Search This Blog

April 10, 2015

Cleveland semis, Frank Talbot

Below are selected errors from my semi-final match against Frank Talbot in Cleveland.


**************
1) I trail 9 away 8 away and am on roll.  Time for a pointless double?

is Frank Talbot

score: 1
pip: 177
9 point match
pip: 149
score: 0

is Bill Calton
XGID=aaa--AD-D--AdCA--c-da---A-:0:0:1:00:0:1:0:9:10
on roll, cube action?

Analyzed in XG Roller++ No double Double/Take
  Player Winning Chances: 66.54% (G:28.62% B:2.39%) 66.38% (G:29.25% B:2.58%)
  Opponent Winning Chances: 33.46% (G:6.32% B:0.25%) 33.62% (G:6.40% B:0.28%)
  Cubeless Equities +0.575 +1.207
Cubeful Equities
     No double:+0.855 (-0.102)
xg Double/Take:+0.957
     Double/Pass:+1.000 (+0.043)
Best Cube action: Double / Take
eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I was reluctant to send a pointless double since these are generally all easy takes.  However this particular position has some poison.  I actually have a  1.5 point board and Frank is already on the bar and my numbers are well diversified.  I am favored to simply make the point or hit a second blot or perhaps both.

Frank well have a lot of play as he is bound to anchor somewhere but this position is just too juicy to pass up.  I should have cubed to capture the significant gammon potential.

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

2) Score is 8 away 8 away.  I have already doubled.

54 to play.

is Frank Talbot

score: 1
pip: 136
9 point match
pip: 167
score: 1

is Bill Calton
XGID=---B-BBB-a--cC-b-bcd-BB---:1:-1:1:54:1:1:0:9:10
to play 54

1.XG Roller++21/16 13/9* eq: -0.308
Player:
Opponent:
45.30% (G:12.36% B:0.34%)
54.70% (G:12.53% B:0.51%)
2.XG Roller++13/8 13/9* eq: -0.403 (-0.095)
Player:
Opponent:
42.72% (G:11.61% B:0.36%)
57.28% (G:14.23% B:0.51%)
3.3-ply13/9* 9/4eq: -0.400 (-0.093)
Player:
Opponent:
42.36% (G:10.91% B:0.35%)
57.64% (G:14.18% B:0.48%)
eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I wasn't ready to give up the security of the double-anchor.  Therefore I went full offense mode with 13/8, 13/9*

Correct was the balanced 21/16, 13/9*.  I need to recognize this is not a backgame.  I would characterie this more as scramble mode, where I am trying to both attack and come around the board at the same time.  Popping out escapes one man from the broken four prime while keeping control over all sectors of the board.  Frank only has a one point board so any return shots are hardly fatal.

Compare the return hits from my play to the computer play.  They are not nearly as devastating after 21/16, 13/9*.

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

Most of my errors came from a long backgame.  The score was 6 away 6 away in the next several positions.

3) I am on roll.  Cube action?

is Frank Talbot

score: 3
pip: 182
9 point match
pip: 140
score: 3

is Bill Calton
XGID=-Aa-cbCBC-B--Ba-aA-eA-b---:0:0:1:00:3:3:0:9:10
on roll, cube action?

Analyzed in Rollout No double Double/Take
  Player Winning Chances: 68.27% (G:28.14% B:0.90%) 68.03% (G:27.92% B:0.87%)
  Opponent Winning Chances: 31.73% (G:6.44% B:0.47%) 31.97% (G:6.49% B:0.48%)
  Cubeless Equities +0.620 +1.294
Cubeful Equities
     No double:+0.957 (-0.043)±0.029 (+0.928..+0.986)
xg Double/Take:+1.027 (+0.027)±0.036 (+0.991..+1.064)
     Double/Pass:+1.000
Best Cube action: Double / Pass
Rollout details
774 Games rolled with Variance Reduction.
Moves: 3-ply, cube decisions: XG Roller
Double Decision confidence:99.8%
Take Decision confidence:92.9%
Duration: 9 minutes 49 seconds
eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I really didn't consider doubling at any point during this game.  I only have a one point board and I haven't assembled a prime yet.  I didn't see Frank as being in any danger until he is in danger of cracking.
 
