation very much too good to double.

Reducing number of indirect shots by moving w8 to 7:

Code:

Cube analysis
Rollout cubeless equity  +0,633

Cubeful equities:
1. Double, pass         +1,000
2. Double, take         +1,075  ( +0,075)
3. No double            +0,770  ( -0,230)
Proper cube action: Redouble, pass

Black's indirect hitting numbers were 66, 64, 55, 44, whereas now only 65.

Moving black from 18 to 21:

Code:

Cube analysis
Rollout cubeless equity  +0,509

Cubeful equities:
1. Double, take         +0,814
2. Double, pass         +1,000  ( +0,186)
3. No double            +0,663  ( -0,151)
Proper cube action: Redouble, take

Again we see the butterfly-effect operating in the backgammon world of cube decisions. Moving black's checker 3 pips is enough to shift the decision from a firm pass to a firm take. As seen in the previous position, the reduction of indirect hits only bear a little responsibility for this drastic change. Another tiny addition is that 51 and 41 don't get him in direct range of the white blot.
The most important reason is likely illustrated by the following position. Moving 15w to 17 and 19:

Code:

Cube analysis
Rollout cubeless equity  +0,708

Cubeful equities:
1. Double, pass         +1,000
2. Double, take         +1,265  ( +0,265)
3. No double            +0,837  ( -0,163)
Proper cube action: Redouble, pass

In the previous position white had a good opportunity to make the 20 or 21-point after he has entered, and when that happens on black's head it might turn the tables. Black had an extra checker to move out of a potential danger zone. In this position white's offensive is less threathening.

Quote

06-07-2015 , 07:59 AM

#209

yogiman

adept

Join Date: Jun 2012 Posts: 784

Position ID: B7crICDYbWDAQQ Match ID: cAkAAAAAEAAE

White - Pips 119

Black - Pips 160
Black on roll. Cube action?
Created with www.BGdiagram.com

Quote

06-07-2015 , 02:27 PM

#210

Karol Szczerek

enthusiast

Join Date: Mar 2015 Posts: 92

I'd go with O'Hagan's Law here.

I count 18 hitters, which are probably all market losers if opp doesn't roll a good response. Let's say, the good responses are ca. 1/4 of his rolls. This leaves ca. 12-13 market losers. There are no anti-jokers (rolls, after which we would have to pass a double). After no-hitting rolls we are doing OK.

A strong double.

Is it a take?

The dead checkers on the 1st point are the main issue here. After we're hit we have some possiblity to anchor-up. Opp's prime isn't decisive yet, and there are not many attackers in the zone. After we're missed, we certainly have some play left. Our board is just 1 point worse, and we have some blocking power in front of opp's 4 back checkers. That all would be enough to take for me. Probably a thin take, since our outfield presence is a mess. But if I had only 2 dead checkers. The third one is a really big liability, shrinking our flexibility. I've learned from bots that such one checker can be a 0,2 - 0,3 equity difference. For me, this guy pushes it below take point.

I pass.

Quote

06-08-2015 , 09:47 AM

#211

uberkuber

veteran

Join Date: Jul 2009 Posts: 3,032

Yeah, that 3rd checker on the ace point is rather annoying to say the least.

Quote

06-08-2015 , 10:07 AM

#212

bleep69

grinder

Join Date: Sep 2008 Posts: 483

We have many good rolls to hit, and a trap that's developing nicely. White on the other hand has 3 guys piled up on his 1 point already. I think even though we are on the bar we have a solid double, with the take looking too risky.

Quote

06-09-2015 , 03:30 AM

#213

yogiman

adept

Join Date: Jun 2012 Posts: 784

White - Pips 119

Black - Pips 160
Black on roll. Cube action?
Created with www.BGdiagram.com

Code:

Cube analysis
Rollout cubeless equity  +0,655

Cubeful equities:
1. Double, pass         +1,000
2. Double, take         +1,122  ( +0,122)
3. No double            +0,862  ( -0,138)
Proper cube action: Double, pass

24w to 19:

White - Pips 124

Black - Pips 160
Black on roll. Cube action?
Created with www.BGdiagram.com

Code:

Cube analysis
Rollout cubeless equity  +0,553

Cubeful equities:
1. Double, take         +0,890
2. Double, pass         +1,000  ( +0,110)
3. No double            +0,740  ( -0,151)
Proper cube action: Double, take

As so often, if white is a good number of pips ahead he has a take. And yes, a killed checker is often a killer.

