Monty vs Mpetch Results - Updated April 12th, 2008
Michael Petch
I will be posting the matches as they are completed at this site:
http://www.capp-sysware.com/analysis
I will save the chat logs, the Jellyfish .MAT file, and the gnu analysis in
this format:
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match#.html
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match#-chat.rtf
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match#.mat
First match results are now posted:
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match1.html
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match1-chat.rtf
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match1.mat
Matches are analyzed at Grandmaster setting for Checker play, and Supremo
for Cube play (2ply cube) using the G11 MET.
Michael Petch
Monty was a good sport and gentleman for the first match, and was actually a
pleasure to play him.
Grunty
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match1.mat
Thanks Michael.
Just a minor point: in the .mat file, last move in game#4, the third
1/0 is missing. You may want to append it so the import won't flag it
as an illegal play.
mpetch
Interesting. You are correct! Gnu didn't complain about it. I have
amended it.I don't think it will have any impact on the analysis.
given it was last roll and no change to th outcome. I assume Snowie
was the one that caught it?
Grunty
Yes, Snowie (old version). No impact though.
Don't know whether Sw4 flags it or ignores it.
Back to Gammon
Michael Petch
Third match results are now posted:
These results are 2ply Checker / 2ply Cube (Given the discussion I will post
the results with those sends - using HUGE move filter) and G11 Met:
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match3.html
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match3-chat.rtf
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match3.mat
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match3.sgf
These results are 3ply Checker / 2ply Cube (using HUGE move filter) and G11
Met:
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match3-2cu3ch.html
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match3-2cu3ch.sgf
Michael Petch
On 4/26/08 3:15 PM, in article C438F78C.57B37%mpe...@capp-sysware.com,
"Michael Petch" <mpe...@capp-sysware.com> wrote:
> results with those sends
Should have read:
results with those settings
Michael Petch
On 4/26/08 3:15 PM, in article C438F78C.57B37%mpe...@capp-sysware.com,
"Michael Petch" <mpe...@capp-sysware.com> wrote:
>
> Third match results are now posted:
>
As a followup - the match wasn't anything spectacular except that the
relative luck between us was slightly favored for Monty but actually very
very close.
During the match the only place where we both commented and really disagreed
was the Cube (By Monty) / Drop (mpetch) in Game 13 Move 16. Monty felt it
was a take and was curious what GnuBG would say. My comment on the play
was:
[1:45 PM] Monty> I thought that was a take
[1:46 PM] mpetch> naw, no guarntee for timing.
[1:46 PM] mpetch> I'd rather make it 4 away 2 away
[1:46 PM] mpetch> which is a far more itnerestign score for the person
trailing to play :)
Monty of course won, and congratulations to him. Last match was a toss up
with equity swinging in favor of both of us many times (Was 13-13 going into
last game).
Thanks for those who came out and watched.
Michael
Raccoon
> "Michael Petch" <mpe...@capp-sysware.com> wrote:
>
> > Third match results are now posted:
>
> As a followup - the match wasn't anything spectacular except that the
> relative luck between us was slightly favored for Monty but actually very
> very close.
Not much excitement, with 9 of 15 games (in 3 matches, 23 of 43 games)
ending with an initial Double and Pass.
> During the match the only place where we both commented and really disagreed
> was the Cube (By Monty) / Drop (mpetch) in Game 13 Move 16. Monty felt it
> was a take and was curious what GnuBG would say.
That would be this position:
GNU Backgammon
Position ID: bC5iAFbYncEBKA
Match ID : cAngAbAAYAAA
+24-23-22-21-20-19------18-17-16-15-14-13-+ O: mpetch
| X X O O O | O | O O X | 11 points
| O O O | | X |
| O | | X |
| | | |
| | | |
| |BAR| |v 15 point match (Cube: 1)
| | | |
| | | |
| X X | | |
| O X X X | | X O | On roll
| O O X X X | | X O | 12 points
+-1--2--3--4--5--6-------7--8--9-10-11-12-+ X: Monty
X on roll leads 3-away 4-away. O has 3 checkers back but only one
anchor, 1 more checker on the bar, 2 more blots in the outfield, and
does not and will not own his 5 point and bar point. X has a 3 point
board with builders on the 6 and 5 points, and 24/36 numbers that put
a second checker on the bar. O's best hope is a series of miracles;
second best hope, perhaps a badly timed ace-deuce game, or just a
deuce-point game. Rarely a 2-3 game. Most of O's losses will be
gammons.
