Doubles: Bill & Steve versus Dmitriy and Prince

Below are selected error from Novi doubles match. Steve and I against Dmitriy and Prince.

************


Opening game of 7 point match. 11 to play.

is Dmitriy and Prince score: 0 pip: 159

7 point match

pip: 159 score: 0 is Bill and Steve

XGID=-b-B--D-B---eE---b-bbb--B-:0:0:1:11:0:0:0:7:10

to play 11

1.	XG Roller++	24/22 6/5(2)	eq: -0.131
	Player: Opponent: 46.44% (G:14.12% B:0.74%) 53.56% (G:15.69% B:0.71%)
2.	XG Roller++	8/7(2) 6/5(2)	eq: -0.189 (-0.058)
	Player: Opponent: 45.46% (G:14.18% B:0.79%) 54.54% (G:17.77% B:1.03%)
3.	4-ply	24/23(2) 6/5(2)	eq: -0.250 (-0.119)
	Player: Opponent: 42.94% (G:13.13% B:0.59%) 57.06% (G:16.11% B:0.77%)

eXtreme Gammon Version: 2.02, MET: Kazaross XG2

We didn't really give this play any thought.

8/7(2), 6/5(2) was out play, blocking sixes and forming a broken four prime.

Better was 24/22, 6/5(2). I am guessing reasons are:

Bar is not a new point. Opp's 6-5-4 structure is very strong and we should try to split now while they lack checkers in the zone.

*************

We trail 7 away 6 away. We have already cubed and have 65 to play.

is Dmitriy and Prince score: 1 pip: 116

7 point match

pip: 125 score: 0 is Bill and Steve

XGID=-bBB-BC---C--A---cbb-dB-b-:1:-1:1:65:0:1:0:7:10

to play 65

1.	XG Roller++	10/5 10/4	eq: +0.376
	Player: Opponent: 65.93% (G:22.05% B:0.75%) 34.07% (G:10.12% B:0.34%)
2.	XG Roller++	13/2	eq: +0.315 (-0.062)
	Player: Opponent: 63.21% (G:20.17% B:0.65%) 36.79% (G:6.77% B:0.19%)
3.	4-ply	13/8 10/4	eq: +0.318 (-0.059)
	Player: Opponent: 64.14% (G:19.61% B:0.69%) 35.86% (G:8.53% B:0.24%)

eXtreme Gammon Version: 2.02, MET: Kazaross XG2

We played super wimpy 13/2.

Better was to slot 13/8, 10/4. The twin tactical themes here appear to be:

Keep you checkers in play; don't kill them.

Opponent's position is realatively weak, so the cost of getting hit is less than it might otherwise be.

Strategically, it looks like we are trying to outprime him, so slotting makes more sense in that context also.


************

Later in the same game. We have a tricky 32 to play.

is Dmitriy and Prince score: 1 pip: 110

7 point match

pip: 110 score: 0 is Bill and Steve

XGID=aABB-BCAa-B------cbb-dB-b-:1:-1:1:32:0:1:0:7:10

to play 32

1.	XG Roller++	10/8* 8/5	eq: +0.785
	Player: Opponent: 74.48% (G:38.99% B:0.64%) 25.52% (G:6.20% B:0.24%)
2.	XG Roller++	10/7 10/8*	eq: +0.747 (-0.038)
	Player: Opponent: 74.09% (G:35.74% B:0.64%) 25.91% (G:5.45% B:0.19%)
3.	XG Roller++	7/4 6/4	eq: +0.663 (-0.121)
	Player: Opponent: 70.91% (G:35.23% B:0.64%) 29.09% (G:6.13% B:0.22%)
4.	XG Roller++	6/1	eq: +0.639 (-0.146)
	Player: Opponent: 69.26% (G:36.54% B:0.46%) 30.74% (G:5.63% B:0.20%)

eXtreme Gammon Version: 2.02, MET: Kazaross XG2

At least four reasonable options here.

We went with 7/4, 6/4 making the board. It worked out for is, as we eventually close dout two checkers and rolled hime to a gammon. But any ace is really bad, and even if he misses we are nine long pips away from covering next time.

