POPA66

Below are selected from an online match against POPA66.


1.  31 to play (9 away 9 away)

is POPA66 score: 0 pip: 161

9 point match

pip: 152 score: 0 is Chutzpah

XGID=-----bEaC--BcD-a-c-e----A-:0:0:1:31:0:0:0:9:6

to play 31

1.	XG Roller++	24/21 8/7*	eq: +0.111
	Player: Opponent: 53.83% (G:10.59% B:0.35%) 46.17% (G:11.29% B:0.34%)
2.	XG Roller++	24/20	eq: +0.067 (-0.044)
	Player: Opponent: 53.01% (G:8.88% B:0.30%) 46.99% (G:11.12% B:0.30%)
3.	XG Roller++	13/10 8/7*	eq: +0.043 (-0.068)
	Player: Opponent: 51.62% (G:12.26% B:0.55%) 48.38% (G:12.78% B:0.56%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

An awkward roll.  My play (24/20) isn't terrible but the concept is wrong.  24/20 makes since strategically since I want to escape my last man but has the obvious defect of coming under the gun.  I have limited options on the other side of the board that don't hit so I reluctantly stepped up.

As you can see, any hitting play was better.  The computer likes 24/21, stepping up to get out but not under direct sixes, while also hitting with 8/7* to slow him down.  I don't want to get hit, but frequently he misses.  And if he misses, I gain a lot.  Also, getting hit isn't so bad anyway.  Probably the biggest issue is I give the opponent his full roll to do something positive.  I was probably overly focused on not wanting to get another man sent back when I have only one back man.  This is true as far as it goes, but the alternatives in this specific position are simply worse.



2.  41 to play (8 away 8 away)

is POPA66 score: 1 pip: 153

9 point match

pip: 155 score: 1 is Chutzpah

XGID=--A-aBCbC---bC--Adbd---B--:0:0:1:41:1:1:0:9:6

to play 41

1.	XG Roller++	23/22 6/2	eq: -0.199
	Player: Opponent: 44.74% (G:12.17% B:0.38%) 55.26% (G:14.26% B:0.63%)
2.	XG Roller++	16/15 6/2	eq: -0.302 (-0.102)
	Player: Opponent: 41.93% (G:11.58% B:0.32%) 58.07% (G:14.02% B:0.61%)
3.	3-ply	23/22 8/4*	eq: -0.291 (-0.092)
	Player: Opponent: 43.65% (G:11.91% B:0.44%) 56.35% (G:17.95% B:1.15%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

This is another instructive error in concept.

I went for the "obvious" 16/15, 6/2.  This "safe" play fails to meet the strategic demands of the position and misses the tactical features of the position as well.

Much better was 6/2, 24/23. 

Strategically, being anchored on the deuce point is very bad.  I am blocked six away and in danger of getting blocked in.  I should WANT to strive for a better anchor and/or to simply jump over my opponent's budding prime. 

This position offers me several tactical opportunities to seek an upgrade immediately.  For the time being, I have a 3 to 1 board advantage so will be risky for opponent to hit loose anywhere.  Also if opponent breaks his bar point to point on me he will leave a man exposed - contact favors me while I have the better home board.  Also important is that opponent lacks a spare on the midpoint.  Breaking the point to hit me entails considerable risk since I have the better board (as already mentioned).

Finally, even if POPA66 points cleanly on the 2 point, strategically this is not the point he wants to make.  The 8 point and 2 point can never be part of the same prime, and he would prefer to make any of the 5/4/3 points instead.  And tactically, if he does point on my head on the 2 point, I have the chance to roll many good numbers.  And 3 as well as 21, 41 and 52 make a better anchor, while any entering number with a six lets me link up to the outfield with 22/16.

3.  61 to play (8 away 8 away)

is POPA66 score: 1 pip: 148

9 point match

pip: 150 score: 1 is Chutzpah

XGID=--B-aBBbC---bC-A-dbc---Ba-:0:0:1:61:1:1:0:9:6

to play 61

1.	XG Roller++	23/16	eq: +0.015
	Player: Opponent: 50.08% (G:13.53% B:0.47%) 49.92% (G:13.19% B:0.43%)
2.	XG Roller++	15/8	eq: -0.126 (-0.141)
	Player: Opponent: 46.17% (G:12.00% B:0.28%) 53.83% (G:12.11% B:0.33%)
3.	XG Roller+	23/22 15/9	eq: -0.018 (-0.033)
	Player: Opponent: 48.67% (G:14.01% B:0.45%) 51.33% (G:12.47% B:0.37%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

Very next roll and I still had the same blinders on.  I moved 15/8.

