Arimaa Forum (http://arimaa.com/arimaa/forum/cgi/YaBB.cgi)
Team Games >> 2007 One vs TheMob >> Move 16
(Message started by: Fritzlein on Oct 19th, 2007, 6:00pm)

Title: Move 16
Post by Fritzlein on Oct 19th, 2007, 6:00pm
Here is the branch of the tree we played:

15b ed4w ec4e Dc3n ha6s
.    16w Ee5w Hb3w rb4s Ra2e
.    .    16b ha5e Dc4w mc5s ed4s
.    .    .    17w rb3e Ha3e Ed5w md2s
.    .    .    .    17b Db4w mc4w Hb3w mcb4s
.    .    .    .    .    18w Ec5s Ec4w Md1n De2n
.    .    .    .    .    .    18b Ha3s mb3w Da4n ma3n
.    .    .    .    .    .    .    19w Eb4s Eb3w Ea3e ma4s
.    .    .    .    .    .    .    .    19b hb5n hb6w ha6e Da5n (=+)
.    .    .    17w Ed5n Ed6n cc7s cc6x Ed7w
.    .    .    .    17b ed3w ec3e Ccn2 Cc3x rb8s (=+)
.    .    16b Dc4w ed4w ha5s mc5w
.    .    .    17w Ed5n Ed6n cc7s cc6x Ed7w
.    .    .    .    17b ec4s ec3n Ccn2 Cc3x rb8s
.    .    .    .    .    18w rb3e Ha3e Rc1n Md2n (unclear)
.    .    16b Dc4w ed4w ha5s dd8s
.    .    .    17w Ed5n mc5e rb3e Ha3e (=)
.    .    16b ha5e Dc4w mc5s dd8s
.    .    .    17w rb3e Ha3e Ed5w (+=)
.    16w Ee5w mc5w Ed5w Dc4s
.    .    16b rb4w mb5n dd8s ed4w
.    .    .    17w Hb3w ra4e Ha3n Dc3w
.    .    .    .    17b hg6e hh6w Rh5n rh8s (=+)
.    .    16b ed4w hg6e hh6w Rh5n
.    .    .    17w Hb3w rb4s mb5s Ec5w (+=)
.    16w Ee5w mc5n Ed5w Dc4s  
.    .    16b hg6e hh6w Ra5n ed4s (=+)

I wonder what the chances are that chessandgo will play one of the three moves contemplated.  If we once again failed to consider his actual move, that is no reason to despair.  One of the beautiful things about Arimaa is that you don't necessarily have to anticipate your opponent's moves to win.

I notice that I had accidentally marked the main line as being advantageous to Gold (+=) instead of advantageous to Silver (=+).  It goes to show how uninfluential the tree is.  The majority voted for a move marked as being refuted!

Title: Re: Move 16
Post by Fritzlein on Oct 19th, 2007, 8:11pm
Thanks mostly to the server outage, our reserve is halfway exhausted.  This is the situation where time management is most critical.

With a full reserve there is no reason to move quickly, whereas with an empty reserve there is no option to move slowly.  Being in the middle as we are, it will sometimes be wise to move quickly and sometimes wise to move slowly.

Title: Re: Move 16
Post by Soter on Oct 31st, 2007, 4:05pm
C&G's move: Md2n Hb3s rb4s Ee5w

Title: Re: Move 16
Post by Fritzlein on Oct 31st, 2007, 8:54pm
Thanks for noticing and posting the move, Soter.

Once again, chessandgo has totally taken me by surprise.  We can throw away the whole analysis tree and start over.  I didn't consider any moves containing Md2n.  That superficially looks strong, because our elephant now needs an extra step to get into its natural square on d3, and if we don't take d3, gold's camel will have access to b3, stopping any hopes we have of advancing our horse.

Without any analysis, just on a gut reaction, I think we have to push the gold camel back to d2 and take d3 with our elephant.  Let me suggest 16b Md3s ed4s ha5e dd8s.  This gets our elephant to its natural square and threatens the exposed dog with capture once again.  If chessandgo uses two steps to frame our rabbit with his horse, that only serves further to cut his dog off.

Although chessandgo has produced another move which appears quite strong, and which I totally overlooked, I am not yet despairing.  He is gaining traction against our rabbit, but it seems that we have adequate compensation in the form of his exposed dog and eventual action against his exposed h5-rabbit.

If folks would suggest a few moves, we can start debating the merits of our alternatives.

Title: Re: Move 16
Post by camelback on Oct 31st, 2007, 11:41pm
I haven't analyzed much but how about this move?

16b Md3s ed4s ha5e hb5s

Title: Re: Move 16
Post by The_Jeh on Oct 31st, 2007, 11:47pm
I haven't examined much either, but what about 16 hg6e hh6w Rh5n ha5e?

Title: Re: Move 16
Post by Soter on Nov 1st, 2007, 7:12am
I think I like your move The Jeh - Gold is bound to lose more (materially) than Silver, at least at first glance.

Title: Re: Move 16
Post by UruramTururam on Nov 1st, 2007, 8:40am
Chessandgo escaped from the tree... again!
Well, at least things look interesting.

My idea: 16b: Dc3w ed4w ha5e hg6e


Title: Re: Move 16
Post by Fritzlein on Nov 1st, 2007, 12:36pm

on 10/31/07 at 23:47:41, The_Jeh wrote:
I haven't examined much either, but what about 16 hg6e hh6w Rh5n ha5e?

