Author |
Topic: Move 5 (Read 5697 times) |
|
Fritzlein
Forum Guru
Arimaa player #706
Gender: 
Posts: 5928
|
 |
Re: Move 5
« Reply #15 on: Jun 19th, 2007, 3:50pm » |
 Quote  Modify
|
I assume that the lack of folks contradicting my latest analysis is due to the Arimaa server and/or forum being down for a good portion of the last few days. Of course, if everyone agrees with me, we can vote right away, but if we aren't unanimous, I am amenable to a delay. We have eaten into our reserve somewhat, but not as much as chessandgo has eaten in to his!
|
|
 IP Logged |
|
|
|
arimaa_master
Forum Guru
Arimaa player #2010
Gender:
Posts: 358
|
|
Re: Move 5
« Reply #16 on: Jun 19th, 2007, 11:04pm » |
Quote Modify
|
on Jun 19th, 2007, 3:50pm, Fritzlein wrote:| I assume that the lack of folks contradicting my latest analysis is due to the Arimaa server and/or forum being down for a good portion of the last few days. Of course, if everyone agrees with me, we can vote right away, but if we aren't unanimous, I am amenable to a delay. We have eaten into our reserve somewhat, but not as much as chessandgo has eaten in to his! |
|
Yes, I am for postponing too.
|
|
IP Logged |
|
|
|
99of9
Forum Guru
Gnobby's creator (player #314)
 
Gender:
Posts: 1413
|
|
Re: Move 5
« Reply #17 on: Jun 20th, 2007, 3:38am » |
Quote Modify
|
I don't have much to contribute here, Fritz's lines seem fine. Is there a reason not to push the camel first? It might give us a little longer to look at the sharp lines associated with pushing his camel onto c6 on turn 6b.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 5
« Reply #18 on: Jun 20th, 2007, 8:24pm » |
Quote Modify
|
on Jun 20th, 2007, 3:38am, 99of9 wrote:| Is there a reason not to push the camel first? It might give us a little longer to look at the sharp lines associated with pushing his camel onto c6 on turn 6b. |
|
Ahhh, good point. If we are going to answer
5b Md4e ec4e Me4n ed4e
6w Ed6e Me5w Md5w Ee6w
with
6b ee4w ed4w ec4w eb4n
then we should instead immediately play
5b ec4n Md4w ec5w Mc4n
to eliminate the possibility that chessandgo will surprise us with a better move 6w. On the other hand, if we are going to answer
5b Md4e ec4e Me4n ed4e
6w Ed6e Me5w Md5w Ee6w
with
6b ee4w ed4w Mc5n ec4n
then we have a reason to prefer this line. It's definitely worth considering. My first thought was that he could ignore the threat and pull a rabbit, but now I'm not so sure. For example:
5b Md4e ec4e Me4n ed4e
6w Ed6e Me5w Md5w Ee6w
6b ee4w ed4w Mc5n ec4n
7w cc7e Mc6n Mc7s rc8s
7b rb7w rc7w Mc6n ed5n
or
5b Md4e ec4e Me4n ed4e
6w Ed6e Me5w Md5w Ee6w
6b ee4w ed4w Mc5n ec4n
7w hb6w Mc6w Mb6s rb7s
7b rb6e Mb5n ec5w rc6s
and either way we have a camel hostage. So maybe his camel would just have to run.
But on the other other hand, those camel hostages are both imperfect, in the first case because our elephant is behind the trap and in the second case because our horse can be blocked into the a-file by a swarm, much reducing the value of the camel hostage.
So maybe 99of9 is exactly right and we should buy time to analyze the sharp lines. Is there any answer to
5b Md4e ec4e Me4n ed4e
better than
6w Ed6e Me5w Md5w Ee6w
that would deter us from this 5b? Arimaa_master's suggestion seems like the only plausible alternative for chessandgo, but after
5b Md4e ec4e Me4n ed4e
6w Ed6e Me5w Md5w Mc5w
6b ee4w ed4w ec4w me7w
I think we stand better. So I have talked myself back into favoring 99of9's move unless someone can come up with a good line for chessandgo from the critical position after the last-mentioned line.
