Arimaa Forum (http://arimaa.com/arimaa/forum/cgi/YaBB.cgi)
Team Games >> 2007 One vs TheMob >> Move 28
(Message started by: RonWeasley on Mar 4th, 2008, 8:01pm)

Title: Move 28
Post by RonWeasley on Mar 4th, 2008, 8:01pm
I told chessandgo that Bomb analysis says his dogs have fleas and they go on the carpet.

Title: Re: Move 28
Post by Soter on Mar 5th, 2008, 5:47am
ROTFL!  :D  ;D

Title: Re: Move 28
Post by RonWeasley on Mar 7th, 2008, 2:09pm
If your name is Tony Martin and you wanted to be added to TheMob, please send the request again.  Otherwise I'm assuming you wanted to be on a list associated with Johns Hopkins APL and broadcast your request to the list recipients instead of the list maintainer (and everybody is laughing at you).

Title: Re: Move 28
Post by The_Jeh on Mar 17th, 2008, 5:26pm
Chessandgo has chosen 28w Ce2e Cf2e Db3n ra3e.

Title: Re: Move 28
Post by Fritzlein on Mar 17th, 2008, 10:52pm
My first impression is that if we attack f3, chessandgo will be able to defend it adequately while capturing our advanced rabbit.  That will leave us with the advantage of being first to advance a dog in a position where an elephant-dog attack is called for, but no extra material.

I am inclined to be more materialistic and save our rabbit from capture in c3.  I would like to contest c3 with our elephant on d3 and full-out western swarm including our horse and more rabbits.  As time permitted, we would also want to advance our dog and perhaps rabbits in the east, but with the idea of complementing our western swarm positionally, not directly taking over f3 or necessarily pulling anything into f6.  It is a no-capture strategy for many moves to come, banking that our EH will keep his EH tied down in the west, and our D will match his D in the east, but somehow our advances will improve our position more than he can improve his.

The trouble in deciding between our two main strategic options is that neither one will result in payoff in the near term.  Building an analysis tree will be essentially useless.  We are in the opposite situation from a few moves ago: now there are almost no forcing lines, and the difference between a correct and an incorrect move will not show up for a long time to come.

I'm not even sure how to discuss properly in a situation like this.  We can toss out moves for discussion (mine is Ee3w Hd6w Hc6w Ra8s), but probably nobody will be able to refute anyone else's move.  In the end I guess it will come down to a gut feeling as to whether it is worth more to hang on to an extra framed rabbit or to give it up and get a dog on a good square around f3.

Thoughts?

Title: Re: Move 28
Post by The_Jeh on Mar 18th, 2008, 12:09am
I completely agree. Rabbit frames are not so bad once the endgame rolls around (something chessandgo told me that he suspected). Our other material equals his, and d3 is a good square for our elephant even without the rabbit. So we might as well keep it - at least maintaining the frame costs chessandgo some resources. He can't advance too much in the west because it loosens the frame on our rabbit, so the swarming advantage is ours.

This is how to talk about these positions. Write generalities that sound nice but that no one can verify.

Title: Re: Move 28
Post by UruramTururam on Mar 18th, 2008, 3:12am

on 03/17/08 at 22:52:05, Fritzlein wrote:
I'm not even sure how to discuss properly in a situation like this.  We can toss out moves for discussion (mine is Ee3w Hd6w Hc6w Ra8s)
[...]
Thoughts?


Well, I agree. The only difference is - I'd advance our dog instead of the rabbit: ee3w hd6w hc6w dd7s

Title: Re: Move 28
Post by Soter on Mar 18th, 2008, 4:43am
I can't give any specific moves now but I agree that our R is worth saving. I'm a bit afraid of chessandgo's E running free on the board while ours is pinned on d3, but such position won't be stable, and we'll be potent enough to generate threats. So the frame looks not so terrifying.

Title: Re: Move 28
Post by jdb on Mar 18th, 2008, 9:42am
In positions with this many pieces exchanged, attacking is generally the correct plan.

It looks like Gold will attack the c6 trap. He already has three heavy pieces on the 4th rank ready to go. If silver attacks the c3 trap, gold's attack on the c6 trap will arrive first. Gold is currently 9 steps from having 3 heavy pieces touching the c6 trap. Silver is more than that for the c3 trap.

This leaves attacking the f3 trap. The dog can arrive at g3 in 5 steps. If gold crosses with the elephant to defend, silver's horse is in a much stronger position.