I think we need to look at the possibility of threatening Gold's exposed h5-rabbit on every move.  That needs to be in the back of our thoughts all the time.  However, it seemed premature for 15b, and on 16b similarly:

16b hg6e hh6w Rh5n ha5e
.    17w rb3e rc3x Hb2n Hb3n Dc4s

I think chessandgo would be relieved to trade rabbits and retreat his dog for an even game.  My hope is to capture his rabbit while still delaying the capture of ours, unless we can get sufficient play against his dog to justify letting go of our rabbit.

Title: Re: Move 16
Post by mdk on Nov 1st, 2007, 2:53pm

on 11/01/07 at 12:36:52, Fritzlein wrote:
I think we need to look at the possibility of threatening Gold's exposed h5-rabbit on every move.  That needs to be in the back of our thoughts all the time.  However, it seemed premature for 15b, and on 16b similarly:

16b hg6e hh6w Rh5n ha5e
.    17w rb3e rc3x Hb2n Hb3n Dc4s

I think chessandgo would be relieved to trade rabbits and retreat his dog for an even game.  My hope is to capture his rabbit while still delaying the capture of ours, unless we can get sufficient play against his dog to justify letting go of our rabbit.


I agree that this gives up any advantage we have. Also I haven't had a chance to look at lines carefully but i think

16b hg6e hh6w Rh5n ha5e
.    17w Ed5n mc5e md5e Ed6s

may actually leave us at a disadvantage

Title: Re: Move 16
Post by Fritzlein on Nov 1st, 2007, 3:10pm

on 10/31/07 at 23:41:18, camelback wrote:
16b Md3s ed4s ha5e hb5s

It makes sense strategically to get our horse in front of our camel, but even so I didn't like this move at first, because it leaves our camel frozen and thus doesn't threaten the c4-dog with immediate capture.  However, I can't immediately find a refutation.  Furthermore your suggestion seems to rule out the reply 17w Ed5n mc5e md5e Ed6s, which will be useful to chessandgo against some moves.  Possibly the camel flip refutes my suggested move of 16b Md3s ed4s ha5e dd8s.  But if chessandgo tries to flip our camel after 16b Md3s ed4s ha5e hb5s, we can simply take his dog.

Title: Re: Move 16
Post by mdk on Nov 1st, 2007, 3:11pm

on 11/01/07 at 08:40:56, UruramTururam wrote:
Chessandgo escaped from the tree... again!
Well, at least things look interesting.

My idea: 16b: Dc3w ed4w ha5e hg6e


16b: Dc3w ed4w ha5e hg6e
.    17w rb3e Hb2n Ed5n mc5e

and unless I am missing something (which is quite possible) we are clearly at a disadvantage. I see no choice but for us to give up our framed rabbit to protect our camel without capturing gold's advanced rabbit in return.

Title: Re: Move 16mc5e
Post by mdk on Nov 1st, 2007, 3:22pm

on 11/01/07 at 15:10:50, Fritzlein wrote:
It makes sense strategically to get our horse in front of our camel, but even so I didn't like this move at first, because it leaves our camel frozen and thus doesn't threaten the c4-dog with immediate capture.  However, I can't immediately find a refutation.  Furthermore your suggestion seems to rule out the reply 17w Ed5n mc5e md5e Ed6s, which will be useful to chessandgo against some moves.  Possibly the camel flip refutes my suggested move of 16b Md3s ed4s ha5e dd8s.  But if chessandgo tries to flip our camel after 16b Md3s ed4s ha5e hb5s, we can simply take his dog.


16b Md3s ed4s ha5e hb5s
.    17w Ed5n mc5e rb3e Hb2n

how do we continue?

sorry if it seems that im simply opposing moves without suggesting a move myself. My time is very limited... first college app due in less than 8 hours now so i'm just offering what i can in the time i have

Title: Re: Move 16mc5e
Post by Fritzlein on Nov 1st, 2007, 3:55pm

on 11/01/07 at 15:22:51, mdk wrote:
16b Md3s ed4s ha5e hb5s
.    17w Ed5n mc5e rb3e Hb2n

how do we continue?

How about this?

16b Md3s ed4s ha5e hb5s
.    17w Ed5n mc5e rb3e Hb2n
.    .    17b Dc4n hb4e Dc5n hb4n

I prefer calmer lines, but sometimes you have to be willing to play sharply.  As you pointed out before, we should be able to handle sharp lines at least as well as he can, since we have more eyeballs on it.  In the above line, can he endanger our camel without sacrificing his dog?

.    .    .    18w Dc6w Db6w md5e Ed6s
.    .    .    .    18b ce7s me5s me4w md4w

Still sharp but we may be ahead.  This is the type of position where refutations abound; I'm just trying to get a tree started, even if my lines are all mistaken.


Quote:
sorry if it seems that im simply opposing moves without suggesting a move myself. My time is very limited... first college app due in less than 8 hours now so i'm just offering what i can in the time i have

To understand the position well, it's also important to look for improvements for us, especially as the tree gets deeper.  On the other hand, it's good to get at least one reply to every move, so we can start a tree.  Furthermore, it's best to look at the sharpest replies first.  If all of our suggestions so far can be refuted outright, that's important to know!  (Which is to say, your contributions are very valuable...)

