Difficulties with "Advanced Backgammon"
Scott Steiner
I recently purchased a copy of Robertie's Advanced Backgammon (both
volumes), I'm finding the books rather difficult though. I've started
playing backgammon beginning of this year and have read Paul Magriel's
Backgammon several times (3-4 times) and have understood the concepts
and noticed that my game improved much after Magriel. But Robertie's
Advanced Backgammon seem to be too difficult for me at the moment so I'm
wondering what people here would recommend me to do. Are there good
books out there targeting an audience whose ability is between Magriel's
Backgammon and Robertie's Adv. Bg.? Another thing I noticed which is
discourging me a little is that Robertie's analysis is quite deep and
extensive requiring much time to think about which isn't the discourging
part but rather that it is hard for me to imagine that I have to come up
with all this analysis in the few seconds I have to make my move! It
looks like it takes 15-20 minutes to solve some of Robertie's problems
but in FIBS people start swearing at you if you take 20 seconds ;-) so
I'm kind of discouraged! So what would you people advise me to do?
Stop being a wimp and sit down and try to absorb Robertie's Adv. Bg. or
maybe look for some other book in the meanwhile? Don't get me wrong, I
think Adv. Bg. is very good, it's just targeted for the more
advanced-expert player IMHO.
thx for any help
Gregg Cattanach
that is its best quality. You can just skip over some of it, and realize
that a lot of what he describes is not necessarily what an expert player
actually does over-the-board. Especially all of the estimates of number of
wins/losses/gammon in 36 games. Some may attempt this much detail, but in
your head you can only do so much.
Keep chugging through these two books, they really do show you the things
you should try to consider while analyzing positions. Lots of book just
give the 'answer' and a flip reason why, but especially in this book, you
can see how he got to the conclusion. Also beware, because perhaps 5% or so
of the positions give the 'wrong' answer in light of what we know now with
the neural nets.
For the intermediate, I'd also recommend Robertie's 501 Positions. Lots
more positions, much less detailed but still interesting analysis, and nice
logical groupings of the positions. And it is formatted like a quiz where
you can guess first and then go look up the answer.
Gregg C.
"Scott Steiner" <nos...@nospam.nospam> wrote in message
news:3D949E53...@nospam.nospam...
Scott Steiner
>
> Robertie's analysis IS extremely detailed throughout 'Adv. Backgammon', but
> that is its best quality. You can just skip over some of it, and realize
> that a lot of what he describes is not necessarily what an expert player
> actually does over-the-board. Especially all of the estimates of number of
> wins/losses/gammon in 36 games. Some may attempt this much detail, but in
> your head you can only do so much.
>
> Keep chugging through these two books, they really do show you the things
> you should try to consider while analyzing positions. Lots of book just
> give the 'answer' and a flip reason why, but especially in this book, you
> can see how he got to the conclusion. Also beware, because perhaps 5% or so
> of the positions give the 'wrong' answer in light of what we know now with
> the neural nets.
I suspected that some answers aren't "right" because I tested some
positions that seemed suspicious to me with gnubg at world class++ and
received different answers. Not that I would trust gnubg blindly but I
wouldn't trust Robertie blindly either ;-). I usually just check a
position with gnubg and see if the answers vary, if they do then I
choose the play that I'm comfortable with more considering my level
play, if the answers don't vary then I take it as God's word ;-)
--
Email replies to following address: bpp at chello dot at
Bob Stringer
<nos...@nospam.nospam> wrote:
If you can find it, I recommend Joe Dwek's "Backgammon for
Profit." It's laid out in somewhat the same style (problem
per page, with explanatory material below each diagram), but
the problems are somewhat easier to comprehend right away.
The positions are useful, and it's much shorter than
Robertie -- 120 problems. That means you can get through it
that much quicker and then go through it again a couple
times to absorb the lessons.
It's dated (1975), but the point is to get you thinking
about the positions, and it does a good job of that.
After that you ought to be in shape to get a handle on
Robertie.
Bob Stringer
--
In order to reply by e-mail, please replace
"1nospam" with "0" (a zero) in my address
Brad Davis
> Robertie's analysis IS extremely detailed throughout 'Adv. Backgammon', but
> that is its best quality. You can just skip over some of it, and realize
> that a lot of what he describes is not necessarily what an expert player
> actually does over-the-board. Especially all of the estimates of number of
> wins/losses/gammon in 36 games. Some may attempt this much detail, but in
> your head you can only do so much.
>
> Keep chugging through these two books, they really do show you the things
> you should try to consider while analyzing positions. Lots of book just
> give the 'answer' and a flip reason why, but especially in this book, you
> can see how he got to the conclusion. Also beware, because perhaps 5% or so
> of the positions give the 'wrong' answer in light of what we know now with
> the neural nets.
you about
25% of these positions have errors. Some errors being very large.
there's no point in telling a newcomer
" Beware 5% of the positions are wrong" if he hasn't got a clue which
ones.
If fact I will go as far as saying people (without the rollouts)
should NOT read these volumes as they contain too many errors.
After reading Magriel's backgammon go on to one of Robertie's newer
publications. I also recommend Robertie's 501.