The computer thinks I have a very strong position that is already a drop.  Let's see if we can see what it sees.  I have a broken four point block with six men already back.  Frank's stacked six point is bad and the made three point is deep.  I may send more checkers back, with the hope of splitting Frank's forces.  And Frank could roll an anti-joker to crack at any moment.

Despite the computer evaluation I don't feel bad at all about not doubling.  There is zero chance Frank would pass this.  Then, after he takes, I have to play the checkers. As the next positions will demonstrate I didn't have a good grasp on how to handle this position and made many very large mistakes.  So I am happy to retain cube access until the position clears itself up.

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

4) Same position.  53 to play.

is Frank Talbot

score: 3
pip: 182
9 point match
pip: 140
score: 3

is Bill Calton
XGID=-Aa-cbCBC-B--Ba-aA-eA-b---:0:0:1:53:3:3:0:9:10
to play 53

1.XG Roller++20/15 17/14* eq: +0.899
Player:
Opponent:
67.74% (G:26.50% B:0.80%)
32.26% (G:6.31% B:0.29%)
2.XG Roller++17/14* 14/9eq: +0.819 (-0.080)
Player:
Opponent:
66.46% (G:25.82% B:0.90%)
33.54% (G:7.08% B:0.30%)
3.3-ply17/14* 6/1eq: +0.786 (-0.113)
Player:
Opponent:
64.24% (G:30.09% B:0.95%)
35.76% (G:7.36% B:0.30%)
eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I hit on the 14 point and slotted the 9 point with 17/14*/9.  My strategy during this phase is to try and build the 9 point.  If i can do this I will establish a 5-point outside prime which will hopefully split Frank's forces and crash his front position.  Basically I am just putting my checkers where I want 'em since while I am unlikely to make the nine next turn the downside is very low.  Frank will find it very difficult to win going forward.

While I think this is the right approach, this is the wrong execution on this dice roll.  Better is 20/15, 17/14.*
This aims major wood at the nine point to either make the nine point naturally and/or hit loose whatever lands there.  The other technical aspect is getting my back man out of harms way.  In particular, Frank would be happy to unload his six point stack on head, get hit, and recycle these men around the board.  By getting my man out of Frank's homeboard Frank can't do this.  With nothing to hit, Frank is in more danger to crash at any moment.

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

5) 56 to play from the air.

is Frank Talbot

score: 3
pip: 181
9 point match
pip: 145
score: 3

is Bill Calton
XGID=-Aa-cbCBCaB--B---A-daab--A:0:0:1:65:3:3:0:9:10
to play 65

1.XG Roller++Bar/20* 8/2* eq: +0.873
Player:
Opponent:
66.25% (G:28.24% B:0.83%)
33.75% (G:7.04% B:0.32%)
2.XG Roller++Bar/20* 20/14eq: +0.783 (-0.090)
Player:
Opponent:
64.77% (G:26.93% B:0.86%)
35.23% (G:6.93% B:0.31%)
3.3-plyBar/20* 17/11eq: +0.742 (-0.132)
Player:
Opponent:
63.93% (G:27.61% B:1.05%)
36.07% (G:7.68% B:0.32%)
eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I moved Bar/20*/14.  I am still hoping to win the battle for the nine point.  While this makes strategical sense it is a tactical mistake.  A common theme in fighting the backgame is to try and recycle men on top of your primed points.  In this position, the spares on top of 8 point and 6 point are not doing anything.  By playing Bar/20*, 8/2* i give Frank 20 rolls that hit on either the ace or deuce point.  I would love to get hit and move these men back around the board where they fight to make the outside prime.

I had hit loose earlier on the ace point with the same idea so I was aware of this concept.  I think I was reluctant to put two checkers so deep on the gambit that they get hit.  Especially since there seems to be some possibility that Frank will pull his game forward.  If Frank ever abandons his five point anchor then all of a sudden my spares on the 8 and 6 are all of a sudden useful in hitting loose in the five point.

Still, as the computer shows 8/2* was the move.  I should mention the computer wants me to double this position too.  While this may be technically correct I am very comfortable waiting.  I am going to piss away much more equity through bad checker plays then I will gain from a theoretically correct double at this moment.