If X doubles, O can pass to trail 4-away 2-away with 33% match equity,
or take and (although not necessarily immediately) redouble and play
for the match. Since O doesn't seem to be winning this game nearly 1/3
of the time, if X doubles O should pass.
Should X double?
Gnubg's comment is: Play on, X! Gammon the petch!
> Monty of course won, and congratulations to him. Last match was a toss up
> with equity swinging in favor of both of us many times (Was 13-13 going into
> last game).
mpetch Monty
MATCH 1-3 Snowie equiv. error rate -6.3 -8.1
MATCH 1-3 Play errors rating loss 392.8 346.6
MATCH 1-3 Cube errors rating loss 26.6 100.3
Michael Petch
Fourth match results are now posted:
These results are 2ply Checker / 2ply Cube (Given the discussion I will post
the results with those sends - using HUGE move filter) and G11 Met:
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match4.html
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match4-chat.rtf
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match4.mat
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match4.sgf
These results are 3ply Checker / 2ply Cube (using HUGE move filter) and G11
Met:
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match4-2cu3ch.html
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match4-2cu3ch.sgf
-----
Bit of commentary. This match imho was not good for either of us. I was way
down, and I had a frustration factor and was hard to be composed. Luck
decided this match, and I just didn't have enough of it. Dice are fickle,
and It happens.
Of interest were a controversial play I made was Game 16 Move 7. An unlucky
roll that stirred up some boos from the crowd. The bot is correct, but this
is a style of play I don't mind (Even though it may not be the right play)-
this was one of those gut feel plays for me.
An interesting mistake (Game 10 move number 36 - bearing off) - I pointed
out after the fact to the kibitzers (I was too focused on gammon to notice
the right play was safe - and it was obvious). I paid for my sin for a
potential loss of 2 extra points.I rolled one of the worst possible rolls,
and Monty capitalized on it with a hit.
I just couldn't get anything going, and Monty won a good match - Congrats
again.
Mike
Michael Petch
On 4/27/08 2:35 PM, in article C43A3FB8.57EA4%mpe...@capp-sysware.com,
"Michael Petch" <mpe...@capp-sysware.com> wrote:
> Bit of commentary. This match imho was not good for either of us.
>
...
> I just couldn't get anything going, and Monty won a good match - Congrats
> again.
>
I should clarify that I meant it was a "good match" in the sense "It wasn't
horribly bad", but it wasn't good from the perspective of the skill in the
other matches in the series.
Raccoon
> Fourth match results are now posted:
>
> down, and I had a frustration factor and was hard to be composed. Luck
> decided this match, and I just didn't have enough of it. Dice are fickle,
> and It happens.
It happens. Thanks for posting these matches. Win a gammon in game 10,
double and win a gammon in game 11, and we'd have a very different
match. There were some sharp plays and doubles. And some blunt
objects.
Game 1: A bit of an own-goal. Second move, 5-1 to play. This is not a
race. Just anchor up 24/18 and play from there. Always look for a play
that makes some improvement -- it's usually possible. The bar point
anchor is a good anchor. 24/18 is the only play that makes any
improvement here.
After the double: 4-4 9/1 6/2(2) gives up the 6 point. You need that
point to contain a hit checker. 5-3 20-15 11/8. 20/17 11/6 would have
remade the 6 point. 5-3 15/10 8/5. 15/7 would have kept 2 checkers to
remake the 6 point with. 5-4 21/12* of course. But the 6 point is
open, so Monty's 6 (and next 6, and next 6) win the game and gammon
for him.
Game 2: Another own-goal.
After double/take: 2-2 to play. No need to make the 2 point. That
wastes pips and Opponent's not going to leave you a shot next roll.
5-4 to play: 6/1 5/1 wastes pips again, and the 4 point is what you
want, not the acepoint. After 6/1 5/1, you'll lose the midpoint next
roll anyway so 13/9 13/8 is the natural play. After 13/9 13/8, of
course you will not be able to hit with direct aces if Opponent runs
from the midpoint next roll. But if you don't play 13/9 13/8, he won't
run one checker, anyway, so you won't be getting any direct shots if
you do keep the midpoint. 13/9 13/8 might even gain you some hits,
since only after 13/9 13/8 might Opponent choose to run.