Title: Re: Move 16
Post by Soter on Nov 1st, 2007, 4:15pm
Is 16b ha5e hb5s rb3w dd8s worth anything?

Title: Re: Move 16
Post by Fritzlein on Nov 1st, 2007, 5:08pm

on 11/01/07 at 16:15:30, Soter wrote:
Is 16b ha5e hb5s rb3w dd8s worth anything?

Actually, getting our rabbit over to a3 was my first inclination, but I got afraid of the gold camel coming to b3, so I didn't mention it.  On a second look, however, our horse isn't in all that much danger.  Also we may have some compensation because his horse is on a bad square at b2.  Maybe we should spend some time thinking about this line.  It seems we would have to give up on hunting his dog and chase his camel instead.  Maybe

16b ha5e hb5s rb3w dd8s
.    17w Cc2e Md3w Mc3w Hb2e
.    .    17b Dc4s ed4w mc5n ha4w

with a threat to take a camel hostage.  Thanks for raising the possibility so I re-examined my gut reaction.

Title: Re: Move 16
Post by Fritzlein on Nov 1st, 2007, 5:25pm
Proto-tree:
16b Md3s ed4s ha5e dd8s
.    17w Ed5n mc5e rb3e Hb2n
.    .    17b hb5e md5s Dc3e md4w (=+)
16b ha5e hb5s rb3w dd8s
.    17w Cc2e Md3w Mc3w Hb2e
.    .    17b Dc4s ed4w mc5n ha4w (=+)
16b Md3s ed4s ha5e hb5s
.    17w Ed5n mc5e rb3e Hb2n
.    .    17b Dc4n hb4e Dc5n hb4n
.    .    .    18w Dc6w Db6w md5e Ed6s
.    .    .    .    18b ce7s me5s me4w md4w (?)
16b hg6e hh6w Rh5n ha5e
.    17w Ed5n mc5e md5e Ed6s (+=)
.    17w rb3e rc3x Hb2n Hb3n Dc4s (=)
16b Dc3w ed4w ha5e hg6e
.    17w rb3e Hb2n Ed5n mc5e (+=)


I hesitate to order the tree this way before mdk has a crack at the top three lines, but I have to move the other two suggestions to the bottom because those (at the moment) stand concretely refuted.  Of course new top-level suggestions are still welcome in addition to growing the branches on the given lines.

Title: Re: Move 16
Post by UruramTururam on Nov 1st, 2007, 8:39pm

on 11/01/07 at 15:11:53, mdk wrote:
16b: Dc3w ed4w ha5e hg6e
.    17w rb3e Hb2n Ed5n mc5e


Yes, this is a painful answer.
C&G had a longer thought before his latest move, I think he calculated it well... But freezing one's own camel is quite an unusual move. I wonder if we have a strong response that leaves the camel frozen...

By now the move that came into my mind is kamikazish orchestrating EMHR against c3:
16b: ha5e Dc4w mc5s hb5e




Title: Re: Move 16
Post by Soter on Nov 2nd, 2007, 8:22am

Quote:
By now the move that came into my mind is kamikazish orchestrating EMHR against c3:  
16b: ha5e Dc4w mc5s hb5e


How about a "cousin" of Ururam's move:
ha5e Dc4w mc5s dd8s?

I know this is no holy grail ( the rabbit frame seems imminent anyway ), but I can't think of anything better now.



Title: Re: Move 16
Post by RonWeasley on Nov 2nd, 2007, 8:42am
Good discussion so far.  Let's try to begin voting Monday morning and ending Tuesday evening.

I'm leaning toward 16b Md3s ed4s ha5e dd8s because the dog threat maintains the initiative.

Title: Re: Move 16
Post by arimaa_master on Nov 2nd, 2007, 12:34pm

on 11/01/07 at 17:25:42, Fritzlein wrote:
Proto-tree:
16b Md3s ed4s ha5e dd8s
.    17w Ed5n mc5e rb3e Hb2n
.    .    17b hb5e md5s Dc3e md4w (=+)
16b ha5e hb5s rb3w dd8s
.    17w Cc2e Md3w Mc3w Hb2e
.    .    17b Dc4s ed4w mc5n ha4w (=+)
16b Md3s ed4s ha5e hb5s
.    17w Ed5n mc5e rb3e Hb2n
.    .    17b Dc4n hb4e Dc5n hb4n
.    .    .    18w Dc6w Db6w md5e Ed6s
.    .    .    .    18b ce7s me5s me4w md4w (?)
16b hg6e hh6w Rh5n ha5e
.    17w Ed5n mc5e md5e Ed6s (+=)
.    17w rb3e rc3x Hb2n Hb3n Dc4s (=)
16b Dc3w ed4w ha5e hg6e
.    17w rb3e Hb2n Ed5n mc5e (+=)


I hesitate to order the tree this way before mdk has a crack at the top three lines, but I have to move the other two suggestions to the bottom because those (at the moment) stand concretely refuted.  Of course new top-level suggestions are still welcome in addition to growing the branches on the given lines.


Since no one answered to my suggestion:

"And here is my first-glance suggestion:

16b Dc4s ed4w mc5n dd8s (=+)?
"

And here (see below) are my "refutations" of two top lines

so I incline to my 16b Dc4s ed4w mc5n dd8s (=+ / =)?