My other advice would be to pay $36 and join GammonLine
The discussion group there is tremendous and many of the world's top
players
will offer their analysis and advice to other top players OR
newcomers.
I have NO financial interest in GammonLine by the way.
Frank Berger
> On Fri, 27 Sep 2002 18:06:40 GMT, Scott Steiner
> <nos...@nospam.nospam> wrote:
>
> If you can find it, I recommend Joe Dwek's "Backgammon for
> Profit." It's laid out in somewhat the same style (problem
> per page, with explanatory material below each diagram), but
> the problems are somewhat easier to comprehend right away.
> The positions are useful, and it's much shorter than
> Robertie -- 120 problems. That means you can get through it
> that much quicker and then go through it again a couple
> times to absorb the lessons.
>
> It's dated (1975), but the point is to get you thinking
> about the positions, and it does a good job of that.
>
> After that you ought to be in shape to get a handle on
> Robertie.
>
> Bob Stringer
On ebay, you can get sometimes a used copy. You can have also
"Backgammon for Profit" as a positiondatabase for BGBlitz. It contains
positions + rollouts.
You can get BGBlitz at http://www.bgblitz.com/download.html
From the books of the 70's area it probably the best, by todays
standards.
ciao
Frank
Ric
"Brad Davis" <backg...@email.com> wrote in message
news:bcf2ceef.02100...@posting.google.com...
> I have rolled out all these positions with Snowie 3.2 and I can tell
> you about
> 25% of these positions have errors. Some errors being very large.
> there's no point in telling a newcomer
> " Beware 5% of the positions are wrong" if he hasn't got a clue which
> ones.
> If fact I will go as far as saying people (without the rollouts)
> should NOT read these volumes as they contain too many errors.
Saying you rolled them out with 3.2 isn't useful information. What settings?
How many games? Search space size? Live cube? Cubeless? Truncated? Etcetera.
In order to roll out all the positions in the two volumes of Advanced
Backgammon using useful settings would take a hell of a long time, even on a
fast computer. So if you'll provide the setting information, then people on
rgb can decide if your comment is valid.
Ric
amni
Because when you study a book like Roberti's you
learn IDEAS which you can apply (at least some of them) on the board.
You are not expected to play against a strong opponent like
Snowie. You play against human players, who plan their
moves using IDEAS rather than elaborate calculations.
amni
backg...@email.com (Brad Davis) wrote in message news:<bcf2ceef.02100...@posting.google.com>...
Douglas Zare
Brad Davis wrote:
> "Gregg Cattanach" <gcattana...@prodigy.net> wrote in message news:<vA1l9.201$5L3.13...@newssvr15.news.prodigy.com>...
> > [a lot of stuff with which I agree, but will cut]
>
> I have rolled out all these positions with Snowie 3.2 and I can tell
> you about>
> 25% of these positions have errors. Some errors being very large.
> there's no point in telling a newcomer
> " Beware 5% of the positions are wrong" if he hasn't got a clue which
> ones.
No, we get much of our information from imperfect sources, and it is valuable both to have such sources and to have indications
of their errors. I wouldn't stop recommending Magriel's _Backgammon_ because of the bad advice about action plays and the opening
5-3. There is too much good stuff, such as the emphasis on the golden point and the unrivaled discussion of duplication. I just
warn my students not to take it as gospel.
> If fact I will go as far as saying people (without the rollouts)
> should NOT read these volumes as they contain too many errors.
I think there are far too many good ideas to skip these books. Incidentally, there are many more than 400 positions discussed, as
many of the problems have multiple variations, and Robertie makes a lot of side comments about nondiagrammed changes in the
position that would affect decisions.
I've often disagreed with the analysis of particular positions. If it is by a lot, then I follow Robertie's other advice, and
play the position out by hand (or more commonly now, with Jellyfish interactive rollouts). Sometimes I get to adjust my
intuition. Sometimes Robertie is dead wrong, but then I understand a common misunderstanding that my opponents will have.
On Gammonline, in response to a position I posted from the New England Backgammon Club, Neil Kazaross pointed out that one can
get some mileage out of errors with Robertie's Rule of 5, for example. Then he demonstrated it, getting a bad pass in a similar
position in a consolation match in the GammonZone Gold Medallion tournament. Robertie's Rule of 5 is a great heuristic, and I
will play better for having seen it rather than just studying correct reference positions.
> > "Scott Steiner" <nos...@nospam.nospam> wrote in message
> > news:3D949E53...@nospam.nospam...
> > > [...] Another thing I noticed which is
> > > discourging me a little is that Robertie's analysis is quite deep and
> > > extensive requiring much time to think about which isn't the discourging
> > > part but rather that it is hard for me to imagine that I have to come up
> > > with all this analysis in the few seconds I have to make my move! It
> > > looks like it takes 15-20 minutes to solve some of Robertie's problems
> > > but in FIBS people start swearing at you if you take 20 seconds ;-) so
> > > I'm kind of discouraged! So what would you people advise me to do?
By the way, instead of trying to make the best play, take a good shot at it, and then analyze later. You aren't supposed to be
able to perform a complicated analysis over the board in most positions.