3-1 to play: 6/5 6/3 keeps a 5-point board. After 13/9, Opponent will
be happy to hit with 11, 12, 13 and 14, as he did one roll later.
Gammon.
Game 5: Good double! You've looked up the name, and got his number.
You know Opponent likes to pass these easy takes.
Game 6: Probably a good double against your gammon-shy opponent, who
may not have noticed that 4-3 and 4-6 don't even complete the 6-prime
cleanly, and he'll have plenty of counterplay after 4-5. Many of your
non-4s are awful.
Game 7 The second 6-5 to play. Opponent's position looks awkward, but
in fact no roll forces him to leave a blot. Probably, he'll get a
better distribution. 22/11 puts you slightly ahead in the race. 5-5
4-4 and 5-3 are his only pointing numbers, the doublets being great
rolls anyway. There's no timing to play a waiting game with the 22-
point anchor and again, the chance that Opponent ever leaves you
something to hit isn't all that large.
The cube. Gnubg would take, rollout says pass. No big deal either way.
But notice that Opponent's pipcount doesn't tell the whole story, with
his pip-wasting checkers on the 2 and 1 points. It's close.
Game 8.
The second 6-5. Ahead in the race, race with bar/14. You're up 25 pips
after the roll. Bar/14 is consistent with the previous play, too. It's
not that you want to get hit, but you're a big favorite if he misses,
and you have some good return rolls if he doesn't.
The cube: against this opponent, you might have cashed 5 rolls
earlier, before the 3-2. Or on the next roll, before the 4-2. After
hitting with 4-2, for sure you'd have gotten a point doubling on the
next roll, before rolling 5-4. With Opponent firmly anchored, it's not
good, but you win some gammons. You're a favorite. Put him to the
test.
The cube (again). He's got one checker on the bar and another blot 6
away from you. Play on! This looks suspiciously like a "damn I wish
I'd doubled before so I better double now" sort of double.
Game 9.
5-5 to play. 15/5* 8/3(2) may look funny with 2 blots in the homeboard
and a stack on the 6 point, but your 5 point is the point you need to
make here -- and stop Opponent from making. 8/3*2(2) 6/1(2) makes a 3
and 1/2 point board, but big deal. Opponent can still hit with 2s, or
simply roll a 5 to anchor. Big swing on this play -- the next-roll 5-1
would have made the 5 point.
4-3 to play. You don't really want to play a 1-3 backgame with your
acepoint already made. Bar/22 13/9 makes a 2nd anchor but leaves you
with 4 awkward checkers on the 24 point, may give you two more men
back, losing more gammons, with no timing for the 1-3 game. 13/9 is
too loose -- outfield control may still be important later -- you do
have a 3-point board, after all. He won't (or shouldn't mind) hitting
more blots, given your suspect timing for a backgame. Bar/21 24/21
makes the best available anchor and puts 1 of the 4 acepoint checkers
to use. He's not likely to prime you by making both the bar point and
the 9 point, but if you play bar/22 13/9, I imagine he's likely to
make at least two out of three of the 9 point, bar point and 4 point.
Also, after bar/21 24/21 he needs 5s everywhere. After bar/22 13/9, he
has a lot more constructive numbers (again, he shouldn't mind sending
both outfield blots back -- you're in danger of crashing immediately).
6-1 and 5-3 to play. You really need to work on your rolling-to-hit-
the-blot technique!
5-2 to play. Not running 24/17 gains on 4-1 1-1 2-2, but look at the
double shots you get after 24/17 and 6-4 5-4 3-4 and single shot on
4-4.
Game 11: Interesting that the cube comes so late with so many threats
in this game, immediately after a no-threat double the game before.
Here, you have a nice gammonish double already before your 3rd roll,
from the bar. You've got a 3 point board, Opponent has 3 checkers on
the acepoint. Probably, you'd have gotten a point right there. You
trail 8-away 4-away. If you double, the cube is not very alive. Your
takepoint on a 4-cube is down around 12%. With this Opponent, you
won't see a 4-cube in unless his winning chances reach 103%, give or
take.
Game 13: The cube. It looks takable at this score. Immediate threats
are few. If things go your way, your 4-cube will be strong.
Game 14:
6-4 to play. Opponent was happy to get 15 ways to hit the slotted bar
point! Just play 20/10. The thing is, you're well enough without
slotting that the when slotting works it doesn't gain so much, but
when it doesn't work (15 rolls!) it loses a lot. The score, and your
gammon chances, give you a good double before that roll -- probably,
you'd have gotten a point right there, too.
