Is there a painless way to count pips?
saratw2
computer gives you the pip count for both sides. It's right in front of you.
BUT - in real life, sitting across from a person you can hear breathing, the
pip count is not visible. The count is not in my head, but it is always in
my opponent's head.
I don't want to show up with a calculator. Is there an intuitive way to know
about how you stand pip wise? I know it is vital in doubling, but having to
count and add (ugh) just takes all the fun out of the game for me.
With the computer you can spend all your efforts on deciding the best move.
Sara
Bob Stringer
<sar...@cox.net> wrote:
It's never painless for most people, but here's a relatively
easy method:
http://www46.pair.com/sengoku/Pip/English/FiveCount.html
Most positions don't require pip counting. Once you get to
a position that requires it, and it looks like the count
will continue to be important, it's a good idea to remember
the difference between the two sides and adjust it by
subsequent rolls. That won't give you the totals, but at
least you have a semi-decent idea of how things are
progressing.
Bob Stringer
Chase
There isn't an "intuitive" way to count pips, but with a little
practice, you can normally get a feel for whether you are ahead or
behind in the race. (Often, just looking at who has the most checkers
back in their opponent's home board will tell you this). Most of the
time, that's all you need to know. I practice this by turning the pip
count feature on my software off. Without doing any actual counting, I
make my best guess at who is leading and by how many pips. Then I
check it with the software. I would recommend starting with Jellyfish
or while watching others play, so you don't take concentration from
your play while you are learning. You'll be terrible at first, but
you'll be amazed at how quickly you get good at it.
For those few occasions where you need an accurate count, there are a
number of pip counting methods that can make it a lot quicker and
easier. Someone just asked about this here in this forum a few weeks
ago and got quite a number of answers. You might try a Google search
to get at those.
Everyone is different, but the method I use is called Cluster
Counting, and it was developed by Jack Kissane. He wrote an article
about it that can be found at
http://www.northcoast.com/~mccool/bg.html#ARTICLES
The idea behind this method is to memorize a handful of checker
formations which seem to come up constantly in play, so that we know
their totals on sight. Then it's just a matter of identifying a few of
these "clusters" and adding their totals. The article gives you five
basic formations, and examples on how to put them to work. Once you
get comfortable with those, you can add your own. For example, one of
the ones I use is that two checkers on the midpoint and one checker on
the 24 point total 50. This is a formation that comes up often, is
easy to spot, and gives a nice round number that is easy to add.
After a few hours of practice with this, I got where I can usually
count one side within about ten seconds. I don't play live much, but
when I do, I set aside ten or fifteen minutes to brush up on my
counting before I go.
Hope this helps.
Chase
_____________________________________________________________________
To respond via email, replace "USERNAME" with "demiga" in my address.
Chase
wrote:
|On Sat, 16 Mar 2002 12:38:51 GMT, "saratw2"
|<sar...@cox.net> wrote:
|
|
|It's never painless for most people, but here's a relatively
|easy method:
|
|http://www46.pair.com/sengoku/Pip/English/FiveCount.html
I wasn't familiar with this method, so I popped over to the page to
have a look. While this sounds easier than doing an actual count, it's
like a lot of other methods that seem to require a lot of steps. Even
if they are easy steps, each step is a chance to make an error. It may
just be the way I think (I am mathematically challenged :), or that
I'm so comfortable with the Cluster Counting approach, but it just
seems so much simpler. Here's the example given on the web page:
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| X X | | X |
| X | | X |
| | | | S
| | | | n
| | | | o
| |BAR| | w
| | | | i
| X | | | e
| X | | |
| X X | | X |
| X X X | | X X |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+
The five count method has you:
- Count the number of checkers in each of the 6 designated groups
- Multiply the total in each group by the group number
- Add those totals together
- Multiply this by 10
- Divide this by 2
- Make adjustments to this by performing a series of mental shifts
None of this is difficult. But it does take time, and there are a lot
of opportunities for error.
If I were counting this using Cluster Counting, here's my thinking
process:
- 10 men symmetrically placed on the 6 point: 60
- 2 men on the 20 point. That's 40 more: 100
- 2 men on the midpoint and 1 man on the 24 point: That's 50 more:
150
- The back checker is really one pip closer to home, so the pip count
is 149.
I realize everyone is different, but I find this a much simpler and
easier approach, which allows me to spend more of my (limited) mental
energy figuring out to do with my checkers and the cube.
Bob Stringer
(Chase) wrote:
>I realize everyone is different, but I find this a much simpler and
>easier approach, which allows me to spend more of my (limited) mental
>energy figuring out to do with my checkers and the cube.
