Forum Archive : Learning

Committing to memory

From:   RobertFontaine
Address:   Robert.Fontaine@shaw.ca
Date:   14 February 2011
Subject:   Committing to Memory
Forum:   BGonline.org Forums

In my spare time I've started to attempt to commit 2nd roll moves to
memory.  I've been making flash cards with mnemosyne and doing rote
drilling to attempt to get the moves to stick. I've started simply by
enumerating the rolls by numeric sequence but as the volume increases it
seems to make more sense to learn and group them by their "nature," i.e. P,
R, $, D etc.

I'm wondering what tricks you use to organize, memorize and recall the 2nd
roll positions and more generally how do you develop a stable collection of
reference positions?

Thanks, Robert

Matt Cohn-Geier  writes:

When I first started doing this in 2008 I wrote down all 441 rolls and
responses and how I would play them. Then I went through and noted what I
got wrong. Then I did it again, writing down all 441 rolls and responses.
Eventually I got them all right. I haven't had a problem remembering it
since, although I did pick up some more information as newer rollouts came

David Rockwell  writes:

I tried for a long time to learn the 2nd move by reply roll. In other
words, I would try to learn the 11 responses to all opening move, etc. Even
though this made sense to me, it didn't work. Then Stick talked me into
learning replies grouped by the opening move. This worked great. I can't
articulate why the one worked so much better than the other, but it did.

Nack Ballard  writes:

Funny, I had the opposite experience. I tried grouping by opening moves,
then switched early on and found it much easier to learn grouping plays by
response roll.

Bob Koca  writes:

I practice both ways. If had to choose one by reply works better for me.

Stick  writes:

I think the best way to memorize them is to break them up into small blocks
of information. I would start by learning all the replies to the opening
rolls that everyone plays the same, 31P, 42P, 53P, 61P, 65R. Be patient.
Don't try to do it all overnight. That may commit it to short term memory
for a hot minute but it will fade and you will end up being aggravated.
Instead on the first day take your time and look at all the replies to 31P.
Look at the position in a bot, it's very easy to set it up and change the
dice to see the actual position, and then look at the rollout results and
make sense of what's going on.

For example, after an opening 31P-62/63/64 you're supposed to
split/split/run respectively. Why is that? Does the difference in the
rollout results from one play to the next make sense? Yes, it does. En
bref, after 31P-62? if you run you leave the most shots, have the smallest
racing lead, are least likely to get your checker to safety, and don't
create as many good outfield point making numbers when not hit. After
31P-63? your racing lead is larger than a 62, it's less shots, it aims at
more of your outfield however your point making numbers are somewhat
duplicated since the 15 point and 13 point are two pips apart as are the 8
point and 6 point. If you look at the rollout of 31P-62 and 31P-63 given
what you noted about the position you would expect that it is not as wrong
to run with a 63 as it is with a 62. This is confirmed by the rollouts:

    * 31P-62 [S R34] /5
    * 31P-63 [S R21] /15

Now if you think about how to play your 64 after 31P-64? you should realize
that you're now ahead more in the race, your checker is a lot more likely
to get to safety when missed, you've minimized shots, and you have
diversified your point making numbers in comparison to 31P-63R. If the
swing from 62 to 63 was roughly .013 you would expect the swing to be at
least that much and likely more than that since it improved in every
aspect. Checking the rollout results it swung a lot and running is actually
correct now.

    * 31P-62 [S R34] /5
    * 31P-63 [S R21] /15
    * 31P-64 [R P19 S21] /5

Now after the first day of doing the responses to an opening 31P you get to
day two and you review the replies to an opening 31P and move on to the
replies to an opening 42P. On day three you review the replies to both 31P
and 42P and add on the opening 53P replies. Etc... If you really want to
understand, memorize, and properly apply the replies this is the best
method I suggest. Not only will you realize things like what I noted above
but in doing in a slow methodical way you will also 'accidentally' commit
to memory how close plays are and that can swing how you play in match
situations or against stronger/weaker opponents.

For example, I can mentally run through all the replies because I've looked
at them and studied them so damn much and tell you what plays are
considered tied. After an opening 31P it would be [11 32]. If you know for
money 31P-11 that S=N and you're trailing 5 away 2 away it's safe to say
that N is clearly correct. The other side of the coin is that with any lead
almost splitting will become more and more clear.