Game 15:
You roll 6-6, he fans ... Lady Luck is a two-faced gal.
Game 16:
54 to play. The roll you mentioned ... such double-hit no-threat
double-blot moves just don't work out well, unless Opponent fans with
6-6. Anything else and you have next to nothing. Something a little
more natural is called for. No play can make much out of this nothing
roll. 13/9 6/1* minimizes hits to 15, so that's probably best.
Game 17:
Your fifth roll 54. "Pure play" wins at DMP. 13/9 13/8! threatening to
make the 9 point while Opponent has a blot in his board.
2-2 to play. 13/11(2) 10/6 gives you 3 spares to play. I guess that's
why you made that play. But 13/9(2) puts 2 checkers exactly where they
should be -- blocking Opponent's sixes. As it turns out, you wouldn't
have been able to keep this point, and Opponent's consecutive 6-6's
would have been winners anyway. But we don't play results, usually ;)
3-1 to play. 11/8 6/5 reduces hitting numbers from 4 to 3, but 4-4
becomes a killer (hits and covers the 2 point).
1-1 to play. Again I see that 8/4 minimizes shots to 3, but 8/6 8/7(2)
(a) makes the bar point, which may be very useful soon (b) is probably
your best attempt to put a 4th builder in place next roll (c) makes
5-5 good but 4-4 bad, instead of 5-5 bad and 4-4 very good (escaping
and hitting your outfield blot) (d) doesn't duplicate the 5s and 1s
you need to enter if he hits your outfield blot -- 8/6 8/7(2) ... 6-2
22/14* ... you return hit with 52 53 54 11 12 13 14 -- 13 numbers.
8/4 .... 6-2 22/14* ... you return hit with 51 52 55 11 12 -- 9
numbers.
5-1 to play. It's time to go now with 20/14 while Opponent has only a
4 and 1/2 point board. Opponent is out of time. 6/1 5/4 loses two
builders you need to attack with when he runs. 20/14 covers the
outfield. Only his 6-5 1-4 1-1 hit happily.
This game was decided by jokers -- your 6-6 followed his 6-6 twice --
but after 5-1 20/14 a more frequent continuation would have seen him
with a crashed board, having to run off his anchor, with your two back
checkers covering the outfield.
Michael Petch
On 4/27/08 11:37 PM, in article
59e29dcd-be10-46cf...@u12g2000prd.googlegroups.com, "Raccoon"
<racg...@yahoo.com> wrote:
> It happens. Thanks for posting these matches. Win a gammon in game 10,
> double and win a gammon in game 11, and we'd have a very different
> match. There were some sharp plays and doubles. And some blunt
> objects.
Thanks for the response. I skimmed your post, but wish to followup later
when I am actually awake to learn from your expertise. At 1am my mind is
shutting off. I'll post a more thorough followup when I can do your post
more justice ;-). I don't mind being critical of my own game, and I accept
constructive criticism well.
Just some quick comments. The 5-1 was just playing to fast (game 1). After
the fact it was obvious there was only one play. Later in the match you
alluded to where my frustration came from and where I started to make more
errors. My mind set was that "I am getting direct shots and pretty much
missing everything". I knew my game was suffering because of that. I began
to be cautious on cubes where I thought Monty had a take because I figured I
was likely going to get the negative luck. This is key to where my failure
is to move my game to the next level - "I start playing thinking I know
what's going to be rolled".
If anyone watches me play (And I am hoping to get match 7 on Fibs) they will
realize that I play very fast. That is another problem with my game. I often
play the first move I think is right. If I sat and looked a second time I
would likely drop my error rate. I play an almost "Automatic Backgammon
Game" - basically I play at my current level on auto pilot.
In this match, I can think of only one play that I actually paused (And
commented in the chat log). Usually with cubes I already know in advance of
my opp giving me a cube what my response will be. Monty's cube in Game 7 - I
actually took a lot of time on it. My gut instinct was to take it, but
decided the drop wasn't so bad. Thanks for letting me know the rollout
result.
Although my error rate was not a whole lot worse than my other matches, I
felt after running the analysis that there were many "no brainers" checker
plays that had I seen them I would *never* have made the play I made! I felt
it was bad for me because I figured I should have had a checker play error
rate closer to 125. I was not happy with my performance. Cubes should have
been better but I became overly cautious given frustration.