1)
16b Md3s ed4s ha5e dd8s

I am a little bit worried about camel flip:

17w Ed5n mc5e md5e Ed6s


2)
16b ha5e hb5s rb3w dd8s

the same camel flip may work (although there could be some sharp lines)

17w Ed5n mc5e md5e Ed6s


3)

16b Md3s ed4s ha5e hb5s
.    17w Ed5n mc5e rb3e Hb2n
.    .    17b Dc4n hb4e Dc5n hb4n
.    .    .    18w Dc6w Db6w md5e Ed6s
.    .    .    .    18b ce7s me5s me4w md4w (?)

I donīt really know the evaluation too but the next move could be  

19w ce7s me5s me4w md4w










Title: Re: Move 16
Post by 99of9 on Nov 2nd, 2007, 10:56pm
I haven't even had time to read the other suggestions, but my suggestion is:
Dc4w ed4w mc5w dd8s

Title: Re: Move 16
Post by mdk on Nov 3rd, 2007, 1:02pm

on 11/01/07 at 17:25:42, Fritzlein wrote:
Proto-tree:
16b Md3s ed4s ha5e dd8s
.    17w Ed5n mc5e rb3e Hb2n
.    .    17b hb5e md5s Dc3e md4w (=+)
16b ha5e hb5s rb3w dd8s
.    17w Cc2e Md3w Mc3w Hb2e
.    .    17b Dc4s ed4w mc5n ha4w (=+)
16b Md3s ed4s ha5e hb5s
.    17w Ed5n mc5e rb3e Hb2n
.    .    17b Dc4n hb4e Dc5n hb4n
.    .    .    18w Dc6w Db6w md5e Ed6s
.    .    .    .    18b ce7s me5s me4w md4w (?)
16b hg6e hh6w Rh5n ha5e
.    17w Ed5n mc5e md5e Ed6s (+=)
.    17w rb3e rc3x Hb2n Hb3n Dc4s (=)
16b Dc3w ed4w ha5e hg6e
.    17w rb3e Hb2n Ed5n mc5e (+=)


I hesitate to order the tree this way before mdk has a crack at the top three lines, but I have to move the other two suggestions to the bottom because those (at the moment) stand concretely refuted.  Of course new top-level suggestions are still welcome in addition to growing the branches on the given lines.


Sorry I haven't responded sooner. There seem to be a number of plausible suggestions at this point.

A couple of general comments first. The camel flip 17w Ed5n mc5e md5e Ed6s seems to be a likely response for many of the moves that do not move the camel. Does this mean that our camel belongs on the western wing? I'm not completely sure. In some lines I think so  in other lines I don't. Also we need to be careful that we do not leave our camel in a position where it will be taken hostage without ample compensation.

Now some specific lines

16b Md3s ed4s ha5e dd8s
.    17w Ed5n mc5e rb3e Hb2n
.    .    17b hb5e md5s Dc3e md4w
.    .    .    18w Ed6s Ed5n hc5e Db4w

are all our pieces safe or are we just going to end up giving up a camel hostage?

We should also look at the line of the camel flip

16b Md3s ed4s ha5e dd8s
.    17w Ed5n mc5e md5e Ed6s
.    .    17b ce7s me5e mf5n hb5s
.    .    .    18w Ed5e Ee5e Ef5e mf6s
.    .    .    .    18b ce6s mf5n Dc4n hb4e (+)
.    .    .    18w Ed5e Ee5e Ef5w mf6s
.    .    .    .     18b hg6s mf5n Dc4n hb4e (and again we look ahead)

so going after the camel on 18w doesn't work. But what if chessandgo tries something different. Is there a way he can launch an EH attack on the western side of the board? If there is he seems to be better as our camel is on the eastern side of the board.

Say he now plays 18w rb3e Hb2n Ed5n Cc2w.
How would you evaluate this position? I'm not really sure how to evaluate it. Am I just imagining that chessandgo has such a threat or is it an actual possibility?

Title: Re: Move 16
Post by mdk on Nov 3rd, 2007, 1:48pm
16b ha5e hb5s rb3w dd8s
.    17w Ed5n mc5e md5e Ed6s
.    .    17b Dc4n hb4e Dc5n Dc6x hc4n
.    .    .    18w me5s Ed5e me4s Ee5s
.    .    .    .    18b ed4w Md3n Md4n ec4e (+)
.    .    .    18w hc5s Ed5w hc4s hc3x Ec5s
.    .    .    .    18b ed4e Md3n Md4n ee4w
.    .    .    .    .    19w Ec4n Md5n Md6w Mc6w
.    .    .    .    .    .    19b ed4w ec4w eb4n dd7s
.    .    .    .    .    .    .    20w Ec5e me5e Ed5e Hg3n (-)
.    .    .    .    .    .    19b ed4w ec4w eb4n me5n
.    .    .    .    .    .    .    20w Ec5n Ec6e cc7s cc6x Hb2n(-)
    (we end up down HCR for D and M hostage)

there are other moves for 19b but i haven't found a line that doesn't leave us down. So unless someone can find a 19b that works for us or a flaw in my analysis we need a different 17b. I'm  sure there are some reasonable suggestions but I think the real thing to consider is if our position is worstened by having our camel on the east.