I suggest
28b d->f3


Title: Re: Move 28
Post by RonWeasley on Mar 18th, 2008, 10:43am
I like h->b6 because it slows any attack on c6.  I like e->d3 because it preserves our extra rabbit and covers any threat of gold trapping pieces on c3, like the horse I want on b6.  With our final step, I'd like to start our e-d attack right away with d->f6.  On the next move, I'm hoping to put  that dog on g3, where it will be hard for the gold horse to get to it without getting caught crossing.  Our e on d3 can still contribute to threats against pieces around f3.  Then we send rabbits to that area.  If the gold E crosses we have time to protect our advanced rabbit by h->c4 followed by the d7 dog.

So 28b ee3w hd6w hc6w df7s

Title: Re: Move 28
Post by mistre on Mar 18th, 2008, 1:23pm
I am going to go against the general consensus who thinks it is worth keeping the rabbit alive in a frame.  My reasoning is that Chessandgo has the west very well defended and our highest secondary piece cannot beat his highest secondary piece.  We will find it very difficult to break the frame and we probably have to give up the rabbit at a later time anyways.

While there is no immediate gain for attacking the f3 trap, I am inclined to go after it because it is under-defended.  

My recommended move is:

28b ee3s ee2e hd6w hc6w

Title: Re: Move 28
Post by jdb on Mar 18th, 2008, 3:26pm
If silver moves the horse to b6, how much does it slow down gold's attack on c6?  If I count correctly, it would then take gold 11 steps to get 3 pieces touching the c6 trap, instead of the original 9 steps.  Silver used two steps to reposition the horse, so it seems about equal.

Also, silver's horse is now really a camel. We took great pains to keep the camel in the centre at the start of the game. Why would we voluntarily move it to the wing now?

Title: Re: Move 28
Post by Fritzlein on Mar 18th, 2008, 6:17pm

on 03/18/08 at 15:26:45, jdb wrote:
Also, silver's horse is now really a camel. We took great pains to keep the camel in the centre at the start of the game. Why would we voluntarily move it to the wing now?

One keeps the deputy (the camel usually, but now the horse) in the center for defense and on the wing to attack.  In the current position we would move the horse to b6 in preparation for getting it to b3.

Title: Re: Move 28
Post by Fritzlein on Mar 18th, 2008, 6:37pm

on 03/18/08 at 09:42:04, jdb wrote:
It looks like Gold will attack the c6 trap. He already has three heavy pieces on the 4th rank ready to go. If silver attacks the c3 trap, gold's attack on the c6 trap will arrive first. Gold is currently 9 steps from having 3 heavy pieces touching the c6 trap. Silver is more than that for the c3 trap.

It seems your count of 9 steps doesn't include any for framing our rabbit.  If he advances all his heavy pieces without first taking four steps to frame our rabbit with his rabbit on b3, then we can start capturing in c3 just with our elephant and rabbit.  Furthermore we will be swarming forward with our own pieces, which will slow him down.  If he has a rabbit on b3, then he can't afford to let any of our pieces break through to b3.  Yes, he already has three pieces on his fourth rank, but we have two on our sixth rank!  :)

That said, I don't have much experience with what happens when two players try to attack through each other.  It's possible that Gold could gain control of c6 without losing control of c3, and also possible that he could trade control of his home trap for ours in such a way that he captures more in c6 than he loses in c3.  I really don't know.


Quote:
This leaves attacking the f3 trap. The dog can arrive at g3 in 5 steps. If gold crosses with the elephant to defend, silver's horse is in a much stronger position.

I suggest
28b d->f3

I have nothing against the dog charge, except that it loses our rabbit.  Moving the dog to f3 does guarantee our dog a strong square, which could be worth more than a framed rabbit for all I know.

Another possibility is 28b ee3w df7s df6s df5e.  If we try to defend our advanced rabbit we still want to be active in the east as well as the west, and I'm not sure how to divide up the steps.

Title: Re: Move 28
Post by jdb on Mar 18th, 2008, 9:46pm

Quote:
That said, I don't have much experience with what happens when two players try to attack through each other.  It's possible that Gold could gain control of c6 without losing control of c3, and also possible that he could trade control of his home trap for ours in such a way that he captures more in c6 than he loses in c3.  I really don't know.


This is what concerns me about focusing the play on the c3/c6 trap. Gold has 9 pieces on the east side(all from gold's POV) and 4 pieces on the west side. Silver's elephant is well placed, so it can count on both sides of the board. This gives silver 8 pieces in the east and 7 pieces in the west.

So assuming that both sides make effective use of all their pieces, gold currently has the material advantage in the east and silver has the material advantage in the west.

I don't know how effective silver's attack on c3 will be when we already have a material deficit on that side of the board.