Quote

06-10-2015 , 11:00 AM

#214

yogiman

adept

Join Date: Jun 2012 Posts: 784

Position ID: NzYBgC3YbUFKQA Match ID: cAkAAAAAEAAE

White - Pips 151

Black - Pips 144
Black on roll. Cube action?
Created with www.BGdiagram.com

Quote

06-11-2015 , 02:56 AM

#215

Kamba

enthusiast

Join Date: Feb 2015 Posts: 93

I will risk great embarrassment and venture an answer here; I tend to learn from the ones I get badly wrong, so here goes.

I am pretty sure White has a take. I am thinking that there is enough chance of white hitting before his home board collapses to give him a good enough chance. If white hits two checkers he's looking even better.

I am totally clueless about the double. It looks great for black on the surface, but i have lost a lot of games from positions like these. OTB its 50/50 for me whether i double these or not.

So:

Surely a take.

Probably a double.

Quote

06-11-2015 , 09:25 AM

#216

bleep69

grinder

Join Date: Sep 2008 Posts: 483

Yogi, how do you find these positions?!

I am at sea here. The only reference point I can bring to bear on this is that XG doesn't double when a good back game seems to be brewing. However, as I typed this, I partially remembered another nugget, that for a back game, we have to be a certain distance behind in pip count, and for a 23 back game, while I don't remember the number, I know it must be high, possibly as high as 100, and we are practically level here, so my one and only reference point has been rendered useless before I begin.

Double/take.

Quote

06-12-2015 , 07:27 AM

#217

yogiman

adept

Join Date: Jun 2012 Posts: 784

Code:

Cube analysis
Rollout cubeless equity  +0,541

Cubeful equities:
1. Double, take         +0,843
2. Double, pass         +1,000  ( +0,157)
3. No double            +0,598  ( -0,246)
Proper cube action: Double, take

White has a take, because there is 45% chance that black will not enter, and 50% chance that white hits a second checker with a 1, 5 or 6. Because of the string of black blots these will almost be certain wins. That's almost the required 25%, and there will be also some gammons among them. So in answer to Bleep69, a backgame is not on the mind of black.

Removing two blots will lead to a pass. 15 to 16:

Code:

Cube analysis
Rollout cubeless equity  +0,750

Cubeful equities:
1. Double, pass         +1,000
2. Double, take         +1,319  ( +0,319)
3. No double            +0,857  ( -0,143)
Proper cube action: Double, pass

Giving life to the dead checker gives a much lower equity for black, but doubling is not bad. (well done, Kamba). 24w to 19:

Code:

Cube analysis
Rollout cubeless equity  +0,234

Cubeful equities:
1. No double            +0,257
2. Double, pass         +1,000  ( +0,743)
3. Double, take         +0,227  ( -0,030)
Proper cube action: No double, take

Putting the dead checker behind black's block doesn't increase white's flexibility very much for the moment, so it is still a double. 24w to 3:

Code:

Cube analysis
Rollout cubeless equity  +0,523

Cubeful equities:
1. Double, take         +0,796
2. Double, pass         +1,000  ( +0,204)
3. No double            +0,589  ( -0,207)
Proper cube action: Double, take

Quote

06-12-2015 , 08:32 AM

#218

bleep69

grinder

Join Date: Sep 2008 Posts: 483

Please call me bleep yogi, all my friends do.

Quote

06-12-2015 , 11:34 AM

#219

yogiman

adept

Join Date: Jun 2012 Posts: 784

Okay, bleep. This reply gives me an excuse for a correction. A backgame is not on the mind of white, instead of black. And black does have some market losers, if he enters with no 36 or 46, because he will hit, cover the 8-point blot or safety the 8-point blot.

Quote

06-14-2015 , 04:50 AM

#220

yogiman

adept

Join Date: Jun 2012 Posts: 784

Position ID: cOfBAEzZbsAGEA Match ID: cAkAAAAACAAE

White - Pips 162

Black - Pips 119
Black on roll. Cube action?
Created with www.BGdiagram.com

Quote

06-15-2015 , 02:14 AM

#221

bleep69

grinder

Join Date: Sep 2008 Posts: 483

I'll go with double/take. Black is ahead in the race with a good board, but white could develop his very quickly. There is bound to be a shot or two coming his way.