Worst was 6/1. This makes a 5 point board without a blot. But the board with hole on 4 point is much weaker than hole on ace point. If Dima&Prince enter from the bar timmediately we are in big trouble and if they enter within a few rolls hey are right back in the game.

The top play per the computer was 10/8*, 8/5. This hits a second blot and diversifies our men best for filling in the ace, four points.

Second best was 10/8*, 10/7 which also hits a second man but is less efficient in the aim to closing the board.


***************

We lead 3 away 6 away and have 36 to play from the air.

is Dmitriy and Prince score: 1 pip: 165

7 point match

pip: 151 score: 4 is Bill and Steve

XGID=-aaB--DaC--BbB-b-c-e----AA:0:0:1:63:4:1:0:7:10

to play 63

1.	XG Roller++	Bar/16	eq: +0.099
	Player: Opponent: 55.61% (G:18.64% B:1.00%) 44.39% (G:11.78% B:0.50%)
2.	XG Roller++	Bar/22 24/18	eq: +0.078 (-0.022)
	Player: Opponent: 55.28% (G:18.71% B:1.20%) 44.72% (G:12.31% B:0.51%)
3.	XG Roller++	Bar/22 13/7*	eq: +0.031 (-0.068)
	Player: Opponent: 53.64% (G:19.95% B:1.33%) 46.36% (G:12.36% B:0.69%)

eXtreme Gammon Version: 2.02, MET: Kazaross XG2

We entered and hit with Bar/22, 13/7*. While we have the better board, this play seems too unlikely to work. We get back with most 1's, 5's and 6's. Probably we did it because we didn't want to give opponents a free chance to anchor on the bar point.

But stategically, we are up in the race so moving the back men makes sense. Either Bar/16 or Bar/22, 24/18 are better plays. They can't do everything in one roll and with the better board, an exchange of hits may favor us.

Another small factor may be that hitting gives up the midpoint and strands our back men somewhat.

**************

We are on roll holding a 2 cube and leading 3 away 6 away. Cube action?

is Dmitriy and Prince score: 1 pip: 3

7 point match

pip: 5 score: 4 is Bill and Steve

XGID=-AB---------------------c-:1:1:1:00:4:1:0:7:10

on roll, cube action?

Analyzed in XG Roller++	No redouble	Redouble/Take
Player Winning Chances:	85.65% (G:0.00% B:0.00%)	85.65% (G:0.00% B:0.00%)
Opponent Winning Chances:	14.35% (G:0.00% B:0.00%)	14.35% (G:0.00% B:0.00%)
Cubeless Equities	+0.713	+1.169
Cubeful Equities
No redouble:	+0.713 (-0.092)
Redouble/Take:	+0.805
Redouble/Pass:	+1.000 (+0.195)
Best Cube action: Redouble / Take

eXtreme Gammon Version: 2.02, MET: Kazaross XG2

We rolled. It turns out this is a large double. I had a similar position a couple weeks ago where I was up 3 away 7 away holding a 2 cube in a 3 roll position. That time I redoubled to put the match on the line and it was a huge error. So, I was gun shy and overcompensated this time, which turns out to be a proper cube. Note also I goofed up the big cube against Ray (prior post) so I need some work on these.

Let's look at the math.

Min Double Point = Risk / (risk + gain)

= [ (No Double & Lose) - (Double & Lose) ] / [(Double & WIn) - ( No Double & Win) + RISK]

= [ (ME 3a4a) - (lose match)] / [(Win match) - (ME 6a Crawford)+ RISK]

= ( 57 - 0 )/ (100 - 89) + 57

= 57 / 68

= 84%

A two roll position is about 86% to win. Though here double ones don't get off, so that should cost a bit, maybe a percent. So let's say we are 85% to win. Therefore, I guess it is a small double.


However, I am absolutely hopeless when it comes to this stuff. It took me twnety minutes with a calculator in hand to come up with this 84% number. I have zero chance of doing this at the table.



Let's try another approach. What is opponent's takepoint? Dmitriy and Prince can drop and play from 6 away Crawford - about 11%.