I'll review the rest of your comments and post a follow-up. Take care,
My only hope is that I am able to put up some fight next match. Monty is
running over me!
Michael
Michael Petch
Fifth match results are now posted:
These results are 2ply Checker / 2ply Cube (Given the discussion I will post
the results with those sends - using HUGE move filter) and G11 Met:
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match5.html
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match5-chat.rtf
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match5.mat
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match5.sgf
These results are 3ply Checker / 2ply Cube (using HUGE move filter) and G11
Met:
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match5-2cu3ch.html
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match5-2cu3ch.sgf
----
I lost again, but did fair well on error rates again. It is the first match
where Monty played better on Cubes (But it was reasonably close).
Game 3, Move 42. Please see the annotation, but here is a cube I knew was
bad, hoping for a slightly different outcome. The annotation has the lesson
people should take to heart!
Game 12, Move 16. The only other move I would have played different (But not
sure it would have made a difference in the race) was this double 6 and go
for a race being slightly behind. I actually thought about this one and
decided to hang back. See the annotation for my reasoning.
I didn't feel I played overly bad, it felt like a better Match than #4.
Congrats Monty!
Raccoon
>See the annotation for my reasoning.
"I wanted to run and make it a race (being down by a bit), but felt I
wasn't rolling well enough to make it a race so I hung back."
Perish the thought. Perish! Perish!
paulde...@att.net
Move number 44: mpetch to play 62
Pip counts: mpetch 13, Monty 6
Position ID: FQAAsBUAAAAAAA Match ID: AQHrAWAAMAAA
mpetch owned the cube. Hopefully, someone can translate the ID info
into a diagram. Anyway, my point is that after mpetch's 62, although
Monty was a huge game favourite, mpetch's game was not remotely close
to resignable. What happened? Were there no further rolls? If so
why? Or was the diagram inaccurate? Presumably all others who
followed the analysis have the same questions I do.
Paul Epstein
Michael Petch
On 5/4/08 5:33 AM, in article
11d196ab-6b31-494a...@b9g2000prh.googlegroups.com,
"paulde...@att.net" <paulde...@att.net> wrote:
> I'm puzzled by this:
>
> Move number 44: mpetch to play 62
>
Its a bug in the client software. There was a similar (more evident issue
raised by raccoon earlier on). In reality Monty rolled a double 6 to bear
off. The roll wasn't recorded (Occasionally occurs on last move of a game)
and game ended.
Thanks for pointing it out.
Michael
Michael Petch
roll in this case.
On 5/4/08 12:12 PM, in article C44358AE.58861%mpe...@capp-sysware.com,
"Michael Petch" <mpe...@capp-sysware.com> wrote:
>
> On 5/4/08 5:33 AM, in article
> 11d196ab-6b31-494a...@b9g2000prh.googlegroups.com,
> "paulde...@att.net" <paulde...@att.net> wrote:
>
>> I'm puzzled by this:
>>
>> Move number 44: mpetch to play 62
>>
> Thanks for pointing it out.
>
> Michael
>
Michael Petch
Sixth match results are now posted:
These results are 2ply Checker / 2ply Cube (Given the discussion I will post
the results with those sends - using HUGE move filter) and G11 Met:
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match6.html
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match6-chat.rtf
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match6.mat
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match6.sgf
These results are 3ply Checker / 2ply Cube (using HUGE move filter) and G11
Met:
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match6-2cu3ch.html
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match6-2cu3ch.sgf
-----
I have put my thought in a number of annotations (I think 6 or 7 of them).
I did manage to win a match finally so good for me! And congrats to Monty
for making it difficult as usual.
I started out the match aggressive on cube play. But after a big lead early
on I started to play conservatively on the cube. I had reasonably better
checker play this time, and the one checker blunder I made was a position I
thought about and made the right choice first, and then played my gut which
was wrong ;-).
I really felt like this was a bad match (Error rate wise) before analyzing
it. I was a bit surprised we faired as well as we did.
paulde...@att.net
>
> Game 3, Move 42. Please see the annotation, but here is a cube I knew was
> bad, hoping for a slightly different outcome. The annotation has the lesson
> people should take to heart!
>
No, I disagree with this self-criticism. I think this practical
(albeit anti-theoretical) double was absolutely correct by you.
If I remember the analysis correctly, the double was good provided
your opponent would pass 38.7% of the time or more?