Title: Re: Move 28
Post by Ice on Mar 18th, 2008, 11:32pm
I might as well put my 2 cents in. btw this is a great idea to have a forum discussion on moves. I love it. I haven't played in about a year so keep in mind that all of my analysis probably has a lot of rust to it.  :)

My feeling is that soon after the rabbit frame we would have to give it up or let chessandgo get an EH on c6.

Personally if I was chessandgo I would bring EH up to c6 as quickly as I could if our E was stuck in the rabbit frame. If he can get the silver horse to c6 and starts advancing gold rabbits, that portion of the board might be in trouble.

That being said moving ee3w costs us one step and delays the advance on c6 by the dog by at least 2 moves so starting out with ee3w that at least makes chessandgo take the time to make the frame might be worth it.

Just out of curiousity what happens if we play with moving our horse up towards the f3 trap with the dog?



Title: Re: Move 28
Post by arimaa_master on Mar 19th, 2008, 8:03am
My feeling about position is that if we keep our rabbit (ee3w) and fill in the c3 trap (hd6w dd7s) - so gold has nothing to achieve and we our safe - so we have time to start to play d+c+r attack at f3 trap (ce7s).


So my proposal is: 28b ee3w hd6w dd7s ce7s

(But I am not saying that already proposed moves is worse or better in general - that is beyond my horizon :)).

Title: Re: Move 28
Post by UruramTururam on Mar 19th, 2008, 5:55pm
Another idea:

28b: ee3s ee2e Cg2n ef2e

If C&G decides to trap our rabbit then he has only 2 more steps. If he attacks our h, the cat may be trapped in f3 if he secures f3 we may try to push the cat towards f6...
But somehow I don't like the resulting position with our elephant so decentralized.

Title: Re: Move 28
Post by RonWeasley on Mar 21st, 2008, 6:44am
Let's plan on voting starting Monday morning.

I still like 28b ee3w hd6w hc6w df7s.  An e-d attack against f3 is supported by our h around c6 to defend against a counter attack by gold's H.  If gold tries E-H against c6, we advance a rabbit through f3 for a goal threat that can't be ignored.

Title: Re: Move 28
Post by RonWeasley on Mar 24th, 2008, 7:50am
One of the candidate moves

ee3w hd6w hc6w ee2e (e->c3, h->b6) should read

ee3s hd6w hc6w ee2e (e->c2, h->b6).

Sorry for the confusion.

Title: Re: Move 28
Post by Fritzlein on Mar 24th, 2008, 10:14am

on 03/24/08 at 07:50:38, RonWeasley wrote:
One of the candidate moves

ee3w hd6w hc6w ee2e (e->c3, h->b6) should read

ee3s hd6w hc6w ee2e (e->c2, h->b6).

Sorry for the confusion.


Do you mean ee3s hd6w hc6w ee2e (e->f2, h->b6)?

Title: Re: Move 28
Post by RonWeasley on Mar 24th, 2008, 10:31am

on 03/24/08 at 10:14:04, Fritzlein wrote:
Do you mean ee3s hd6w hc6w ee2e (e->f2, h->b6)?



Yesh (hic!).

Title: Re: Move 28
Post by UruramTururam on Mar 25th, 2008, 9:18am
Uffff... It was hard to choose the order this time.  :o

Title: Re: Move 28
Post by Fritzlein on Mar 25th, 2008, 9:55am
It is especially hard because we don't even know what the top two contenders are.  Sometimes one can assume that one of two moves is surely going to win, so the only thing that really matters in a ballot is the relative order of those two.  Other votes will at most affect the order of the losers.  But this time, since I didn't know which choices were going to be near the top, it seemed like every position in the ranking could be meaningful.

Of course, that said, I just voted my gut feeling anyway...

Title: Re: Move 28
Post by RonWeasley on Mar 25th, 2008, 1:39pm
What an unruly mob!  We have a tie.

ee3w hd6w hc6w df7s

and

df7s df6s df5s df4s

I am not at the right computer to start the run-off, so expect the run-off to begin Wednesday morning and end Thursday morning.  The forum is open for all who wish to lobby for either move.  If the run-off results in a tie, I will flip a galleon (if I can get Harry to lend me one).

Title: Re: Move 28
Post by Fritzlein on Mar 25th, 2008, 3:04pm
Wow, before the beatpath tiebreaker was applied, ee3w df7s df6s df5e was in a circular three-way tie with the other top two options.  And not only was there a three-way tie, but also for no pair of moves was the voting unanimous.  There are no zeros in the preference matrix.  That hasn't happened since move 19 when there were only four choices on the ballot.  There is nothing the unruly Mob agrees on, not even that the "best" proposal is better than the "worst" proposal.  I think a coin flip is entirely appropriate to decide the election and save reserve time.