Quote

06-15-2015 , 09:52 AM

#222

Alpha Fish

Pooh-Bah

Join Date: Aug 2008 Posts: 3,832

this looks scary as hell, I drop pretty quickly

Quote

06-16-2015 , 02:47 AM

#223

Kamba

enthusiast

Join Date: Feb 2015 Posts: 93

I think I would drop this one.

I might take as white with another one or two checkers on the mid point. White's position here just looks a little too awkward for my liking.

Quote

06-17-2015 , 07:17 AM

#224

yogiman

adept

Join Date: Jun 2012 Posts: 784

Code:

Cube analysis
Rollout cubeless equity  +0,667

Cubeful equities:
1. Double, pass         +1,000
2. Double, take         +1,133  ( +0,133)
3. No double            +0,916  ( -0,084)
Proper cube action: Double, pass

Ideally we use as little memory as possible, and amplify the reference positions we know with our deductive reasoning. From Kit Woolsey's book we can infer the directive that
if white has a two-point holding game and black has no rear checkers left but the 13-point, it is virtually always a pass for white

The following two statements lead to the right cube analysis:
A)If black gets the rear checker into white's outfield while white is on the bar the above directive is practically fulfilled.
B)If black covers the 1-point, he will very likely be able to move the rear checker ahead, which also means that above directive is practically fulfilled.

ad A) black gets the rear checker into the outfield with 16, 26, 52, 53, 55, which gives 9 possibilities (=25%). If he does get his rear checker out white's only hope is to throw a 1:
white doesn't hit with 25:

Code:

Cube analysis
Rollout cubeless equity  +0,617

Cubeful equities:
1. Double, pass         +1,000
2. Double, take         +1,060  ( +0,060)
3. No double            +0,946  ( -0,054)
Proper cube action: Double, pass

white hits with 15:

Code:

Cube analysis
Rollout cubeless equity  +0,512

Cubeful equities:
1. Double, take         +0,772
2. Double, pass         +1,000  ( +0,228)
3. No double            +0,659  ( -0,113)
Proper cube action: Double, take

ad B) black covers with 4's, 31, 22, 33 or 66, which gives 16 possibilities (=45%)

The chance that white doesn't throw a 1 is 70%, multiplied by the probability that black gets his rear checker out: 70% times 25% = 17% (A). The chance that black covers immediately is 45% (B). If we do A + B then at least 17 plus 45 is about 60% of black's numbers leads to a pass for white. Of the remaining 40% certainly no half of it that will lead to a win for white, so I think the pass is very clear.

In response to Kamba, as we can directly infer from the directive, putting two extra checkers on the midpoint turns out to be of no significance:

[/URL]

Code:

Cube analysis
Rollout cubeless equity  +0,669

Cubeful equities:
1. Double, pass         +1,000
2. Double, take         +1,114  ( +0,114)
3. No double            +0,895  ( -0,105)
Proper cube action: Double, pass

Putting the builder out of direct range leads also to a pass:

Code:

Cube analysis
Rollout cubeless equity  +0,632

Cubeful equities:
1. Double, pass         +1,000
2. Double, take         +1,080  ( +0,080)
3. No double            +0,911  ( -0,089)
Proper cube action: Double, pass

Black has only 6 shots that will cover, which is 16%. However, the chance that white will not enter is 45%, after which he will be in direct range (=about 60%). So the chance that black will cover is 16 + 45% times 60% = 16 + 27 = 43%. A + B = 17 + 43, so will give about the same number of passes as in the original position. The difference is that white has somewhat more opportunity to hit the 1-point blot.

3-point board for white:

Code:

Cube analysis
Rollout cubeless equity  +0,463

Cubeful equities:
1. No double            +0,689
2. Double, pass         +1,000  ( +0,311)
3. Double, take         +0,679  ( -0,011)
Proper cube action: No double, take

This might be doubled, but it is a huge take. This is rather difficult to infer. If white hits, he will become a serious contender. The chance that white hits is the chance that black doesn't cover multiplied by the chance that white hits: (100 – 45)% times 30% = 18%. Black might fail to cover a second time: 55% times 55% times 30% = 10%. Let's make an estimate (and please correct me) that with a 3-board half of those 30% hits will lead to victory, than he has got 15% wins. Of the remaining 70% that white enters on the 2-point white should at least be able to win 10%, which adds up to the necessary 25% to take. Ofcourse in reality it will be more, but at least this calculation leads to the right decision.