Title: Re: Move 16
Post by mdk on Nov 3rd, 2007, 1:56pm

on 11/02/07 at 12:34:06, arimaa_master wrote:
Since no one answered to my suggestion:

"And here is my first-glance suggestion:

16b Dc4s ed4w mc5n dd8s (=+)?
"

And here (see below) are my "refutations" of two top lines

so I incline to my 16b Dc4s ed4w mc5n dd8s (=+ / =)?

[edit]...

3)

16b Md3s ed4s ha5e hb5s
.    17w Ed5n mc5e rb3e Hb2n
.    .    17b Dc4n hb4e Dc5n hb4n
.    .    .    18w Dc6w Db6w md5e Ed6s
.    .    .    .    18b ce7s me5s me4w md4w (?)

I donīt really know the evaluation too but the next move could be  

19w ce7s me5s me4w md4w


What did you mean to suggest for 19w arimaa master?

One thing I would consider is 19w Ed5n hc5e hd5e Ed6s

our position is left imbalanced in that both our horses are on the same side and our camel suddenly seems much more vulnerable to being taken hostage. Additionally, gold's advanced dog doesn't look so bad because there isn't a horse on that side of the board to take it hostage.

So there is my crack at the top three Fritz. Hope my contribution helps bring about a better decision. Fortunately the mob seems to be filled with suggestions as there are 5 more moves that have been posted since the original tree but unfortunately I do not have time at the moment to take a look at those. Hopefully later. I'm looking forward to hearing some responses.

Title: Re: Move 16
Post by arimaa_master on Nov 3rd, 2007, 4:32pm

on 11/03/07 at 13:56:41, mdk wrote:
What did you mean to suggest for 19w arimaa master?


I am not sure that I follow your question coz I suggested

19w ce7s me5s me4w md4w  (but it is visible) - but anyway I think that we are not ahead after that nor after your suggested move - so I agree with you.

Title: Re: Move 16
Post by arimaa_master on Nov 3rd, 2007, 4:44pm

on 11/01/07 at 20:39:31, UruramTururam wrote:
Yes, this is a painful answer.
C&G had a longer thought before his latest move, I think he calculated it well... But freezing one's own camel is quite an unusual move. I wonder if we have a strong response that leaves the camel frozen...

By now the move that came into my mind is kamikazish orchestrating EMHR against c3:
16b: ha5e Dc4w mc5s hb5e



Very interesting move,
I donīt think that it is bad move, but at least I donīt see any advantage for us after:

17w hc5n Ed5w mc4s Ec5s
17b Md3e ed4s Cc2e mc3s
18w rb3w Db4s Ec4s Me3e

Title: Re: Move 16
Post by arimaa_master on Nov 3rd, 2007, 5:12pm

on 11/02/07 at 08:22:23, Soter wrote:
How about a "cousin" of Ururam's move:
ha5e Dc4w mc5s dd8s?

I know this is no holy grail ( the rabbit frame seems imminent anyway ), but I can't think of anything better now.


I think that we could be in serious trouble after:

17w Ed5w mc4s Ec5s De2n


Title: Re: Move 16
Post by arimaa_master on Nov 3rd, 2007, 5:37pm

on 11/02/07 at 22:56:36, 99of9 wrote:
I haven't even had time to read the other suggestions, but my suggestion is:
Dc4w ed4w mc5w dd8s



Very interesting move.

I thought about this:

17w rb3e Hb2n Db4w Ed5w
17b hg6e hh6w Rh5n mb5n
18w Md3s rc3e rd3e Md2n
18b ec4e ed4e Md3n rg7e
19w Ec5e re3e rf3x De2n Md4s

now we could flip the dog but it is unclear after:

19b ee4w De3n De4n ed4e
20w De5e Ed5e Df5s Ee5e
20b ha5n Da4n Da5e ha6s
21w Hb3n Db5e Dc5s Df4s


so maybe it would be better to recapture the rabbit by:

19b hg6s Rh6w Rg6w Rf6x hg5n


but now I found this variation:

16b Dc4w ed4w mc5w dd8s
17w rb3e Hb2n Db4w Ed5w
17b hg6e hh6w Rh5n mb5n
18w Md3s rc3e rd3e Md2n
18b ec4e ed4e Md3n rg7e
19w Ec5e re3e rf3x De2n Md4s
19b ee4e De3n De4n ef4w
20w De5e Df5s Ed5e Ee5e
20b ha5n Da4n ha6n Da5n
21w Df4s Ef5w Ee5w Ed5n
21b hg6s Rh6w Rg6w Rf6x hg5n (=+)

where we have dog hostage for our good but this could be complicated at any point so no guarantee at all :)



Title: Re: Move 16
Post by mdk on Nov 3rd, 2007, 6:34pm

on 11/03/07 at 16:32:50, arimaa_master wrote:
I am not sure that I follow your question coz I suggested

19w ce7s me5s me4w md4w  (but it is visible) - but anyway I think that we are not ahead after that nor after your suggested move - so I agree with you.


your 19w is a move for silver (the move that Fritz suggests for 18b). so i thought you meant to suggest a 19w that chessandgo might play

Title: Re: Move 16
Post by arimaa_master on Nov 4th, 2007, 3:20am

on 11/03/07 at 18:34:01, mdk wrote:
your 19w is a move for silver (the move that Fritz suggests for 18b). so i thought you meant to suggest a 19w that chessandgo might play


You are right, of course - sorry for that (I simply ended with

18w Dc6w Db6w md5e Ed6s

but I thought that I ended with silver move - donīt know why :)

so I find

18b ce7s me5s me4w md4w  and thought that is 19w

of course this was 18b suggested by Fritzlein before (I donīt have an explanation how I could miss it even after you pointed out at that).