Reference positions in general are a different beast. While I have a good
idea of opening phase of the game blitzes for example, or holding games or
bear off positions, most parts of the game you can't memorize reference
positions for.  I find it's more useful when you hit a position that you
have trouble with playing it out over and over and over and over and over
and over and over and over and over and over and over and over and over and
over again until you understand what is going on.

Here is a position that I've recently used to demonstrate how important it
is to pick up a second checker, how you best go about picking up that
second checker, what you are trading off on certain plays (gammon losses v.
wins, this is handy to know if the position comes up at dmp), and when you
should go about recubing such positions.

       13  14  15  16  17  18      19  20  21  22  23  24
      |     X   X             |   |                     O |  OOO
      |                       |   |                       |  OOO
      |                       |   |                       |  OOO
      |                       |   |                       |  OO
      |                       |   |                       |  OO
      |                       |   |                       |
      |                       | O |                       |
      |                       |   |                       |
      |                       |   |                       |
      |                 X   X |   | X   X   X   X         |
      |             X   X   X |   | X   X   X   X         |
       12  11  10   9   8   7       6   5   4   3   2   1

Reference positions that you should 'know' but no need committed to
absolute memory are 'holy shit, I win a TON more games with two checkers
closed out compared to one checker closed out'. From memory I believe it's
21.5% with one closed out compared to 68.5% with both men closed out. The
exact percentages aren't important. What is important is that you remember
I need that second checker!
Did you find the information in this article useful?          

Do you have any comments you'd like to add?     



Advancing beyond intermediate  (James Eibisch, July 1998) 
Beginners' mistakes  (Alan Webb+, Nov 1999)  [Long message] [Recommended reading]
Best way for a beginner to learn  (Koyunbaba+, July 2007)  [Long message]
Committing to memory  (RobertFontaine+, Feb 2011) 
Getting better than "awful"  (Morph+, May 2004) 
How to excel in backgammon  (Max Levenstein+, Aug 2011) 
How to improve  (N Merrigan, Jan 2007) 
How to improve  (Albert Steg, Feb 1996) 
How to improve cube handling  (RealNick+, Jan 2011) 
How to learn and improve  (Hristov, Aug 2005) 
Lowering your error rate  (Stick Rice+, Apr 2009) 
Maintaining your game  (Robert-Jan Veldhuizen, Apr 2005)  [GammOnLine forum]
Matchqiz and Jellyfish  (Gilles Baudrillard, May 1997) 
Missing candidate plays  (Klaus Evers+, Apr 2009) 
Most efficient way to learn  (Stick+, May 2007)  [Long message]
Practice and preparation  (Ian Shaw+, Mar 2004)  [GammOnLine forum]
Practice/study plan  (Marcus Brooks+, Nov 1995) 
Reference positions  (Chuck Bower, July 1999) 
Study Methodology  (Phil Simborg, Dec 2012) 
Study method  (Jason Lee+, Jan 2012) 
Study plan  (Tenland+, Nov 2012) 
Taking your game up a level  (CW+, Aug 2002) 
Taking your game up a level  (Ron Karr, Aug 1996) 
The backgammon cake  (Daniel Murphy, Nov 1997) 
The best way to learn  (Chuck Bower+, Oct 2003)  [GammOnLine forum] [Long message]
Three steps to better play  (David Montgomery, July 1998) 
Using Jellyfish tutor  (Stephen Hubbard, Sept 1997) 
What more can I do?  (Alison Wylie+, Apr 2000) 
Zen in the art of backgammon  (Robban+, Aug 2009) 

[GammOnLine forum]  From GammOnLine       [Long message]  Long message       [Recommended reading]  Recommended reading       [Recent addition]  Recent addition

  Book Suggestions
Computer Dice
Cube Handling
Cube Handling in Races
Extreme Gammon
Fun and frustration
GNU Backgammon
Luck versus Skill
Magazines & E-zines
Match Archives
Match Equities
Match Play
Match Play at 2-away/2-away
Opening Rolls
Pip Counting
Play Sites
Probability and Statistics
Source Code
Strategy--Bearing Off
Strategy--Checker play


Return to:  Backgammon Galore : Forum Archive Main Page