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

6) 61 to play.

is Frank Talbot

score: 3
pip: 190
9 point match
pip: 151
score: 3

is Bill Calton
XGID=--AabcCBBbB--BA--A-d-ab-A-:0:0:1:61:3:3:0:9:10
to play 61

1.XG Roller++24/18 14/13eq: +0.669
Player:
Opponent:
65.07% (G:18.22% B:0.44%)
34.93% (G:6.76% B:0.29%)
2.XG Roller++24/18 17/16eq: +0.602 (-0.067)
Player:
Opponent:
63.68% (G:18.16% B:0.46%)
36.32% (G:7.84% B:0.36%)
3.XG Roller+24/17eq: +0.644 (-0.025)
Player:
Opponent:
64.44% (G:18.94% B:0.51%)
35.56% (G:7.59% B:0.29%)
eXtreme Gammon Version: 2.10, MET: Kazaross XG2

Now the game has entered a new phase.  Frank has won the battle for my nine point.  This is a huge lifeline for Frank, which maintains his timing indefinitely.

I was confused as to what my proper strategy should be.  My approach was spread my blots around and try to force Frank to hit me and vacate the point.  After forcing him off the point, I would resume the initial phase of trying to conquer my nine point.  If I don't get hit, at least I keep my checkers in front of him and may make other points.

Therefore I played 24/18, 17/16.   The computer prefers 24/18, 14/13.  I am not confident as to what is going on here.  Maybe my idea is just too unlikely to work.  Perhaps I should just give up on trying to bust his timing?  I mean I tried hard but it just didn't work.  Should I just start bringing my guys around and see what happens?  If he vacates the nine point, I can still fight for it.  If not, I try to bear in safely and give a good double later.  If I get hit in the process, his offense isn't exactly ready so I should have several opportunities to escape before he will be anywhere near a cube.  The above reflects my current thinking and is subject to change.

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

7) 61 to play.

is Frank Talbot

score: 3
pip: 189
9 point match
pip: 145
score: 3

is Bill Calton
XGID=--A-bcCBBbBa-BaAA-Ad--b---:0:0:1:61:3:3:0:9:10
to play 61

1.XG Roller++16/10 15/14* eq: +0.571
Player:
Opponent:
61.87% (G:20.88% B:0.50%)
38.13% (G:7.64% B:0.34%)
2.XG Roller++18/11* eq: +0.498 (-0.073)
Player:
Opponent:
61.10% (G:18.84% B:0.47%)
38.90% (G:9.18% B:0.48%)
3.XG Roller+15/14* 14/8eq: +0.589 (+0.018)
Player:
Opponent:
62.33% (G:21.68% B:0.59%)
37.67% (G:7.76% B:0.31%)
eXtreme Gammon Version: 2.10, MET: Kazaross XG2

Consistent with my strategy, I moved 18/11*.  This place chums the water with my blots as bait.  And hopes to perhaps make the 11 point later.

The computer likes 16/10, 15/14*.  Once again I don't have a good feel for what the computer is trying to do.  Has it simply given up on trying to bust the timing?  is it trying to just race home safe while Frank has a very limited offense in place?

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

8) 54 to play.

is Frank Talbot

score: 3
pip: 186
9 point match
pip: 137
score: 3

is Bill Calton
XGID=--AabbCBBbBA-BaBa--d--b---:0:0:1:54:3:3:0:9:10
to play 54

1.XG Roller++15/11 15/10eq: +0.596
Player:
Opponent:
63.67% (G:16.54% B:0.27%)
36.33% (G:6.21% B:0.27%)
2.XG Roller++15/11 13/8eq: +0.478 (-0.118)
Player:
Opponent:
60.59% (G:18.07% B:0.51%)
39.41% (G:8.66% B:0.33%)
3.XG Roller+15/10 11/7eq: +0.549 (-0.047)
Player:
Opponent:
62.04% (G:18.76% B:0.48%)
37.96% (G:7.75% B:0.28%)
eXtreme Gammon Version: 2.10, MET: Kazaross XG2

Once again I went for maximum blots with 15/11, 13/8 while the computer prefers 15/11, 15/10.

The more I look at this and related positions the more I am thinking the bust the timing strategy just didn't work so I should shift gears and try to run home safe before Frank builds his board.