Also, you don't have to understand the book in one reading. I don't think that is a good thing to try with Magriel's
_Backgammon_, either.
Douglas Zare
Scott Steiner
>
> "Gregg Cattanach" <gcattana...@prodigy.net> wrote in message news:<vA1l9.201$5L3.13...@newssvr15.news.prodigy.com>...
> > Robertie's analysis IS extremely detailed throughout 'Adv. Backgammon', but
> > that is its best quality. You can just skip over some of it, and realize
> > that a lot of what he describes is not necessarily what an expert player
> > actually does over-the-board. Especially all of the estimates of number of
> > wins/losses/gammon in 36 games. Some may attempt this much detail, but in
> > your head you can only do so much.
> >
> > Keep chugging through these two books, they really do show you the things
> > you should try to consider while analyzing positions. Lots of book just
> > give the 'answer' and a flip reason why, but especially in this book, you
> > can see how he got to the conclusion. Also beware, because perhaps 5% or so
> > of the positions give the 'wrong' answer in light of what we know now with
> > the neural nets.
> I have rolled out all these positions with Snowie 3.2 and I can tell
> you about
> 25% of these positions have errors. Some errors being very large.
> there's no point in telling a newcomer
> " Beware 5% of the positions are wrong" if he hasn't got a clue which
> ones.
> If fact I will go as far as saying people (without the rollouts)
> should NOT read these volumes as they contain too many errors.
> After reading Magriel's backgammon go on to one of Robertie's newer
> publications. I also recommend Robertie's 501.
Can you name Robertie's newer publications that you mean? Is Robertie's
501 a newer publication too? Doesn't 501 have a good amount of errors
too?
thx
Gregg Cattanach
"Scott Steiner" <nos...@nospam.nospam> wrote in message
> Brad Davis wrote:
> Can you name Robertie's newer publications that you mean? Is Robertie's
> 501 a newer publication too? Doesn't 501 have a good amount of errors
> too?
>
> thx
His newest book "Modern Backgammon" was released in the last couple months.
Personally, I think it's the best thing he's written. He's created 4 guiding
principles for the game, and the illustrated positions and the fully
analyzed match are really well chosen. And yes, "501 Positions" has some
errors, but not a lot of big ones. It's still very informative and well
laid out, (but I don't like those oval checkers :)
Gregg C.
Scott Steiner
Is it true that 501 was released 1999 or 2000? If this is true then
that is fairly new, so I guess the number and magnitude of errors
shouldn't be so dramatic by modern standards like you said.
And yes, I've heard about the tiny book format with the squashed oval
checkers ;-). What the heck, as long as it improves my game I'll even
take flat checkers ;-)
Brad Davis
I used a pruning approach.
I rolled them all out 2 ply Huge 720 games truncated at 8 to start
with.
Then I Re-Ran those that weren't conclusive or invloved more difficult
plays
3 Ply Full 100% UNtruncated 720 games. All cubeless
And yes it did take a hell of a long time. There are actually 456
positions.
In case you're wondering if I'm suitably qualified to decide which
positions to Re-Run, my online ELO fluctuates around 1950.
Whilst my rollouts may not be 100% conclusive, the number of errors is
certainly nearer 25% than 5% which in my opinion is too many.
Let's face facts Backgammon has moved on quite a way since these books
were written.
>
> Ric
Ric
unless you've got a Cray in your study! ;-)
A question that does occur to me: did you allow for positions that Snowie is
considered not to be good at? I don't know how many of those there might be
in the books or how that might affect your numbers.
Be interesting to hear comments here from others on your work.
Ric
Brad Davis
> Sounds like you did a reasonable study. You must be about 97 years old now,
> unless you've got a Cray in your study! ;-)
I had home,work and laptop running for months!
>
> A question that does occur to me: did you allow for positions that Snowie is
> considered not to be good at? I don't know how many of those there might be
> in the books or how that might affect your numbers.
Yes I did. As an obvious example I ingnored the Kauder Paradox
position.
In the deep back game positions I gave Robertie the benefit of the
doubt
if Snowie rollouts indicated he's erred slightly. Although Snowie
doesn't
evaluate back games well at all at least she plays them pretty well.
Interpreted cubeless numbers are a better guide as Snowie's cube
handling is way off in those positions.
I also allowed for positions where recirculating and capturing
additional checkers was involved.
My Snowie 4 is due any day, I may rerun these type of positions but I
doubt
they'll make much difference to my statement that around 25% are
wrong.
I did think about releasing my data but I know Robertie is a bit
touchy
about these matters. Especially as the old matra still seems to be
"Read Magriel's Backgammom and then Robertie's Advanced backgammon"
Steve Bortnyck
> touchy
> about these matters. Especially as the old matra still seems to be
> "Read Magriel's Backgammom and then Robertie's Advanced backgammon"
Seems a shame doesn't it? Robertie is good enough to know the game has
changed, and what used to be "right" isn't anymore. And what is now, may
not be later.
I would think as a writer I would be flattered by the attention, no one
bothers to look at Becker's book :)
Steve Bortnyck