As discussed above, 23/16 is much better.  I need to pop out for both strategical and tactical reasons.  Note getting out is much more flexible for the future and gives me all sorts of playability.   If missed, my next several rolls will play naturally and easily. 

Tactically, opponent must break midpoint to hit and will be very risky where I have a 3 point board and he has a 1 point board with a blot.

A very poor play.


4.  Cube action (5 away 4 away)

is POPA66 score: 5 pip: 160

9 point match

pip: 133 score: 4 is Chutzpah

XGID=aBa--BBABA--dCA--bbe----A-:0:0:1:00:4:5:0:9:6

on roll, cube action?

Analyzed in XG Roller++	No double	Double/Take
Player Winning Chances:	69.14% (G:34.36% B:0.57%)	68.08% (G:40.02% B:0.50%)
Opponent Winning Chances:	30.86% (G:6.01% B:0.27%)	31.92% (G:7.29% B:0.49%)
Cubeless Equities	+0.641	+1.681
Cubeful Equities
No double:	+0.869 (-0.131)
Double/Take:	+1.423 (+0.423)
Double/Pass:	+1.000
Best Cube action: Double / Pass

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I misjudged this position.  I thought I was "too good" and took another roll.

The computer evaluation shows this is a proper Double / Pass.

I actually don't feel too bad about this one since the ingredients of a "too good" are there. 

a) Is opponent is dropping in a flash?  yes
b) Is there significant gammon potential?  yes
c) what is the risk of playing on?  not too big - nothing too bad can happen next time
d) worst case I figured I would double later on if he rolled something good

Apparently there are some sequences where opponent has a double and a take next time. 

I am unsure if I am interpreting this correctly but I will take a shot.  From the cube position, I went to dice distribution and looked at the Equity after all 1st and 2nd roll combinations.

I think the red bars indicate scenarios where I am still too good.  Similarly, I believe the green bars indicate scenarios where the position is a double and a take.

Restating the above, it looks like there are enough sequences where POPA66 has a take to justify simply cashing the game now.

Still, I don't feel bad about it.  The opponent may drop the subsequent take anyway, happy to just lose one point.  Or, the opponent correctly takes, and the most likely variation is I get rewarded for my mistake with a 2 or 4 point win.

5. 66 to play  (3 away 4 away and I own the cube)

is POPA66 score: 5 pip: 131

9 point match

pip: 172 score: 6 is Chutzpah

XGID=---BaAC-B--AbBba-b-cbbBB--:1:1:1:66:6:5:0:9:6

to play 66

1.	XG Roller++	22/4* 11/5	eq: -0.537
	Player: Opponent: 36.46% (G:11.62% B:0.54%) 63.54% (G:23.35% B:1.47%)
2.	XG Roller++	22/16(2) 11/5 8/2	eq: -0.645 (-0.109)
	Player: Opponent: 28.32% (G:5.83% B:0.15%) 71.68% (G:17.62% B:1.00%)
3.	3-ply	22/16(2) 13/7(2)	eq: -0.628 (-0.091)
	Player: Opponent: 29.80% (G:5.42% B:0.19%) 70.20% (G:18.57% B:1.01%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

So much for the back game.  Now what?

I tried 22/16(2) 11/5, 8/2.  I was too afraid to try and win going forward but that was the best play.  Probably I was overly spooked about getting gammoned for the match at this score.


22/4*, 11/5.   This is clearly the DMP play.  Cover the 5 point and hit him loose on the 4 point.  The computer shows this wins about 8% more often in exchange for getting gammoned 5+% more often.

This result is confusing me a bit.  His gammon value is almost 1.0 at this score so I would think I need to be gammon averse and the tradeoff above doesn't appear to justify the extra risk.

Maybe the fact that I win 6% more gammons factors in to the mix, as well as the fact that I might win some games with a recube later on?  I don't know.  Perhaps one of my readers can help me with this one.