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

9) 11 to play

is Frank Talbot

score: 3
pip: 176
9 point match
pip: 128
score: 3

is Bill Calton
XGID=--A-bbCBCcBB-AaA---da-b---:0:0:1:11:3:3:0:9:10
to play 11

1.XG Roller++15/14* 14/12 13/12eq: +0.641
Player:
Opponent:
64.00% (G:18.85% B:0.28%)
36.00% (G:6.03% B:0.26%)
2.XG Roller++15/14* 14/12 8/7eq: +0.558 (-0.083)
Player:
Opponent:
62.19% (G:18.62% B:0.47%)
37.81% (G:7.42% B:0.30%)
3.XG Roller+15/14* 14/13 8/7 2/1eq: +0.605 (-0.036)
Player:
Opponent:
63.16% (G:19.23% B:0.28%)
36.84% (G:6.57% B:0.25%)
eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I can be stubborn.  I haven't given up on the blots everywhere strategy and went with 15/14*/12, 8/7

Superior was 15/14*, 14/12, 13/12.  Besides simply trying to come home safe, there is a tactical element at work here too.  I give double 3 in particular as a horror roll.

**********

10) 52 to play.


is Frank Talbot

score: 3
pip: 200
9 point match
pip: 123
score: 3

is Bill Calton
XGID=-A-bbbCBCcBBB------b-bb---:0:0:1:52:3:3:0:9:10
to play 52

1.XG Roller++8/1eq: +0.157
Player:
Opponent:
50.26% (G:23.51% B:0.43%)
49.74% (G:11.07% B:0.47%)
2.XG Roller++12/10 12/7eq: +0.063 (-0.094)
Player:
Opponent:
48.45% (G:21.75% B:0.41%)
51.55% (G:11.90% B:0.52%)
3.XG Roller+10/8 6/1eq: +0.126 (-0.031)
Player:
Opponent:
49.59% (G:23.22% B:0.53%)
50.41% (G:11.57% B:0.46%)
eXtreme Gammon Version: 2.10, MET: Kazaross XG2

Finally I gave up and am just trying to come home safe.  Thus 12/10, 12/7.

But I missed another tactical opportunity.  8/1 is anti-positional but keeps my mini three prime and makes any combination of 3-2-1 rolls awkward for Frank to play.

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

11) Double 1's.  What a great shot but how to play it?

is Frank Talbot

score: 3
pip: 178
9 point match
pip: 91
score: 3

is Bill Calton
XGID=-AAbbaEFBa-a--a----babb---:0:0:1:11:3:3:0:9:10
to play 11

1.XG Roller++7/5* 6/5 2/1eq: +0.486
Player:
Opponent:
54.62% (G:33.43% B:0.67%)
45.38% (G:7.42% B:0.24%)
2.XG Roller++7/5* 6/5(2)eq: +0.328 (-0.158)
Player:
Opponent:
52.65% (G:27.78% B:0.86%)
47.35% (G:9.12% B:0.33%)
3.3-ply6/5*(3) 2/1eq: +0.379 (-0.107)
Player:
Opponent:
52.90% (G:31.77% B:0.52%)
47.10% (G:8.84% B:0.29%)
eXtreme Gammon Version: 2.10, MET: Kazaross XG2

Best is simply 7/5*, 6/5 and 2/1.  This makes the 5 point and cleans up the blots.  I hope for a couple more good rolls and a strong cube later.

Now that I have a four prime, I reverted back to the play pure strategy.  I moved 7/5*, 6/5(2).  I am still hoping to be hit and recycle.  But a four prime isn't very good and I have no men to extend the prime.  Just the wrong idea.

***********

12) Next roll.  53 to play

is Frank Talbot

score: 3
pip: 183
9 point match
pip: 87
score: 3

is Bill Calton
XGID=aAAbbCCEBa-a--a----babb---:0:0:1:53:3:3:0:9:10
to play 53

1.XG Roller++8/5 7/2eq: +0.498
Player:
Opponent:
53.57% (G:35.77% B:1.29%)
46.43% (G:8.07% B:0.29%)
2.XG Roller++6/1 5/2eq: +0.252 (-0.246)
Player:
Opponent:
49.14% (G:33.37% B:0.59%)
50.86% (G:10.35% B:0.34%)
3.3-ply7/2 5/2eq: +0.323 (-0.175)
Player:
Opponent:
50.82% (G:33.44% B:0.94%)
49.18% (G:9.42% B:0.35%)
eXtreme Gammon Version: 2.10, MET: Kazaross XG2

Now I just said the heck with it and made the board with 8/5, 7/2.  Tough game backgammon.  Last move, leaving the blots was horrible.  This time, covering the blots is horrible.

