Discussion:
Nice boxcars: what's your play?
(too old to reply)
Tim Chow
2020-05-29 15:27:09 UTC
Permalink
XGID=-AaBBbC-B--BbB-A--acbbb---:1:-1:1:66:0:0:0:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X X O | | O O O O | +---+
| X | | O O O O | | 2 |
| | | O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| O X X | | X O X X |
| O X X | | X O X X O X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 112 O: 138 X-O: 0-0
Cube: 2, O own cube
X to play 66

---
Tim Chow
Paul Epstein
2020-05-29 17:59:49 UTC
Permalink
Post by Tim Chow
XGID=-AaBBbC-B--BbB-A--acbbb---:1:-1:1:66:0:0:0:0:10
X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X X O | | O O O O | +---+
| X | | O O O O | | 2 |
| | | O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| O X X | | X O X X |
| O X X | | X O X X O X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 112 O: 138 X-O: 0-0
Cube: 2, O own cube
X to play 66
Well, duh!
I'm going to point on the two point and make my
acepoint.
95% confidence.
I won't be shocked if I'm wrong because it is a quiz after all.
Of course, I don't mean that I would bet at those odds.
What I'm saying is that, if this position was not a quiz, I would
find the fact that my play is wrong to be as surprising as a 5% event.

Paul
b***@gmail.com
2020-05-29 18:56:42 UTC
Permalink
Post by Paul Epstein
Post by Tim Chow
XGID=-AaBBbC-B--BbB-A--acbbb---:1:-1:1:66:0:0:0:0:10
X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X X O | | O O O O | +---+
| X | | O O O O | | 2 |
| | | O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| O X X | | X O X X |
| O X X | | X O X X O X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 112 O: 138 X-O: 0-0
Cube: 2, O own cube
X to play 66
Well, duh!
I'm going to point on the two point and make my
acepoint.
95% confidence.
I won't be shocked if I'm wrong because it is a quiz after all.
Of course, I don't mean that I would bet at those odds.
What I'm saying is that, if this position was not a quiz, I would
find the fact that my play is wrong to be as surprising as a 5% event.
Paul
Is that your play at dmp?

Stick
Paul Epstein
2020-05-29 21:20:29 UTC
Permalink
Post by b***@gmail.com
Post by Paul Epstein
Post by Tim Chow
XGID=-AaBBbC-B--BbB-A--acbbb---:1:-1:1:66:0:0:0:0:10
X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X X O | | O O O O | +---+
| X | | O O O O | | 2 |
| | | O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| O X X | | X O X X |
| O X X | | X O X X O X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 112 O: 138 X-O: 0-0
Cube: 2, O own cube
X to play 66
Well, duh!
I'm going to point on the two point and make my
acepoint.
95% confidence.
I won't be shocked if I'm wrong because it is a quiz after all.
Of course, I don't mean that I would bet at those odds.
What I'm saying is that, if this position was not a quiz, I would
find the fact that my play is wrong to be as surprising as a 5% event.
Paul
Is that your play at dmp?
Stick
It's hard to say what I would have done at DMP because the fact
that you're asking the question introduces a bias.

My belief is that, at DMP, 13/7(2) 15/3 is correct for maximum safety.
Hard to say what my belief would have been if I'd seen the position
OTB at DMP for the first time.

So, if I'm right it breaks your DMP rule.

Paul
Tim Chow
2020-05-30 16:09:43 UTC
Permalink
Post by Paul Epstein
So, if I'm right it breaks your DMP rule.
A good example here of a proof by contradiction.

---
Tim Chow
Tim Chow
2020-05-31 14:59:06 UTC
Permalink
XGID=-AaBBbC-B--BbB-A--acbbb---:1:-1:1:66:0:0:0:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X X O | | O O O O | +---+
| X | | O O O O | | 2 |
| | | O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| O X X | | X O X X |
| O X X | | X O X X O X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 112 O: 138 X-O: 0-0
Cube: 2, O own cube
X to play 66

There are two natural candidate plays here: blitzing, and playing safe.
Blitzing clearly wins a lot more gammons. But O's board is strong. Does
the blitzing play lose too often? According to the rollout below, yes.
The fact that the blitzing play leaves some return shots is part of the
story, but not all of it; see Variant 1.

See also Variant 2, where I have weakened O's forward position. This
illustrates my point about a "proof by contradiction." Had I posted
Variant 2, and Paul had proposed blitzing, Stick would of course *not*
have asked, "Is that your play at DMP?" even though one might naively
think that it would be a reasonable question to ask. So the fact that
Stick asked the question proves not only that Paul's play in the original
position was not the bot's DMP play, but also that it was not the bot's
money play.