So I donīt have a better move for 19w than yours.


Title: Re: Move 16
Post by mdk on Nov 4th, 2007, 2:31pm

on 11/02/07 at 12:34:06, arimaa_master wrote:
Since no one answered to my suggestion:

"And here is my first-glance suggestion:

16b Dc4s ed4w mc5n dd8s (=+)?
"
...[edit]


16b Dc4s ed4w mc5n dd8s
.    17w Cc2e Cc3s rb3e Hb2n

and chessandgo's dog is safe and our rabbit is framed.

however if gold's phant goes to the eastern side of the board our horse and camel can advance on the western side of the board. this takes time though and while these pieces were advancing chessandgo could threaten pieces on the eastern side of the board. so i'm really not sure if we are better or worse after this move.

Title: Re: Move 16
Post by Fritzlein on Nov 4th, 2007, 8:45pm
You guys are doing a great job without me.  I see we now have nine candidate moves, and each one has at least one suggested response for chessandgo.  Here are are the suggested lines so far in tree form, for what it is worth:

16b Md3s ed4s ha5e dd8s
.    17w Ed5n mc5e md5e Ed6s
.    .    17b hb5s Dc4n hb4e dd7s
.    .    .    18w me5e Ed5e mf5s Ee5e
.    .    .    .    18b ed3w ec3e Cc2n Cc3x ed3e
.    .    .    18w me5s Ed5e me4s Ede5s
.    .    .    .    18b De2s me3s ed3e dd6s
.    .    17b ce7s me5e mf5n hb5s
.    .    .    18w rb3e Hb2n Ed5n Cc2w
.    .    .    18w Ed5e Ee5e Ef5e mf6s
.    .    .    .    18b ce6s mf5n Dc4n hb4e
.    .    .    18w Ed5e Ee5e Ef5w mf6s
.    .    .    .     18b hg6s mf5n Dc4n hb4e
.    17w Ed5n mc5e rb3e Hb2n
.    .    17b hb5e md5s Dc4w md4w
.    .    .    18w Ed6s Ed5n hc5e Db4w
16b ha5e hb5s rb3w dd8s
.    17w Ed5n mc5e md5e Ed6s
.    .    17b Dc4n hb4e Dc5n Dc6x hc4n
.    .    .    18w hc5s Ed5w hc4s hc3x Ec5s
.    .    .    .    18b ed4e Md3n Md4n ee4w
.    .    .    .    .    19w Ec4n Md5n Md6w Mc6w
.    .    .    .    .    .    19b ed4w ec4w eb4n dd7s
.    .    .    .    .    .    .    20w Ec5e me5e Ed5e Hg3n
.    .    .    .    .    .    19b ed4w ec4w eb4n me5n
.    .    .    .    .    .    .    20w Ec5n Ec6e cc7s cc6x Hb2n
.    .    .    18w me5s Ed5e me4s Ee5s
.    .    .    .    18b ed4w Md3n Md4n ec4e
.    17w Cc2e Md3w Mc3w Hb2e
.    .    17b Dc4s ed4w mc5n ha4w
16b Md3s ed4s ha5e hb5s
.    17w Ed5n mc5e rb3e Hb2n
.    .    17b Dc4n hb4e Dc5n hc4n
.    .    .    18w Dc6w Db6w md5e Ed6s
.    .    .    .    18b ce7s me5s me4w md4w
.    .    .    .    .    19w Ed5n hc5e hd5e Ed6s
16b hg6e hh6w Rh5n ha5e
.    17w Ed5n mc5e md5e Ed6s
.    17w rb3e rc3x Hb2n Hb3n Dc4s
16b Dc4w ed4w ha5e hg6e
.    17w rb3e Hb2n Ed5n mc5e
16b Dc4w ed4w mc5w dd8s
.    17w rb3e Hb2n Db4w Ed5w
.    .    17b hg6e hh6w Rh5n mb5n
.    .    .    18w Md3s rc3e rd3e Md2n
.    .    .    .    18b ec4e ed4e Md3n rg7e
.    .    .    .    .    19w Ec5e re3e rf3x De2n Md4s
.    .    .    .    .    .    19b ee4e De3n De4n ef4w
.    .    .    .    .    .    .    20w De5e Df5s Ed5e Ee5e
.    .    .    .    .    .    .    .    20b ha5n Da4n ha6n Da5n
.    .    .    .    .    .    .    .    .    21w Df4s Ef5w Ee5w Ed5n
.    .    .    .    .    .    .    .    .    .    21b hg6s Rh6w Rg6w Rf6x hg5n
.    .    .    .    .    .    19b hg6s Rh6w Rg6w Rf6x hg5n
.    .    .    .    .    .    19b ee4w De3n De4n ed4e
.    .    .    .    .    .    .    20w De5e Ed5e Df5s Ee5e
.    .    .    .    .    .    .    .    20b ha5n Da4n Da5e ha6s
.    .    .    .    .    .    .    .    .    21w Hb3n Db5e Dc5s Df4s
16b Dc4s ed4w mc5n dd8s
.    17w Cc2e Cc3s rb3e Hb2n
16b ha5e Dc4w mc5s hb5e
.    17w hc5n Ed5w mc4s Ec5s
.    .    17b Md3e ed4s Cc2e mc3s
.    .    .    18w rb3w Db4s Ec4s Me3e
16b ha5e Dc4w mc5s dd8s
.    17w Ed5w mc4s Ec5s De2n


I create this tree as the first step in ordering the moves for my own votes, and for adding analysis, but since I'm not sure how far I will get in analyzing and evaluating, I thought I would at least share the raw tree.

