? "100 Backgammon Puzzles" by Paul Lamford ?

[Harold Simon]

Hi, all

I can buy this book from a reseller for $9.95 (publication date not specified,
original price $12.95). The price is not important but its instructive value
is to me. Entertainment is nice, but a chance to learn is better.

Evaluations from anyone who has read it will be appreciated.

TIA

Hal

[Joxer]

On Fri, 19 Oct 2001 04:30:56 GMT, hals...@iccas.com (Harold Simon)
sallied forth:

It helped me. Can't really say much more than that, but for such a small
sum (though I think I paid even less) there's little reason not to buy it
- all the problems were apparently checked by either Snowie and/or the
Fish for accuracy, though the relevant analyses are understandably not
included. Apart from anything else, it showed me to be a pretty good
positional player but a lousy cube judger - I already knew this, but did
not realise just how lousy! Now all I need is the time and determination
to improve that aspect...
--
Colin B.

Grrr. Arrgh.

[Douglas Zare]

Harold Simon wrote:

This book is fun. There are 100 positions, each followed by a short answer and a
quick tip. The puzzles are challenging, sometimes too challenging. That's because
some of them are slightly wrong, despite being checked by Snowie 2 and Jellyfish.

I think the book is worthwhile for a strong player who will not be misled by the
overly simplistic tips, some of which are quite wrong. For example, following an
example in which a loose hit was wrong, it advises not to hit if your opponent has
the stronger board. That's wrong. Not every hit is a loose hit, and hitting
doesn't necessarily lead to a hitting battle, or the worst of one. Hitting can be
a great way to get back into the game, and may be the best way to slot a point.

Douglas Zare

[Mark Driver]

As others have mentioned its a great little book. Its probably the most read
book I own for the simple reason that it fits nicely into a jacket pocket to
read on the bus or train etc.

50 puzzles are on checker play, the other 50 on cube action.

Each puzzle is devoted to one page with the answer at the bottom of the
following page (not the best format IMO I would prefer to have all the
answers at the end of each section).

The author's analyses are quite brief, (dont expect anything akin to 'News
Ideas in Backgammon' for instance), but instructive nevertheless. Each
problem is accompanied by 'handy hints and tips' section which summarize the
salient points in memorable soundbites. The quiz format allows the reader to
estimate their ELO rating from a table at the back of the book. The
introductory section also contains Paul Lamford's PRAT (Position Race And
Threats) mnemonic for cube decisions which is simple but effective. Check
out the MSO website where Allan Webb wrote an article featuring the PRAT
concept.

All in all a very good book for the price. Btw I noticed a UK book shop
knocking out the original edition copies at two pounds-fifty, (Sterling) a
while back, so you might find it cheaper via bookfinder.com or Abebooks etc.

Harold Simon wrote in message <3bcfab7...@news.cis.dfn.de>...

[Tore Fredriksen]

The cube positions in this book is definitely much more difiicult than
the checkerplay posistions

[Michael Schell]

Here is a list of problems in "100 Backgammon Puzzles" where Snowie rollouts
produce results that are at variance with Lamford's suggested plays. Many of
the differences are minor, with a second candidate play performing about the
same as the preferred play. Some differences are significant (e.g., #41).
Many are minor. In at least one case (#100), Snowie's rollout results are
not dependable. I have not included problems such as #55 where Snowie's
3-ply evaluation disagrees with Lamford but where rollouts favor his choice
over Snowie's.

Checker play problems:

#17
7/3 5/4 and Lamford's 5/1 4/3 come out tied at depth 7.

#24
8/4* and Lamford's 7/4*/3 come out tied at depth 7. 8/4* wins more games but
gives up more gammons.

#41
23/13 clearly prevails over 23/17 5/1 in full rollouts. The latter rates as
a blunder in fact.

#46
Bar/23 5/4 and Bar/23 3/2 come out tied at depth 7.

Cube decisions:

#64
Lamford says no redouble/take. Full rollouts split on whether this is a
redouble or not, with the cube-adjusted numbers saying no redouble/take and
the live cube numbers saying double/take. It's extremely close in either
case. White is 67.1% (16.2%/5.3%) cubeless, putting this on the margin.

#67
In this race position, White leads 100-112, making it double/take. Full
rollouts concur. Lamford says "If the lead was just one pip more, then Black
should pass; two pips fewer and White should wait before doubling." Here,
the rollouts disagree, giving White a legitimate double at 102-112 and
103-112. 104-112 is not quite a double. Snowie agrees, though, that it's a
pass at 99-112.

#76
Lamford says no double/take. Full rollouts make this a very narrow double
and an easy take.

#77
Lamford says no double. Full rollouts put this right at the threshold, with
White at about 66%.

#78
Lamford says "This is a correct, but only marginal double and easy take".
Full rollouts have this as a clear double and take. Not to double would be a
blunder.

#79
Lamford says it is premature to redouble. Snowie figures you lose your
market on 1-1, 2-1 or 4-1, or a 3-1 or 6-1 followed by a fan, or a 4-2 then
a 1 picking up the second blot. Ergo, double/take even though you're only
58% to in. Full rollouts concur.

#86
Lamford says play on for a gammon. Full rollouts say cash now.

#92
Lamford says no double/take. Full rollouts split, with cube-adjusted numbers
saying double/take and the live cube numbers saying no double/take. Another
borderline case, evidently.

#97
Lamford says play on for a gammon. Full rollouts (not score-based) split,
with the cube-adjusted numbers saying cash, and the live cube numbers saying
play on.

#98
Lamford says redouble/take. Full rollouts (both score-based and non
score-based) say that redoubling is a moderate error. White is about 27% to
win here.

#99
Lamford says redouble/take. Snowie 3-play says redouble/pass, and this is
confirmed by rullouts with the bearoff database turned off. White is 66.9%
to win, and Black will not have a recube opportunity, so his take point is
34.0% according to Snowie. Perhaps with a different MET this would be a
take.

#100
In this trick position (all 30 checkers in White's home board), Lamford says
no double/take. Snowie full rollouts say White is 75.8% to win (with 41.0 %
backgammons and 61.7% gammons), but playing out this position manually
against Snowie demonstrates that the bot doesn't understand how to play the
Black side here. I would trust Lamford over Snowie in this case.

- Michael Schell
www.cribbageforum.com