XG recommends 8/5, 7/2.  I am quite confused comparing this play to the prior one and am ready to just shrug my shoulders and move on.

The builder distro is obviously superior this way and getting hit isn't fatal but I have to admit I just don't get it.

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

13) I lead 4 away 5 away.  21 to play.

is Frank Talbot

score: 4
pip: 148
9 point match
pip: 151
score: 5

is Bill Calton
XGID=--aB--EaB---cD---cbe---AA-:0:0:1:21:5:4:0:9:10
to play 21

1.XG Roller++24/22 23/22eq: -0.114
Player:
Opponent:
44.37% (G:11.19% B:0.29%)
55.63% (G:10.28% B:0.36%)
2.XG Roller++13/11 8/7* eq: -0.276 (-0.161)
Player:
Opponent:
42.79% (G:13.57% B:0.49%)
57.21% (G:18.23% B:1.27%)
3.XG Roller+24/21eq: -0.197 (-0.083)
Player:
Opponent:
43.71% (G:11.06% B:0.37%)
56.29% (G:13.60% B:0.46%)
eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I hit with 13/11, 8/7*  While I do have him outboarded this play is very unlikely to work!  It is so many return shots I can't even count that high.

The conservative 24/22, 23/22 was called for.  Making an anchor with a window out is always niec and the race is close so I don't want to get hit.

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

14) 4 away 4 away.  54S- 64.

is Frank Talbot

score: 5
pip: 158
9 point match
pip: 167
score: 5

is Bill Calton
XGID=-a---aE-C---dE---d-e----B-:0:0:1:64:5:5:0:9:10
to play 64

1.XG Roller++8/2 6/2eq: -0.218
Player:
Opponent:
45.17% (G:11.37% B:0.33%)
54.83% (G:14.33% B:0.66%)
2.XG Roller++24/14eq: -0.277 (-0.059)
Player:
Opponent:
45.03% (G:8.81% B:0.30%)
54.97% (G:15.14% B:0.51%)
3.XG Roller+13/3eq: -0.309 (-0.091)
Player:
Opponent:
43.66% (G:10.42% B:0.34%)
56.34% (G:15.31% B:0.76%)
eXtreme Gammon Version: 2.10, MET: Kazaross XG2

I always get confused on how to play 64 reply roll.  I used to have a rule of thumb that was to always make the 2 point when opponent splits to the 5 point and you miss.  As I recall there is an exception or two to this rule and sometimes 24/14 is right.  But I don't remember.

I ran with 24/14 while 8/2, 6/2 is right.  Do any of my readers have insight into this reply roll?

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

15) 4 away 4 away.  Cube action?


is Frank Talbot

score: 5
pip: 38
9 point match
pip: 37
score: 5

is Bill Calton
XGID=--B-DAB------------babd-a-:0:0:1:00:5:5:0:9:10
on roll, cube action?

Analyzed in XG Roller++ No double Double/Take
  Player Winning Chances: 72.62% (G:0.00% B:0.00%) 72.59% (G:0.00% B:0.00%)
  Opponent Winning Chances: 27.38% (G:0.00% B:0.00%) 27.41% (G:0.00% B:0.00%)
  Cubeless Equities +0.452 +1.066
Cubeful Equities
     No double:+0.707 (-0.077)
xg Double/Take:+0.784
     Double/Pass:+1.000 (+0.216)
Best Cube action: Double / Take
eXtreme Gammon Version: 2.10, MET: Kazaross XG2

With only a one pip lead I judged this to be too easy of a take and just rolled.

However it is a big Double and I should have gone for it.

Keith count comes up 43 to 39 which implies Double / Take and works for this position.  A double is a double right?

1 comment:

  1. Hi Bill,

    regarding 11 versus 12: In 12 it is not that it's bad to cover the blots, but that you create an inflexible position for the bears by taking away your spares on the 5 and 6. It is in particular bad not to have spares on the 5 when the bears starts in this position because they are the only 3s you can play against this structure.

    The difference is likely coming from the fact that it is a lot worse have two checkers sent back than one. In 11 you cover in order not to have 2 blots, if you could, given the roll, playing with one blot would ok. In 12

    ReplyDelete