P.S. I may have to retract my criticism of our camel pull on 4b. It seems to offer us more play than I realized.
P.P.S. If the discussion doesn't pick up a little, there's no point in delaying the vote, because we'll still have to vote a couple of days later, with nothing to show for the delay but a reduced reserve. Whoever wants more time should raise some issues and questions.
|
| « Last Edit: Jun 20th, 2007, 8:43pm by Fritzlein » |
IP Logged |
|
|
|
The_Jeh
Forum Guru
Arimaa player #634
Gender:
Posts: 460
|
|
Re: Move 5
« Reply #19 on: Jun 21st, 2007, 12:19am » |
Quote Modify
|
What if after 5b Md4e ec4e Me4n ed4e Chessandgo plays
6w Ed6e Me5w Ee6w me7s
If we try to push his camel back towards the f6 trap, he can then push our camel onto it, which takes away our camel's influence on the west side of the board. I don't know -- just thought I'd venture something...
We shouldn't be fooled by Chessandgo's low reserve time. He is well aware that his ability to replenish it is much greater than ours.
|
| « Last Edit: Jun 21st, 2007, 12:20am by The_Jeh » |
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 5
« Reply #20 on: Jun 21st, 2007, 8:20am » |
Quote Modify
|
on Jun 21st, 2007, 12:19am, The_Jeh wrote:What if after 5b Md4e ec4e Me4n ed4e Chessandgo plays
6w Ed6e Me5w Ee6w me7s |
|
Nice find! That is a strong 6w for chessandgo that I hadn't even considered. His camel on d5 is actually safer than it would be on c5. Meanwhile, for him to drag our camel forward costs us a lot of time in getting our camel to the west where it wants to be. This is a big improvement over what I thought was his best move.
The best continuation I see for us would be 6b ee4n me6n ee5s Md5e, which is useless due to repetition. Other moves for us seem to end up in chessandgo pulling our a8 rabbit for no compensation, at least when I play them out.
That makes me flip-flop (again!) to a clear preference for dragging his camel with 5b ec4n Md4w ec5w Mc4n.
|
|
IP Logged |
|
|
|
RonWeasley
Forum Moderator
Forum Guru
Harry's friend (Arimaa player #441)
Gender:
Posts: 882
|
|
Re: Move 5
« Reply #21 on: Jun 21st, 2007, 2:00pm » |
Quote Modify
|
I just looked at variations after 5b ec4n Md4w ec5w Mc4n where the M goes to c6 and a D goes to c5 or d5. Mainly I was worried about getting our c6 flooded. However, with our e so close to the M and not needed elsewhere, it seems like we can eventually get the M hostage. A reply I used is 6b rb7w cc7w me7w md7w. The forward D becomes vulnerable if the M tries to escape.
So I like this move best and I would be ready to vote. Vote tomorrow?
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 5
« Reply #22 on: Jun 22nd, 2007, 12:36am » |
Quote Modify
|
on Jun 21st, 2007, 2:00pm, RonWeasley wrote:
Yes, let's vote. New analysis could still change my mind, but I'm comfortable voting on what we know so far. Furthermore, it doesn't look like new analysis is flooding in. Furtherfurthermore, we will burn at least a day of reserve even if we vote right away, probably closer to two days.
|
|
IP Logged |
|
|
|
99of9
Forum Guru
Gnobby's creator (player #314)
Gender:
Posts: 1413
|
|
Re: Move 5
« Reply #23 on: Jun 22nd, 2007, 1:11am » |
Quote Modify
|
yes, let's vote
|
|
IP Logged |
|
|
|
RonWeasley
Forum Moderator
Forum Guru
Harry's friend (Arimaa player #441)
Gender:
Posts: 882
|
|
Re: Move 5
« Reply #24 on: Jun 22nd, 2007, 7:11am » |
Quote Modify
|
The only way to set up the vote this time is at my work computer which I won't access over the weekend. So I can end the vote and move on 5pm Friday, or I have to wait until Monday morning. So if we get enough votes (10) by COB Friday, I will use that move. Otherwise I will move Monday and TheMob will lose that time off the clock.