I'm not sure I like having a two-way runoff.  If we narrow it to two options, then saving the rabbit should beat the dog charge since it won 5-4 in the first election.  As long as we are going to burn the reserve necessary to have a runoff, I would actually prefer it to be a three-way runoff, for fear that we will now chose a move that was beaten in the first round by a move that got left off the second ballot.

One might argue that a three-way runoff will just result in another circular preference, but it won't necessarily.  People now have time to change their minds and/or to lobby with renewed vigor.  Most importantly, people who didn't vote the first time can jump in and alter the outcome.  Note that the preference of any ONE additional voter could have decided the outcome in the first round, so greater participation is in itself quite likely to clear up the mess.  But of course I bow to the Mob Coordinator's decision if he wants a two-way runoff in preference to a coin flip or a three-way runoff.


   .    .    .    .    .    .    .    .    .    .  1. 2. 3. 4. 5. 6. 7. 8.
1. ee3w hd6w hc6w df7s (e->d3, h->b6, d->f6)     .   -  5  4  6  7  6  7  8
2. df7s df6s df5s df4s (d->f3)    .    .    .    .   4  -  6  5  6  6  7  6
3. ee3w df7s df6s df5e (e->c3, d->g5)  .    .    .   5  3  -  5  5  5  7  6
4. ee3w hd6w hc6w dd7s (e->d3, h->b6, d->d6)     .   2  4  4  -  6  7  7  7
5. ee3w hd6w hc6w ra8s (e->d3, h->b6, r->a7)     .   2  4  4  2  -  6  6  6
6. ee3w hd6w dd7s ce7s (e->d3, h->c6, d->d6, c->e6)  2  3  4  1  3  -  6  7
7. ee3s ee2e Cg2n ef2e (e->g2, C->g3)  .    .    .   2  2  1  2  3  3  -  5
8. ee3w hd6w hc6w ee2e (e->f2, h->b6)  .    .    .   1  3  2  2  3  2  3  -

Title: Re: Move 28
Post by UruramTururam on Mar 25th, 2008, 5:17pm
I'd opt to choose the first move because it won the direct competition with the second one, but I think we should let Ron decide.

Title: Re: Move 28
Post by Fritzlein on Mar 25th, 2008, 5:42pm

on 03/25/08 at 17:17:09, UruramTururam wrote:
I'd opt to choose the first move because it won the direct competition with the second one, but I think we should let Ron decide.

Also using the dumb tiebreaker of summing all the numbers in the row gives the first option a slight win, so it wouldn't be too upsetting if Ron made an executive decision in its favor, in spite of the "Mob Rules" not specifying any such thing.  (Can there be rules to a Mob, or only rule by a Mob?)

I guess my main point was that if we have a revote, we might as well include the top three.  It doesn't take any more time off our clock and it gives all the top options a chance.  And now that I think about it, with our reserve fully replenished, a re-vote seems like a much better option than an executive decision.  This is not a time scramble situation.  Let's let the Mob have its say even if it takes some time to figure out what it is saying.

Title: Re: Move 28
Post by mistre on Mar 25th, 2008, 11:06pm
Definitely a revote.  Of the top 2 moves, I have changed my mind on which I like best.

Title: Re: Move 28
Post by RonWeasley on Mar 26th, 2008, 6:44am
Wow, I didn't notice we were having a three-way.  That's really humiliating to somebody.

Anyway, I just looked at the result rankings and saw a tie.  We will run-off with the top three.

Title: Re: Move 28
Post by woh on Mar 26th, 2008, 7:32am
With the results the way they are, I would put option 2 on top.


on 03/25/08 at 17:42:58, Fritzlein wrote:
Also using the dumb tiebreaker of summing all the numbers in the row ...

If you use this between the results of the top three only then the second choise is the winner.

Using Schulze between the top 3 also favors the second option.

Title: Re: Move 28
Post by aaaa on Mar 26th, 2008, 8:12am
I would like to point out that Schulze is not supposed to be run again with candidates eliminated as it would violate the monotonicity criterion, unless the eliminated candidates are all outside the Smith set.

So I would advise a revote between all members of the Smith set.
Also, if you're going to break a tie randomly, do it by repeatedly choosing a random ballot before picking a random candidate.

Title: Re: Move 28
Post by RonWeasley on Mar 26th, 2008, 11:15am
Fortunately the top three candidates constitute the Smith set for this election.