Or they can take and rewhip for the match, when they are a tiny bit better than a two-roll position. I woudl hvae estimated 15%.

So we know the position is an easy take. But is it a double? This is the tough part. This is like a "last roll" as far as the cube goes, since next turn the game will either be over or they are dropping. So we lose the market on the next exchange if things go well.

The key seems to be how close is opponent to the drop point? ie, how close are the winning chances (15% in my head, actually 14%) to the drop point (11%)?

The closer to the drop point, the more efficient the cube, and the further away the less efficient. Realistically, this is the best I can hope to achieve over the board. Figure out opponent's drop point, estimate win chances, and note the gap betweeen the two and that make a judgment call. Given all this data, I still would have held the cube as the four percent gap from 15 to 11 seems large enough that you are risking too much for not enough gain. But I guess it is a double after all.

sixteen cube from Novi

This sequence is from my consolation match against Ray.  I can't say we got to 16 legitimately, as I made some poor cube decisions on the way to 16.  But it is entertaining.

First game of a match to 9.  Ray sent the initital cube, which was a very takeable backgame.

He left a shot which I hit.  Now I am on roll.  Cube action?

is Bill score: 0 pip: 151

9 point match

pip: 91 score: 0 is Ray

XGID=-bCCaCC--Ba-a-a---abbbb--A:1:-1:-1:00:0:0:0:9:10

on roll, cube action?

Analyzed in Rollout	No redouble	Redouble/Take
Player Winning Chances:	71.06% (G:13.37% B:0.58%)	72.12% (G:14.47% B:0.73%)
Opponent Winning Chances:	28.94% (G:11.74% B:1.75%)	27.88% (G:11.95% B:2.73%)
Cubeless Equities	+0.426	+0.959
Cubeful Equities
No redouble:	+0.796 (-0.058)	±0.010 (+0.786..+0.805)
Redouble/Take:	+0.854	±0.014 (+0.839..+0.868)
Redouble/Pass:	+1.000 (+0.146)
Best Cube action: Redouble / Take
Rollout details
1296 Games rolled with Variance Reduction. Moves: 3-ply, cube decisions: XG Roller
Double Decision confidence:		100.0%
Take Decision confidence:		100.0%
Duration: 7 minutes 51 seconds

eXtreme Gammon Version: 2.02, MET: Kazaross XG2

I correctly redoubled and Ray took. I rolled an ugly double 5's and Ray hit me with 16 from the bar. That's right, 16, not 26. A couple rolls later and I am starting down this 8 cube. Take or Drop?

is Ray score: 0 pip: 77

Unlimited Game Jacoby Beaver

pip: 127 score: 0 is Bill

XGID=--ABBCC-------a-b-BccBcc--:2:-1:-1:00:0:0:3:0:10

on roll, cube action?

Analyzed in Rollout	No redouble	Redouble/Take
Player Winning Chances:	76.67% (G:11.53% B:0.09%)	76.69% (G:11.32% B:0.08%)
Opponent Winning Chances:	23.33% (G:1.66% B:0.04%)	23.31% (G:1.78% B:0.05%)
Cubeless Equities	+0.633	+1.259
Cubeful Equities
No redouble:	+0.923 (-0.077)	±0.003 (+0.920..+0.926)
Redouble/Take:	+1.087 (+0.087)	±0.012 (+1.075..+1.098)
Redouble/Pass:	+1.000
Best Cube action: Redouble / Pass
Rollout details
1296 Games rolled with Variance Reduction. Moves: 3-ply, cube decisions: XG Roller
Double Decision confidence:		100.0%
Take Decision confidence:		100.0%
Duration: 4 minutes 00 second

eXtreme Gammon Version: 2.02

I took though the computer shows it to be a significantly sized pass. I confess I did not attempt to calculate a takepoint here. Even if I manage to get the percentages from the ME table correctly I can't divide the numbers in my head. I always get it wrong anway, so I don't even bother to try. Lastly, Ray left a shot during the bearoff. He has one checker off and I have a last gasp shot. Cube action?

is Ray score: 0 pip: 32

9 point match

pip: 80 score: 0 is Bill

XGID=-BCBBBB--------A----aAfbe-:3:1:1:00:0:0:0:9:10

on roll, cube action?