It's really not so bad if we lose that time. I think we will make it up on exchange moves in the future. But the discussion appears to favor a single move, so today's vote appears to be easy.
Please post something here today if you don't like my plan for ending the vote! I will monitor this forum and we can always negotiate.
I personally didn't look hard at the 5b ec4w Db3e eb4s eb3w. It looks viable, but I didn't like leaving the gold camel by itself in the center and our e off to the side. This is just my personal preference not based on any analysis because ... I'm too lazy!
|
|
IP Logged |
|
|
|
99of9
Forum Guru
Gnobby's creator (player #314)
Gender:
Posts: 1413
|
|
Re: Move 5
« Reply #25 on: Jun 22nd, 2007, 7:29am » |
Quote Modify
|
on Jun 22nd, 2007, 7:11am, RonWeasley wrote:| I personally didn't look hard at the 5b ec4w Db3e eb4s eb3w. It looks viable, but I didn't like leaving the gold camel by itself in the center and our e off to the side. This is just my personal preference not based on any analysis because ... I'm too lazy! |
|
I looked at it a bit, and you're right, the camel becomes too dangerous. I'm pretty convinced that we should definitely go for the camel pull.
|
|
IP Logged |
|
|
|
RonWeasley
Forum Moderator
Forum Guru
Harry's friend (Arimaa player #441)
Gender:
Posts: 882
|
|
Re: Move 5
« Reply #26 on: Jun 22nd, 2007, 3:54pm » |
Quote Modify
|
Well, it's about 5pm and we have only 7 votes. I wanted 10 to close the election early, so I will extend the time until Monday morning. Who knows, maybe we'll still have 7 votes. It was worth a try, but not everybody reads their e-mail every day. Thank you to those who voted early.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 5
« Reply #27 on: Jun 23rd, 2007, 10:29am » |
Quote Modify
|
on Jun 22nd, 2007, 7:11am, RonWeasley wrote:| It's really not so bad if we lose that time. I think we will make it up on exchange moves in the future. |
|
That brings to mind a discussion I had with MrBrain before this year's World Championship. He said he left Arimaa due to frustrations over the time control. Whenever he lost he was frustrated because he felt he wouldn't have made the game-losing mistake if only he had had a bit more time to think about the move.
Partly he just wanted more time for every move, which is understandable. Wanting to play slow or fast is mostly a matter of taste. But partly (and this was the more interesting part of our discussion) he was annoyed not to have full control over time management. If the time control were two minutes per move, he would like to be able to spend twenty minutes on a single move, perhaps having saved up by using only one minute per move on the previous twenty moves. The idea is that the outcome of a game usually hinges on a small number of critical decisions.
My contention is that Arimaa has very few non-critical moves. In order to feel comfortable moving quickly, it must be true either that the best move isn't much different than the second best move, or that the best move is very easy to find. It seems to me that in Arimaa the best move is usually hard to find, and it usually does make a difference.
Admittedly, in some positions making a sub-optimal moves merely costs a tempo, whereas in other positions making a sub-optimal move costs a whole piece, or a goal. It makes sense to me to think longer in the latter, sharp positions. But you can't dismiss the tempo-losing moves as unimportant, because if you let your opponent accumulate tempos, then you will land in a position where you will eventually lose material even if you find the best move every time (assuming your opponent doesn't blunder).
Furthermore, given that, in Arimaa, captures can usually be postponed by the defender, positions often stay sharp for ten or twenty consecutive moves. For example, when one side is holding a camel hostage, the question is usually on what move the defender will abandon the camel to its fate in order to gain material elsewhere. Given that the defending elephant is constantly threatening to leave, the position is sharp on every single move until it does leave.