Title: Re: Move 16
Post by Fritzlein on Nov 5th, 2007, 9:06am
For my top move, if he flips our camel I was intending to respond very sharply by leaving our camel hanging and attacking the dog instead.

16b Md3s ed4s ha5e dd8s
.    17w Ed5n mc5e md5e Ed6s
.    .    17b hb5s Dc4n hb4e dd7s

The tactics are very hard to wade through, but from what I have looked at so far, we have the upper hand.  For example:

.    .    .    18w me5e Ed5e mf5s Ee5e
.    .    .    .    18b ed3w ec3e Cc2n Cc3x ed3e

or

.    .    .    18w me5s Ed5e me4s Ede5s
.    .    .    .    18b De2s me3s ed3e dd6s

Please double check, but superficially this makes me lean more toward my first hunch 16b Md3s ed4s ha5e dd8s.

Title: Re: Move 16
Post by RonWeasley on Nov 5th, 2007, 10:23am
OK, I looked at other responses to the camel flip and the the dog attack appears to hold.  I was worried when arimaa_master warned us about the camel flip, but Fritzlein's reply seems adequate.  That exposed dog is valuable to us, so I also like Fritzlein's top move.

Anybody object if we start the vote today?

Title: Re: Move 16
Post by Fritzlein on Nov 5th, 2007, 1:42pm
Oh, since the voting has already started, please note that the tree I gave above is not ordered in terms of my evaluations or (presumably) in terms of anyone's evaluations.  It is just a collection of analysis.  If anyone who has looked at all the moves (mdk? arimaa_master?) would care to give a ranking, that could be very helpful.

Title: Re: Move 16
Post by mdk on Nov 6th, 2007, 3:00pm

on 11/05/07 at 09:06:39, Fritzlein wrote:
For my top move, if he flips our camel I was intending to respond very sharply by leaving our camel hanging and attacking the dog instead.

16b Md3s ed4s ha5e dd8s
.    17w Ed5n mc5e md5e Ed6s
.    .    17b hb5s Dc4n hb4e dd7s

The tactics are very hard to wade through, but from what I have looked at so far, we have the upper hand.  For example:

.    .    .    18w me5e Ed5e mf5s Ee5e
.    .    .    .    18b ed3w ec3e Cc2n Cc3x ed3e

or

.    .    .    18w me5s Ed5e me4s Ede5s
.    .    .    .    18b De2s me3s ed3e dd6s

Please double check, but superficially this makes me lean more toward my first hunch 16b Md3s ed4s ha5e dd8s.


Sorry I didn't respond sooner. school and college apps are really taking up all my time. In the line
16b Md3s ed4s ha5e dd8s
.    17w Ed5n mc5e md5e Ed6s
.    .    17b hb5s Dc4n hb4e dd7s
.    .    .    18w me5s Ed5e me4s Ede5s
.    .    .    .    18b De2s me3s ed3e dd6s
.    .    .    .    .    19w rb3w Hb2n Ee4w Rc1e
how do we continue so that we don't end up losing material? I've tried but I haven't been able to find a line that works. I seriously hope that I am missing something because if i am not this is probably too late to affect the elections.

[edit]
i kept looking and i guess 19b Cf2n me2e mf2e ee3w might allow us to escape with giving up a camel hostage for a dog, although i'm not completely sure about what would follow.

however another possible move is 19w is rb3w Hb2n Ee4w Rg1e.   Again the postion seems to give up a camel hostage. This may be a perfectly sound 16b and I am simply having these concerns because I have no successful experience with playing such positions. I certainly hope it is if we play this

Title: Re: Move 16
Post by RonWeasley on Nov 6th, 2007, 6:11pm
I think camel hostage for a dog favors silver.  I also think chessandgo believes that.  We would have to flood the hostage trap swiftly.

Title: Re: Move 16
Post by mdk on Nov 6th, 2007, 11:27pm
I guess sometimes i overvalue camel hostages in my mind. but you are right, chessandgo won't want to give up the dog for a camel hostage and if we play it properly should be better off with such an exchange.

Title: Re: Move 16
Post by RonWeasley on Nov 7th, 2007, 7:04am
With 10 votes reporting after extending the deadline to Wednesday morning, we have a tie!

Md3s ed4s ha5e dd8s

and

ha5e hb5s rb3w dd8s

The preference matrix favors the first of these moves, but each was ranked first by 5 voters, so it's not obvious I want to use it as a tie breaker.  The voting site suggests breaking ties randomly.  I think I agree, but I'll wait a few hours to allow mobsters to make comments and give advice.  If I get no objections, I will flip a galleon at noon EST to make the choice.