I was aware of cluster counting, but hadn't liked it.
However, I looked at it when I was starting to play, about 3
years ago. I came across Sho's method just recently and am
very comfortable with it. Maybe I'd like cluster counting
now that I'm more comfortable with the game in general.
I'll give it another try and see.
Thanks.
Bob Stringer
saratw2
I'll start with is the one by Michael Crane at
http://www.msoworld.com/mindzine/news/classic/bg/pipcount1.html
Also I like your idea of practicing with the pip count turned off and seeing
how far off my guess - er, count is.
Still am not up to any big mathematical calculations, but will try the
above.
Thanks.
A_C
http://www.math.columbia.edu/~zare/bgarticles.html
I think you can get the artikel if you talk to Douglas Zare.
/Andreas A_C on FIBS
Sho Sengoku
> The five count method has you:
>
> - Count the number of checkers in each of the 6 designated groups
> - Multiply the total in each group by the group number
> - Add those totals together
The first section of that page, "Introduction to 'Five-Count'" is
mainly to explain just basic of the system, how the system can
provide accurate (not approximate) pip count. It isn't intended
to show efficient ways to apply the system. I made the section 2,
3, and 4 for that purpose.
This is how I would count the position I used in the first
section to introduce my counting method:
http://www46.pair.com/sengoku/Pip/English/A/000.html
> - Multiply this by 10
> - Divide this by 2
I usually don't have to do this math because I already memorized
the result of most numbers. There are not many to remenber:
http://www46.pair.com/sengoku/Pip/English/EF_2.html
> - Make adjustments to this by performing a series of mental shifts
Not true. My method doesn't require any mental shifts.
In an adjustment phase, you don't have to mentaly move any
checkers. Please check these pages:
http://www46.pair.com/sengoku/Pip/English/Section_2.html
http://www46.pair.com/sengoku/Pip/English/EF_3.html
----[About Mirrors and other patterns]----
In a straight count method including the Cluster Count, the
"mirrors" (2checkers at one point and 2 checkers at opposing
point) always make 50 pips.
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| | | X |
| | | X |
| | | | Pip = 50
| | | |
| | | |
| | | X |
| | | X |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+
But those checkers must be at exactly mirrored position, or
you have to do some mental shifting.
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| | | X |
| | | X |
| | | | Pip = 49
| | | | (50 +1 -2)
| | | <---- --> |
| | | -2 +1 |
| | | X X |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+
"Mirrors" also make my "five count" system easier in a
group counting phase, because they always make 10 as
a group count. Big difference from straight count methods
is that those four checkers don't have to be at exact
mirrored positions. As long as two are in a group
and another two are in a opposing group, they make
10. No mental shift is required.
G.5 : Group 4 : Group 3
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| : | | : X |
| : | | : X |
| : | | : | Group
| : | | : | count = 10
| : | | : | 3x2 + 2x2
| : | | : |
| : | | : X X |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+
G.0 : Group 1 : Group 2
G.5 : Group 4 : Group 3
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| : X | | : |
| : X | | : |
| : | | : | Group
| : | | : | count = 10
| : | | : | 4x2 + 1x2
| : X | | : |
| : X | | : |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+
G.0 : Group 1 : Group 2
If you don't have two checkers in a group, you can
"borrow" checkers from lower group to make that
"mirror" pattern, for example:
G.5 : Group 4 : Group 3
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| : | | : X |
| : | | : X |
| : | | : | Group
| : | | : | count = 10
| : | | : |
| : | | X: |
| : | | X: X |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+
G.0 : Group 1 : Group 2
During a group counting process in my "five count"
I always look for 10s (as I stressed out in my Web pages.)
Here's some common patterns that make 10s (besides
"mirrored" positions):
G.5 : Group 4 : Group 3
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| : X | | : X |
| : | | : X |
| : | | : | Group
| : | | : | count = 10
| : | | : | (6+4)
| : | | : |
| : | | : |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+
G.0 : Group 1 : Group 2
G.5 : Group 4 : Group 3
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| : | | : X |
| : | | : X |
| : | | : X | Group
| : | | : | count = 10
| : | | : | (9 + 1)
| : | | : |
| : X | | : |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+
G.0 : Group 1 : Group 2
G.5 : Group 4 : Group 3
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| : X | | : X |
| : | | : |
| : | | : | Group
| : | | : | count = 10
| : | | : | (5+5)
| : | | : |
| : X | | : X |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+
G.0 : Group 1 : Group 2
G.5 : Group 4 : Group 3
+24-23-22-21-20-19-------18-17-16-15-14-13-+
| X : | | : X |
| : | | : X |
| : | | : X | Group
| : | | : X | count = 20
| : | | : X |
| : | | : |
| : | | : |
+-1--2--3--4--5--6--------7--8--9-10-11-12-+
G.0 : Group 1 : Group 2
Sho
Chase
impression I get from reading about it on your page is that there are
quite a number of (admittedly easy) steps involved in getting a final
count. I've noted your comments below, and will certainly take
another, closer look at your system.