1. Rollout¹ 15/3 13/7(2) eq:+0.548
Player: 76.22% (G:12.76% B:0.09%)
Opponent: 23.78% (G:3.16% B:0.09%)
Confidence: ±0.006 (+0.543..+0.554) - [100.0%]

2. Rollout¹ 13/1 8/2*(2) eq:+0.432 (-0.116)
Player: 68.66% (G:22.54% B:0.19%)
Opponent: 31.34% (G:5.25% B:0.14%)
Confidence: ±0.006 (+0.426..+0.438) - [0.0%]

¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller

eXtreme Gammon Version: 2.19.207.pre-release

---------
Variant 1
---------

XGID=-AaBBbC-B--BbA-B--acbbb---:1:-1:1:66:0:0:0:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X X O | | O O O O | +---+
| X | | O O O O | | 2 |
| | | O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| O X X | | X O X X |
| O X X | | X O X X O X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 114 O: 138 X-O: 0-0
Cube: 2, O own cube
X to play 66

1. Rollout¹ 15/9(2) 13/1 eq:+0.484
Player: 75.16% (G:7.76% B:0.05%)
Opponent: 24.84% (G:2.93% B:0.08%)
Confidence: ±0.005 (+0.479..+0.489) - [100.0%]

2. Rollout¹ 13/1 8/2*(2) eq:+0.431 (-0.053)
Player: 69.19% (G:20.25% B:0.11%)
Opponent: 30.81% (G:4.31% B:0.12%)
Confidence: ±0.006 (+0.424..+0.437) - [0.0%]

¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller

eXtreme Gammon Version: 2.19.207.pre-release

---------
Variant 2
---------

XGID=-AaBBbC-B--BbB-A-aacbba---:1:-1:1:66:0:0:0:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X X O O | | O O O O | +---+
| X | | O O O | | 2 |
| | | O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| O X X | | X O X X |
| O X X | | X O X X O X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 112 O: 143 X-O: 0-0
Cube: 2, O own cube
X to play 66

1. Rollout¹ 13/1 8/2*(2) eq:+0.614
Player: 73.45% (G:27.30% B:0.28%)
Opponent: 26.55% (G:3.53% B:0.13%)
Confidence: ±0.006 (+0.608..+0.620) - [100.0%]

2. Rollout¹ 15/3 13/7(2) eq:+0.578 (-0.036)
Player: 76.17% (G:15.92% B:0.11%)
Opponent: 23.83% (G:3.31% B:0.11%)
Confidence: ±0.005 (+0.572..+0.583) - [0.0%]

¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller

eXtreme Gammon Version: 2.19.207.pre-release

---
Tim Chow
b***@gmail.com
2020-05-31 15:39:33 UTC
Permalink
Post by Tim Chow
XGID=-AaBBbC-B--BbB-A--acbbb---:1:-1:1:66:0:0:0:0:10
X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X X O | | O O O O | +---+
| X | | O O O O | | 2 |
| | | O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| O X X | | X O X X |
| O X X | | X O X X O X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 112 O: 138 X-O: 0-0
Cube: 2, O own cube
X to play 66
There are two natural candidate plays here: blitzing, and playing safe.
Blitzing clearly wins a lot more gammons. But O's board is strong. Does
the blitzing play lose too often? According to the rollout below, yes.
The fact that the blitzing play leaves some return shots is part of the
story, but not all of it; see Variant 1.
See also Variant 2, where I have weakened O's forward position. This
illustrates my point about a "proof by contradiction." Had I posted
Variant 2, and Paul had proposed blitzing, Stick would of course *not*
have asked, "Is that your play at DMP?" even though one might naively
think that it would be a reasonable question to ask. So the fact that
Stick asked the question proves not only that Paul's play in the original
position was not the bot's DMP play, but also that it was not the bot's
money play.
1. Rollout¹ 15/3 13/7(2) eq:+0.548
Player: 76.22% (G:12.76% B:0.09%)
Opponent: 23.78% (G:3.16% B:0.09%)
Confidence: ±0.006 (+0.543..+0.554) - [100.0%]
2. Rollout¹ 13/1 8/2*(2) eq:+0.432 (-0.116)
Player: 68.66% (G:22.54% B:0.19%)
Opponent: 31.34% (G:5.25% B:0.14%)
Confidence: ±0.006 (+0.426..+0.438) - [0.0%]
¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller
eXtreme Gammon Version: 2.19.207.pre-release
---------
Variant 1
---------
XGID=-AaBBbC-B--BbA-B--acbbb---:1:-1:1:66:0:0:0:0:10
X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X X O | | O O O O | +---+
| X | | O O O O | | 2 |
| | | O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| O X X | | X O X X |
| O X X | | X O X X O X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 114 O: 138 X-O: 0-0
Cube: 2, O own cube
X to play 66
1. Rollout¹ 15/9(2) 13/1 eq:+0.484
Player: 75.16% (G:7.76% B:0.05%)
Opponent: 24.84% (G:2.93% B:0.08%)
Confidence: ±0.005 (+0.479..+0.489) - [100.0%]
2. Rollout¹ 13/1 8/2*(2) eq:+0.431 (-0.053)
Player: 69.19% (G:20.25% B:0.11%)
Opponent: 30.81% (G:4.31% B:0.12%)
Confidence: ±0.006 (+0.424..+0.437) - [0.0%]
¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller
eXtreme Gammon Version: 2.19.207.pre-release
---------
Variant 2
---------
XGID=-AaBBbC-B--BbB-A-aacbba---:1:-1:1:66:0:0:0:0:10
X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X X O O | | O O O O | +---+
| X | | O O O | | 2 |
| | | O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| O X X | | X O X X |
| O X X | | X O X X O X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 112 O: 143 X-O: 0-0
Cube: 2, O own cube
X to play 66
1. Rollout¹ 13/1 8/2*(2) eq:+0.614
Player: 73.45% (G:27.30% B:0.28%)
Opponent: 26.55% (G:3.53% B:0.13%)
Confidence: ±0.006 (+0.608..+0.620) - [100.0%]
2. Rollout¹ 15/3 13/7(2) eq:+0.578 (-0.036)
Player: 76.17% (G:15.92% B:0.11%)
Opponent: 23.83% (G:3.31% B:0.11%)
Confidence: ±0.005 (+0.572..+0.583) - [0.0%]
¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller
eXtreme Gammon Version: 2.19.207.pre-release
---
Tim Chow
Again, part of the reason for the rule is to get people away from thinking about gammons won and lost. It often steers them wrong.

Stick
Tim Chow
2020-05-31 17:49:22 UTC
Permalink
Post by b***@gmail.com
Again, part of the reason for the rule is to get people away from thinking
about gammons won and lost. It often steers them wrong.
The lesson is more convincing if you don't put your thumb on the scale
with such strong selection bias.

---
Tim Chow
Paul Epstein
2020-05-31 22:03:59 UTC
Permalink
Post by Tim Chow
Post by b***@gmail.com
Again, part of the reason for the rule is to get people away from thinking
about gammons won and lost. It often steers them wrong.
The lesson is more convincing if you don't put your thumb on the scale
with such strong selection bias.
To add info to this debate seems empirically straightforward.
Suppose a player, at all match scores, always made the DMP play.
What would the strength of such a player be?
On the one hand, their play would always be correct at some score.
On the other hand, any play which required sacrificing wins for gammons
would be eschewed.
I think that, if such an approach still results in a strong player (say 5.0
approx) then Stick might have a point. But, if the approach is hopeless,
and I have to listen to Stick saying "I don't mean to consider situations
where deviating from DMP is obvious", then I begin to lose patience.

Paul
b***@gmail.com
2020-06-01 14:17:15 UTC
Permalink
Post by Paul Epstein
Post by Tim Chow
Post by b***@gmail.com
Again, part of the reason for the rule is to get people away from thinking
about gammons won and lost. It often steers them wrong.
The lesson is more convincing if you don't put your thumb on the scale
with such strong selection bias.
To add info to this debate seems empirically straightforward.
Suppose a player, at all match scores, always made the DMP play.
What would the strength of such a player be?
On the one hand, their play would always be correct at some score.
On the other hand, any play which required sacrificing wins for gammons
would be eschewed.
I think that, if such an approach still results in a strong player (say 5.0
approx) then Stick might have a point. But, if the approach is hopeless,
and I have to listen to Stick saying "I don't mean to consider situations
where deviating from DMP is obvious", then I begin to lose patience.
Paul
I have never tried it across all scores but I have often done the experiment of playing dmp games then changing the exported code to make them money games and analyze it from there. You have to of course rid yourself of the issues that come when the cube gets involved and many other obvious such plays that one would make at dmp but not money games, and if you can do that always making the dmp play you're the best player in the world.