To make matters worse, the game tends to get sharper and sharper as pieces are exchanged, which loosens up the board and brings goal threats into play. Sometimes, after twenty moves of tension resolve in a flurry of captures, the game transitions directly into a tense endgame where every step counts. This means that for most moves the difference between the best and second-best move is substantial.
In this game, I feel the only time there wasn't much at stake between the best and second-best move was on our opening setup when we were deciding between dogs behind the traps and cats behind the traps. That difference hardly mattered, but every move since has had a tangible difference between the top two. If the difference wasn't in tempo, it was in big strategy decisions about what type of position we would like to have.
If you believe, like me, that there is almost always an important difference between the right move and the not-quite-right move, then then only time you can comfortably spend less time on a move is when it is obvious which move is the right move. The only such situation that occurs to me is forced recaptures. And even in capture situations there can sometimes be a question of which side of the trap to end on, or in which order to capture threatened pieces.
I will be very interested to count, over the course of this game, the number of times the Mob is able to make a snap decision. I predict that there will only be two moves from here forward in which either it is immediately obvious which move is best, or in which the difference between the top moves is insignificant.
Moves like the present certainly must count as critical moves. Whether to push or flip or drag the camel (not to mention counter-attacking) has a significant impact. Even if we all end up agreeing that one move is better than the others, it isn't superficially obvious which. If The_Jeh hadn't found a good move for chessandgo at the last minute, I wouldn't have favored the better move, so this was a move where we definitely needed our thinking time.
When will the next move come where we don't need our thinking time? I urge our fearless leader to be on the lookout for both types (i.e. we all agree, or the disagreement doesn't matter), but I don't really anticipate either.
|
| « Last Edit: Jun 25th, 2007, 8:05am by Fritzlein » |
IP Logged |
|
|
|
RonWeasley
Forum Moderator
Forum Guru
Harry's friend (Arimaa player #441)
Gender:
Posts: 882
|
|
Re: Move 5
« Reply #28 on: Jun 25th, 2007, 7:32am » |
Quote Modify
|
The winning move is 5b ec4n Md4w ec5w Mc4n. 11 out of 15 voted which means 4 votes came in over the weekend.
Welcome to mdk who just registered to TheMob.
Note that C&G has only 3+ days to move, so it will happen soon. I suggest we try to move in 6 days for a while to replenish our reserve and to see if we can move that fast. Remember that this previous move took extra time due to my absence from the right computer and the Forum down time. I think my posting a schedule has helped and I'll keep doing that.
For the past few moves, we've had 11 votes. Does this mean we've had only 11 active mobsters (now there are 12)? Mobsters going inactive is inevitable, but this has a bearing on how many votes constitutes a quorum if we are short on time. I propose that under ordinary circumstances, if we get 8 (instead of 10) or more votes by the stated deadline, I will end the vote and move. This can, of course, evolve as the number of active mobsters changes.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 5
« Reply #29 on: Jun 25th, 2007, 8:00am » |
Quote Modify
|
For the record:
. . . . . . . . . . 1. 2. 3. 4.
1. ec4n Md4w ec5w Mc4n (pull camel and freeze horse) - 9 9 10
2. Md4e ec4e Me4n ed4e (push camel to e5) . . 2 - 8 5
3. ec4w Md4w Mc4n eb4e (camel flip to c5) . . 2 2 - 5
4. ec4w Db3e eb4s eb3w (e attack west) . . . 1 3 4 -
That makes it our clearest decision yet. One possibility for deciding whether ten votes or eight is enough to close the voting after one day would be the amount of separation among the top vote-getters.
Predictions: We will some day have a unanimous first choice. It will be the move on which we goal. This will be move sixty-seven. (It's amazing what you can learn with a magic 8-ball and binary search.  )
|
|
IP Logged |
|
|
|
|
Home
Help
Search
Members
Login
Register
Reply
Notify of replies
Send Topic
Print