Analyzed in XG Roller++	No redouble	Redouble/Take
Player Winning Chances:	29.41% (G:0.00% B:0.00%)	29.41% (G:0.00% B:0.00%)
Opponent Winning Chances:	70.59% (G:0.26% B:0.00%)	70.59% (G:1.42% B:0.00%)
Cubeless Equities	-0.412	-0.464
Cubeful Equities
No redouble:	-0.378
Redouble/Take:	-0.464 (-0.086)
Redouble/Pass:	+1.000 (+1.378)
Best Cube action: No redouble / Take
Percentage of wrong pass needed to make the double decision right: 5.6%

eXtreme Gammon Version: 2.02, MET: Kazaross XG2

Another poor cube action on my part. I doubled and of course Ray took. I got lucky and rolled the ace. Ray entered as soon as I opened the board and the race was close. It went down to the last shake and I managed to win. On the cube, I had no idea what the minimum doubling point was. I fugured I have little to lose and I win almost 30% so why not? While true, I failed to consider that I have little to gain. If I hit the shot and win I am up 8-0 crawford which is pretty darn good so there really is little point to putting the match on the line. Dmitriy advised me that the minimum double point at any match score (and any cube level) where the players are tied is always 50%. I am going to call that "Dmitriy's rule" and try to remember that for the future. In any case, I don't understand why if I am 30% to win and the minimum double point is 50% ... why is my cube error so small? (barely a whopper) 

Practice match with Mr X - 2

I played another training match with Mr X yesterday.  Below are selected blunders from this live encounter.

***********

Opening game of 11 point match. 55 to play.

is Mr X score: 0 pip: 148

11 point match

pip: 144 score: 0 is Bill

XGID=-a--a-ECB---bC-Aacbe---A--:0:0:1:55:0:0:0:11:10

to play 55

1.	XG Roller++	8/3(2) 6/1*(2)	eq: +0.424
	Player: Opponent: 58.19% (G:22.48% B:0.89%) 41.81% (G:9.16% B:0.37%)
2.	XG Roller++	7/2(2) 6/1*(2)	eq: +0.364 (-0.059)
	Player: Opponent: 57.04% (G:24.73% B:0.80%) 42.96% (G:10.62% B:0.53%)
3.	XG Roller++	15/10 13/8 6/1*(2)	eq: +0.279 (-0.145)
	Player: Opponent: 55.03% (G:19.74% B:0.60%) 44.97% (G:9.82% B:0.39%)

eXtreme Gammon Version: 2.02, MET: Kazaross XG2

I moved 15/10, 13/8, 6/1(2).  This brings my man around into attack position should Mr X not anchor or hit me. 

But it violoates two key fundamentals from this position.

First, making a second inner point increases his dancing numbers from 4 to 9 - very important considersation here.  A dance and Mr X is screwed.

Second, it's almost always wrong to give the opponent good sixes from the bar in the early game.  In and a six hits me.

Perhaps a third fundamental would be once you've gone deep, commit to it.  I've made the ace point already so let's just do this thing.  It's too late to look pretty this game.

*************

Later in the same game. A techincal play - how best to move these 33's?

is Mr X score: 0 pip: 141

11 point match

pip: 143 score: 0 is Bill

XGID=-B-aa-BCC-A-bB--acbd---AaA:1:-1:1:33:0:0:0:11:10

to play 33

1.	XG Roller++	Bar/22 23/20 7/4*(2)	eq: +0.330
	Player: Opponent: 63.53% (G:27.60% B:1.28%) 36.47% (G:9.78% B:0.53%)
2.	XG Roller++	Bar/22 10/4* 7/4	eq: +0.316 (-0.014)
	Player: Opponent: 62.87% (G:27.29% B:1.27%) 37.13% (G:9.20% B:0.34%)
3.	4-ply	Bar/22 13/10 7/4*(2)	eq: +0.258 (-0.072)
	Player: Opponent: 60.57% (G:27.67% B:1.59%) 39.43% (G:10.09% B:0.52%)

eXtreme Gammon Version: 2.02, MET: Kazaross XG2

I played Bar/22, 7/4*(2) - obvious - plus 10/7 with the last 3.