Stick
b***@gmail.com
2020-06-01 14:14:57 UTC
Permalink
Post by Tim Chow
Post by b***@gmail.com
Again, part of the reason for the rule is to get people away from thinking
about gammons won and lost. It often steers them wrong.
The lesson is more convincing if you don't put your thumb on the scale
with such strong selection bias.
---
Tim Chow
It's only selection bias to others potentially, not me. I can't help if things aren't obvious to other people and they don't believe me when I say they are to me.

Stick
Tim Chow
2020-06-02 00:35:11 UTC
Permalink
Post by b***@gmail.com
It's only selection bias to others potentially, not me.
By selection bias, I mean that you never ask, "Is that your play at DMP?"
when someone like Paul makes a non-DMP play in response to a quiz, and
the non-DMP play is the bot play. You only ever ask that question when the
non-DMP play is not the bot play.

In other words the "selection" I'm talking about is your selection of which
problems to comment on.

---
Tim Chow
Paul Epstein
2020-06-02 11:50:36 UTC
Permalink
Post by b***@gmail.com
Post by Tim Chow
Post by b***@gmail.com
Again, part of the reason for the rule is to get people away from thinking
about gammons won and lost. It often steers them wrong.
The lesson is more convincing if you don't put your thumb on the scale
with such strong selection bias.
---
Tim Chow
It's only selection bias to others potentially, not me. I can't help if things aren't obvious to other people and they don't believe me when I say they are to me.
I don't understand this. The situation seems as follows:
You suggest a rule for others to follow.
The rule uses the word "obvious".
The people trying to follow the rule don't know what is "obvious" and
so can't follow your rule.
You then say "Well, these exceptions are obvious to me."
I doubt that anyone disbelieves you -- that isn't the issue.

I and I think also Tim assumed that the rule was attempted for the
guidance of a wide range of players including intermediates.
If you say, that obvious means "obvious to Stick" then it only works as
a rule for world-class players.
If you present it as a rule for world-class players, then no probs.
But it makes no sense to say it's a rule for intermediates, but that it's
not the rule's fault if intermediates can't follow it because they don't
know what is "obvious".

Paul
b***@gmail.com
2020-06-02 17:56:10 UTC
Permalink
Post by Paul Epstein
Post by b***@gmail.com
Post by Tim Chow
Post by b***@gmail.com
Again, part of the reason for the rule is to get people away from thinking
about gammons won and lost. It often steers them wrong.
The lesson is more convincing if you don't put your thumb on the scale
with such strong selection bias.
---
Tim Chow
It's only selection bias to others potentially, not me. I can't help if things aren't obvious to other people and they don't believe me when I say they are to me.
You suggest a rule for others to follow.
The rule uses the word "obvious".
The people trying to follow the rule don't know what is "obvious" and
so can't follow your rule.
You then say "Well, these exceptions are obvious to me."
I doubt that anyone disbelieves you -- that isn't the issue.
I and I think also Tim assumed that the rule was attempted for the
guidance of a wide range of players including intermediates.
If you say, that obvious means "obvious to Stick" then it only works as
a rule for world-class players.
If you present it as a rule for world-class players, then no probs.
But it makes no sense to say it's a rule for intermediates, but that it's
not the rule's fault if intermediates can't follow it because they don't
know what is "obvious".
Paul
Just like any backgammon saying, the more you implement it and practice with it the more useful it will become. I tell everyone when in doubt, make the five point. I tell them when you're not sure, pay later. If you don't know whether it's a take or not then you double. Ad nauseam...there are plenty of these shortcuts that have their uses and the more fine tuned your own game is the more you can make use of such maxims. If you told a complete novice to always make the five point when he had the option, you've improved his game. When I tell you to make the dmp play when it's clear except for obvious exceptions, theoretically if you follow it I've improved your game. It's up to you to hone in on when it applies and when it doesn't.