Title: Re: Move 16
Post by Fritzlein on Nov 7th, 2007, 8:18am
Here's an alternative way to break the tie:
The first Mob member who didn't vote, who expresses a preference between the top two in a post in this thread, gets to decide.  I voted, so I'm excluded.

Title: Re: Move 16
Post by Fritzlein on Nov 7th, 2007, 9:06am
For the record
    .    .    .    .    .    .    .    .    .    .  1. 2. 3. 4. 5. 6. 7. 8. 9.
1. Md3s ed4s ha5e dd8s (push M, h->b5, d->d7)    .   -  5  7  6  8  8  8  9  9
2. ha5e hb5s rb3w dd8s (h->b4, r->a3, d->d7).    .   5  -  4  4  5  5  7  6  8
3. Md3s ed4s ha5e hb5s (push M, d->b4) .    .    .   3  4  -  5  6  7  6  8  7
4. Dc4w ed4w mc5w dd8s (push D->b4, m->d5, d->d7)    4  4  3  -  7  5  7  5  5
5. Dc4s ed4w mc5n dd8s (push D->c3, m->c6, d->d7)    2  3  1  1  -  4  5  5  5
6. ha5e Dc4w mc5s hb5e (h->b5, push D->d4, h->a5)    2  3  1  3  4  -  5  5  4
7. hg6e hh6w Rh5n ha5e (pull R, h->b5) .    .    .   2  1  2  1  3  3  -  4  4
8. ha5e Dc4w mc5s dd8s (h->b5, push D->d4, d->d7)    1  1  0  3  3  2  4  -  4
9. Dc4w ed4w ha5e hg6e (push D->b4, h->b5, h->h6)    1  0  1  3  3  3  3  3  -


I see what you mean about the tiebreak, Ron.  If you compare the top two rows, they tie 5-5 head-to-head, but the top row has a higher number than the second row for every column except head-to-head.  Still, I guess we can't break ties in a way we never officially specified.

Title: Re: Move 16
Post by The_Jeh on Nov 7th, 2007, 9:31am
What would be wrong with just using the preference matrix? It seems like a valid way to break the tie to me.

Title: Re: Move 16
Post by Janzert on Nov 7th, 2007, 9:58am
I'm actually very surprised the preference matrix isn't used to break a tie. I can't find on the sight where it suggests a random choice to break a tie, but does it give a reason not to use the preference matrix?

Janzert

Title: Re: Move 16
Post by UruramTururam on Nov 7th, 2007, 11:19am
Yep; since #1 wins against every of #3-#9 more times than #2 does it can be clearly declared better by a tiebreaker.

Title: Re: Move 16
Post by RonWeasley on Nov 7th, 2007, 11:23am
The random selection has chosen Md3s ed4s ha5e dd8s.  I didn't have a coin so I flipped a candy bar from Halloween.  Then I ate it.  Chaos rules TheMob!!

This is the same move that placed first in the preference matrix.  If you go the voting site for this election, under the box is a note recommending breaking the tie with a random process.  The better placement of this move indicates that it ranked better among those who didn't rank it over the other move.  Interpreting this could be tricky because in some cases it can depend on strong feelings about moves other than these top two.  If we did a run-off to remove the effects of the losing moves, these results predict we'd get another tie, and we'd just end up losing more time.

I feel fortunate that this move got selected so we can debate the process without those favoring the preference matrix feeling the wrong choice got made.

Title: Re: Move 16
Post by Fritzlein on Nov 7th, 2007, 10:05pm
Furthermore, we gained three and a half hours of reserve.  This position is reasonably complex now, but it could get even more complex in the near future.  It's good to spend only an average amount of time on average positions, so that we can later opt to spend more than average time if we enter wild complications.

Title: Re: Move 16
Post by chessandgo on Mar 11th, 2010, 3:24am
Hi all,

I'm currently analysing the 2007 Mob game for my book, so I might be digging up some old threads.


on 11/06/07 at 15:00:24, mdk wrote:
16b Md3s ed4s ha5e dd8s
.    17w Ed5n mc5e md5e Ed6s
.    .    17b hb5s Dc4n hb4e dd7s
.    .    .    18w me5s Ed5e me4s Ede5s
.    .    .    .    18b De2s me3s ed3e dd6s
.    .    .    .    .    19w rb3w Hb2n Ee4w Rg1n


This line might have been my best chance actually, as a horse frame on c3 is forced. I'm not sure whether silver can do enough mayhem with his camel on the east meanwhile, like 19s ee3w Cf2n x 20g Rh1n Rf1e De1e Rb1n and gold will frame on c3 next move. What do you think, would silver have had to withdraw the camel on 17s (which would have been a huge relief for me obviously)?

oh and btw I also like the other 16s move that you tied for first place.

Title: Re: Move 16
Post by chessandgo on Mar 11th, 2010, 3:57am
Oh, reading the move 17 thread, arimaa_master has:

17w Ed5e mc5e Ee5n md5e
17b hb5s rb3w hb4s Dc4w
18w me5s Ee6s me4s Ee5s

As far as I can remember this was rather the line that I feared, I guess that

18b De2s me3s De1w me2n is very good for silver.



Arimaa Forum » Powered by YaBB 1 Gold - SP 1.3.1!
YaBB Đ 2000-2003. All Rights Reserved.