(I'm a bit busy/lazy so please correct the figure if I'm wrong.)
Now, I'd like to suggest a thought experiment. Imagine an online
bookmaker who is able to set odds instantly. Let us imagine further
that the bookmaker is only setting the odds to break even, and is not
using any bid-ask spread. Suppose the bookmaker was setting odds on
the event: monty will reject. Surely, the market probability would be
greater than 38.7% (probably about 50%); after all, monty had a
significant history of weak passes. So your double was correct in my
opinion.
In backgammon, you can't go by the results alone to determine if your
choices were right or wrong, and you can't reproach yourself for not
being able to read your opponent's mind.
Paul Epstein
Michael Petch
Third match results are now posted:
These results are 2ply Checker / 2ply Cube (Given the discussion I will post
the results with those settings - using HUGE move filter) and G11 Met:
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match7.html
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match7-chat.rtf
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match7.mat
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match7.sgf
These results are 3ply Checker / 2ply Cube (using HUGE move filter) and G11
Met:
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match7-2cu3ch.html
http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match7-2cu3ch.sgf
----
I lost the final Match so the results for wins is Monty 5, mpetch 2. I
played a good match on both checkers and cubes. Monty had another good match
with a low error rate.
I've put 2 annotations in, both on the last game. One was an early checker
play move, and the other was a cube.
One thing I did feel in this match was that Monty was cubing earlier in the
hopes I would drop (Just a feeling). Most of my decisions on takes/passes
were correct (only one small error). Usually when I am playing with a
significant lead in a match I will drop some takes (Like last match), but
generally when I am trailing I try to give up as little equity on cube
decisions as possible.
My opinion though is this. I may have lost the series 5-2, but I believe
that over the long haul at present skill I should be able to split matches
down the middle with Monty (My feeling is we are closely matched in skill).
As I predicted after the first couple matches, likely skill isn't the
deciding factor between us, but which of us gets luckier.
I played at a level I suggested I would before this began, and I'm not at
all displeased with the results!
Congrats Monty, and thanks for being a good sport!
Michael Petch
On 5/10/08 2:37 PM, in article C44B63B5.58D2A%mpe...@capp-sysware.com,
"Michael Petch" <mpe...@capp-sysware.com> wrote:
> Third match results are now posted:
Should have been "Seventh match results are now posted:"
bob
Someone help me out please with game 4 move 9. Seems to me that 6/4
is better for the 2. The blot is only hit with a 56 and it is not a
devastating hit. 6/4 gives slightly better structure in home board and
leaving it on 11 aims for the 5 point. My gnubg rollout agrees with
snowie evaluation though that it is definitely wrong.
I have a bot question concerning game 5 move 5. It gave your move on
3 ply an equity of -1.000
But then for the ensuing cube decision it gave a 2 ply equity of -.884
Seems to me like those equities should have been the same.
Bob Koca
Michael Petch
On 5/10/08 4:59 PM, in article
72cc34e0-a03e-4711...@r66g2000hsg.googlegroups.com, "bob"
<bob_...@hotmail.com> wrote:
> Someone help me out please with game 4 move 9. Seems to me that 6/4
> is better for the 2. The blot is only hit with a 56 and it is not a
> devastating hit. 6/4 gives slightly better structure in home board and
> leaving it on 11 aims for the 5 point. My gnubg rollout agrees with
> snowie evaluation though that it is definitely wrong.
>
After the fact I would agree. Can I say brain phart?
> I have a bot question concerning game 5 move 5. It gave your move on
> 3 ply an equity of -1.000
> But then for the ensuing cube decision it gave a 2 ply equity of -.884
> Seems to me like those equities should have been the same.
>
Might be a good question for the Gnubg discussion forum, I actually see your
point (Looking at the 2cu3ch version, The values you are looking at are
slightly different than what I see but the comment still applies).
As for the actual play (which you didn't ask about but I'll comment), I must
admit given how much game was left making the 23 point is a better choice.
Covering my 4pt will likely occur before Monty has to leave the shot anyway.
Mike
Grunty
> http://www.capp-sysware.com/analysis/mpetch_vs_Monty-match7.mat
Michael,
this file is missing.
Michael Petch
On 5/10/08 9:56 PM, in article
81a16704-2abf-4357...@z72g2000hsb.googlegroups.com, "Grunty"
<grunti...@yahoo.com> wrote:
Human user stupidity on my part :). I have fixed it, and file is there now.