Stick
Paul Epstein
2020-06-02 18:09:49 UTC
Permalink
Post by b***@gmail.com
Post by Paul Epstein
Post by b***@gmail.com
Post by Tim Chow
Post by b***@gmail.com
Again, part of the reason for the rule is to get people away from thinking
about gammons won and lost. It often steers them wrong.
The lesson is more convincing if you don't put your thumb on the scale
with such strong selection bias.
---
Tim Chow
It's only selection bias to others potentially, not me. I can't help if things aren't obvious to other people and they don't believe me when I say they are to me.
You suggest a rule for others to follow.
The rule uses the word "obvious".
The people trying to follow the rule don't know what is "obvious" and
so can't follow your rule.
You then say "Well, these exceptions are obvious to me."
I doubt that anyone disbelieves you -- that isn't the issue.
I and I think also Tim assumed that the rule was attempted for the
guidance of a wide range of players including intermediates.
If you say, that obvious means "obvious to Stick" then it only works as
a rule for world-class players.
If you present it as a rule for world-class players, then no probs.
But it makes no sense to say it's a rule for intermediates, but that it's
not the rule's fault if intermediates can't follow it because they don't
know what is "obvious".
Paul
Just like any backgammon saying, the more you implement it and practice with it the more useful it will become. I tell everyone when in doubt, make the five point. I tell them when you're not sure, pay later. If you don't know whether it's a take or not then you double. Ad nauseam...there are plenty of these shortcuts that have their uses and the more fine tuned your own game is the more you can make use of such maxims. If you told a complete novice to always make the five point when he had the option, you've improved his game. When I tell you to make the dmp play when it's clear except for obvious exceptions, theoretically if you follow it I've improved your game. It's up to you to hone in on when it applies and when it doesn't.
Thanks for these thoughtful comments.
I'm an approx 6.0 player.
I'm sure that I would double if I didn't know if it was a take or not.
Re the 5 point and paying later, I'm not so confident. I've made so
many blunders from making the 5 point when I should have played something
different and paying later when I should have paid now.

But I am a good player by most normal standards although most of the
regulars here are even better than me.

Paul
Tim Chow
2020-06-04 04:37:56 UTC
Permalink
Post by Paul Epstein
Re the 5 point and paying later, I'm not so confident. I've made so
many blunders from making the 5 point when I should have played something
different and paying later when I should have paid now.
Mark Dvoretsky, the famous chess coach, used to say, "Serious work with
young players usually begins with a diagnosis of the student's strength
and weaknesses." If one is writing a book or trying to issue some kind
of one-size-fits-all advice, then spouting maxims is understandable. On
the other hand, if there is a specific player whose play you are trying
to improve (yourself, or your student), then far more important than any
maxims is the development of a profile of the specific strengths and
weaknesses of that player.

If you have identified that when it comes to making the 5pt, or paying now
versus later, the standard cliche isn't working for you, then you have
gained an important insight about yourself. Ignore the cliche and focus
on the data you have about yourself. Figure out what your incorrect biases
are, and work to correct them. Of course, if you don't have access to a bot,
this is difficult. But if you're serious about improving, you can collect
examples where someone else has provided bot information, and study those.

Maxims are fine up to a point, but they are no substitute for hard data about
your personal predilections.

---
Tim Chow
Paul Epstein
2020-06-04 11:50:32 UTC
Permalink
Post by Tim Chow
Post by Paul Epstein
Re the 5 point and paying later, I'm not so confident. I've made so
many blunders from making the 5 point when I should have played something
different and paying later when I should have paid now.
Mark Dvoretsky, the famous chess coach, used to say, "Serious work with
young players usually begins with a diagnosis of the student's strength
and weaknesses." If one is writing a book or trying to issue some kind
of one-size-fits-all advice, then spouting maxims is understandable. On
the other hand, if there is a specific player whose play you are trying
to improve (yourself, or your student), then far more important than any
maxims is the development of a profile of the specific strengths and
weaknesses of that player.
If you have identified that when it comes to making the 5pt, or paying now
versus later, the standard cliche isn't working for you, then you have
gained an important insight about yourself. Ignore the cliche and focus
on the data you have about yourself. Figure out what your incorrect biases
are, and work to correct them. Of course, if you don't have access to a bot,
this is difficult. But if you're serious about improving, you can collect
examples where someone else has provided bot information, and study those.
Maxims are fine up to a point, but they are no substitute for hard data about
your personal predilections.
The reason I don't have a bot is that, far from being serious about
improving, I'm actually trying to limit my backgammon time. If I do
become serious about improving, I will obtain a bot.
Your post is correct, of course, and I think Dvoretsky was a great coach.
A particularly impressive aspect to Dvoretsky's success is that he
was never a world-class player himself. This suggests that the success
of his trainees is largely due to his coaching.
One trick that coaches have is to obtain great junior players, then
observe that they often become great adult players, and then claim that
this is due to your coaching. This trick is unlikely to work if you
don't have an initial branding based on being an elite GM.

Paul

Loading...