Chase
Win At Betting
Generally, if you have to count pips, the count doesn't matter much.
FREE!! Learn TOTAL SEXUAL CONTROL of women!
Ron & Ann Barry
If you ever do get to the point of actually counting pips, by far the most
elegant, easily remembered technique, and the one that requires the fewest
number of steps, is that which Doug Zare developed some time back. I don't have
the Web page handy just now, but I'm sure Doug can provide it for you.
Regards, Ron Barry.
Win At Betting
I've seen it run its pipcount into the 200s before kicking my butt.
Positional play is everything.
Want to get laid by beautiful women?
Douglas Zare
Ron & Ann Barry wrote:
> Sara,
>
> If you ever do get to the point of actually counting pips, by far the most
> elegant, easily remembered technique, and the one that requires the fewest
> number of steps, is that which Doug Zare developed some time back. I don't have
> the Web page handy just now, but I'm sure Doug can provide it for you.
Thanks. The URL is
http://www.gammonvillage.com/backgammon/news/article_display.cfm?resourceid=282 .
Recently GammonVillage switched to subscriptions. I believe that there will be more
free samples in the future, the way that some of Walter Trice's columns are now
available to everyone. This article was reprinted in a BIBA publication, too.
The basic idea is to count half-crossovers needed to bear your checkers in. Every
checker on your 456 counts 0 (which is good, since you have a lot there), every
checker on your 789 counts 1, every checker on you 101112 counts 2, ..., every
checker on your 192021 counts 5. The error from assuming that every checker on the
123 is on the 2, every checker on the 456 is on the 5, etc. is remarkably low,
particularly in positions which arise from normal play. Normally, when I play OTB
or watch matches on GamesGrid as I exercise (unable to click the calculator
button), I only perform the count of half-crossovers, and don't bother to make the
adjustments of -1, 0, or +1 for each checker to get an exact count except in racing
doubling decisions. To get an exact count, you multiply the HC's to bear in by 3
and add to 75, then for each checker on the high point of its triplet (6 of 456)
add 1, and for each checker on the low point, subtract 1.
As Sara mentioned, the pip count is extremely important in modern backgammon, and
anyone who does not have an accurate feel for the race will be eaten alive by any
strong player who does. Which system you use is not very important. What matters is
that you choose a system that you will use when you need to.
Douglas Zare
Win At Betting
A minor advantage in pips alone is not sufficient to justify doubling.
If you can't see a clear edge in pips, you don't need to worry about the count,
and if you can see one, you don't need to count it to offer the cube (assuming
you have no positional weaknesses).
Pip-counting is for amateurs.
Douglas Zare
Win At Betting wrote:
> [...]
> >As Sara mentioned, the pip count is extremely important in modern backgammon,
> >and
> >anyone who does not have an accurate feel for the race will be eaten alive by
> >any
> >strong player who does. Which system you use is not very important. What
> >matters is
> >that you choose a system that you will use when you need to.
>
> A minor advantage in pips alone is not sufficient to justify doubling.
This doesn't contradict what I said, or what any decent backgammon player believes.
Even tremendous advantages in the race do not by themselves warrant doubling, e.g.,
against a well-timed backgame, or when one of your checkers is primed.
You claim to understand something of chess, although with your apparent
overconfidence in backgammon... At any rate, the race is slightly less important in
backgammon than the material advantage in chess. Both are indicators of how the
endgame should go if nothing special happens. Thus, the pip count is extremely
important in many positions, e.g., holding games. It's important even in positions
in which you expect some hitting to occur, e.g., it is one of the safe-play versus
bold play criteria.
> If you can't see a clear edge in pips, you don't need to worry about the count,
> and if you can see one, you don't need to count it to offer the cube (assuming
> you have no positional weaknesses).
False.
> Pip-counting is for amateurs.
Then I'd be honored to play a "professional" player such as yourself for money.
Douglas Zare
Southpaw
>
>
> Want to get laid by beautiful women?
Yep, now that seems like a lot more fun...