The computer preferes Bar/22, 7/4*(2) - and 23/20 with the last 3.

A very small difference but it is interesting to figure out what is going on here.  Not sure I understand, but here goes:

• Being on the duece point sucks.  "Always move the back checker"

• This is about diversifing my numbers.  4's and 6's hit on his side of the board while 5's, 3's and 2's aim at the 5 point on my side of the board.

• I don't really care about getting hit so much here.  Any non-double hit by Mr X will leave himself severely exposed elsewhere.

• I am not trying to prime him - I am attacking.

*********************

 Later in the same game.  36 to play from the air.

is Mr X score: 0 pip: 180

11 point match

pip: 158 score: 0 is Bill

XGID=bB-aBaBBA---bB--Acbd---AAA:1:-1:1:63:0:0:0:11:10

to play 63

1.	XG Roller++	Bar/16	eq: +0.170
	Player: Opponent: 57.51% (G:26.87% B:1.41%) 42.49% (G:10.07% B:0.48%)
2.	XG Roller++	Bar/22 16/10	eq: +0.146 (-0.024)
	Player: Opponent: 56.23% (G:28.14% B:1.62%) 43.77% (G:10.68% B:0.57%)
3.	3-ply	Bar/22 13/7	eq: +0.086 (-0.084)
	Player: Opponent: 54.77% (G:26.89% B:1.59%) 45.23% (G:11.64% B:0.65%)

eXtreme Gammon Version: 2.02, MET: Kazaross XG2

I played Bar/22, 16/10.  Bringing more wood into the zone.

Computer suggests Bar/22, 22/16 instead. 

Again, not a terrible mistake but let's try to figure out why.  Once again, not entirely sure I know why, but here is my guess:

Attack is unlikely to succeed here.  Therefore I need to get my back guys into the game (outfield control)

I don't want to get my position too disconencted.  Where I am all offense, then suddenly Mr X Anchors and my back men are stranded.

So I am thinking my priorites were off - attacking is not my main game plan anymore since Mr X is about to anchor.

***************

I trail 10 away 7 away and Mr X is on roll.  Cube action?

is Mr X score: 4 pip: 135

11 point match

pip: 130 score: 1 is Bill

XGID=aA----EbCC---B---bdbbbA---:0:0:-1:00:1:4:0:11:10

on roll, cube action?

Analyzed in XG Roller++	No double	Double/Take
Player Winning Chances:	68.51% (G:26.63% B:0.71%)	68.66% (G:26.90% B:0.72%)
Opponent Winning Chances:	31.49% (G:6.31% B:0.23%)	31.34% (G:6.23% B:0.22%)
Cubeless Equities	+0.586	+1.160
Cubeful Equities
No double:	+0.703 (-0.089)
Double/Take:	+0.792
Double/Pass:	+1.000 (+0.208)
Best Cube action: Double / Take

eXtreme Gammon Version: 2.02, MET: Kazaross XG2

This looked super scary and I dropped.

Let's start with the basics

• Position - Large advantage Mr X.  I am totally undeveloped while he has an intimidating 5 prime.

• Race - Close but slight advantage to me.

• Threats - Any ace is great and Mr X will also hit with most 4's and also double 2's, and double 5's is a superjoker.

All in all, pretty scary but let's look at the virtues of this position.

• Even if Mr X does hit the ace I may make an anchor and play from there.

• Also he has three men to bring around.  Sometimes this could be difficult.

• While my position is undeveloped I do have all my checkers in play so the potential to make structure later.

• Oh, and Mr X might miss!! Also I am at the edge of the prime and may jump out immediately with a six.

The computer has it as a .06 take for money. Note trailing by 3 makes this a much larger take. My wins with a cube on 4 are very valuable.  So the match score has a big impact.