For TheMob members like me who haven't looked this up before, it's the smallest set of candidates that are preferred over every candidate outside the Smith set.  In theory it's possible for all candidates to be in this set if you are liberal with my definition.  One method for resolving ties would be to have run offs among the candidates of the incrementally decreasing Smith sets until there is a clear winner or no candidates fail to make the Smith set.  That would take about an extra day per run off, which seems worth it to me.

Unfortunately we don't have access to the ballots so we have to revert to a different methodology for breaking a tie having an all-inclusive Smith set.  A random outcome generator is the fastest means available and I propose choosing from only the "tied" candidates given by the voting site (instead of the entire list).

Title: Re: Move 28
Post by Fritzlein on Mar 26th, 2008, 12:08pm

on 03/25/08 at 23:06:54, mistre wrote:
Definitely a revote.  Of the top 2 moves, I have changed my mind on which I like best.

I actually reversed my preference among these three votes, swapping the top and bottom.  In my case a re-vote is essentially equivalent to randomizing a ballot.  :P

At least the three options are close enough in quality (as far as I know) that I will be happy no matter which one wins.

Title: Re: Move 28
Post by RonWeasley on Mar 27th, 2008, 7:16am
And our winner is ee3w df7s df6s df5e!  A close one.  Thanx for voting.

Title: Re: Move 28
Post by Fritzlein on Mar 27th, 2008, 7:26pm
It is impossible to tell without individual ballot reporting turned on, but there is a fair chance the winning option didn't get 6 of 11 first place votes.  It probably won by virtue of being a compromise option between saving the rabbit and attacking with the dog.

   .    .    .    .    .    .    .    .    .  1. 2. 3.
1. ee3w df7s df6s df5e (e->d3, d->g5)  .    .   -  7  6
2. df7s df6s df5s df4s (d->f3)    .    .    .   4  -  6
3. ee3w hd6w hc6w df7s (e->d3, h->b6, d->f6)    5  5  -


Title: Re: Move 28
Post by aaaa on Mar 27th, 2008, 9:04pm
What does Bomb say? Does the software allow you to determine which of those three moves it prefers?

Title: Re: Move 28
Post by Fritzlein on Mar 28th, 2008, 7:58am

on 03/27/08 at 21:04:17, aaaa wrote:
What does Bomb say? Does the software allow you to determine which of those three moves it prefers?

At 16 steps depth (4 ply) Bomb would have preferred to play ee3s ee2e Cg2n ef2e, with an evaluation of +0.75 for Silver.

There is no way to tell which of the three moves we voted among Bomb would most prefer except by trying out each of them and letting Bomb think to an equal depth in each case.  If we let Bomb think to a depth of 12 steps (3 ply) in each case, its evaluation should be comparable to the evaluation of its preferred move, so we can tell how much of a mistake Bomb thinks we are making.

+0.14 ee3w df7s df6s df5e
+0.11 ee3w hd6w hc6w df7s
+0.01 df7s df6s df5s df4s

So at that particular depth, Bomb likes the move we chose the best of the three moves in the runoff, but thinks we made an error of magnitude 0.61, more than half a rabbit, by rejecting ee3s ee2e Cg2n ef2e in the first ballot.  We had Bomb's preferred move in seventh place of our eight options.

I was about to make the obligatory comments about Bomb having no clue in general, and particularly no clue in a quiet position like this one, but to be fair, we have no clue either.  I would guess that giving up our rabbit and playing for a cat hostage is the wrong idea in a position like this, because it invites chessandgo to regroup and eventually swarm.  I would suspect that with MHR gone from each side, we'd rather be rolling forward than pushing a hostage to our side.  But who knows?  My own understanding of the position is too weak to poke fun at Bomb too vigorously in this circumstance.

Title: Re: Move 28
Post by Fritzlein on Mar 28th, 2008, 8:18am
Note that Bomb's decision to go after the cat hostage is strategic, not tactical.  After the move sequence

28b ee3s ee2e Cg2n ef2e
29w rb3e rc3x Db4s Ec4e Rf1n
29b Cg3n eg2n Cg4n eg3n
30w Rg1n Ed4e Ee4n Ee5n
30b hd6w hc6w eg4w ef4n
31w Dd2e De2n De3e Df3e
31b df7s rg6e df6e rh6s

Bomb thinks to depth 12 steps and concludes that Silver is still ahead by +0.68.  It wasn't just that Bomb was pushing the loss of Silver's rabbit over the horizon, or seeing some tactical compensation.  Bomb really likes that cat hostage strategically, which I don't quite agree with in a position like this.



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