Author |
Topic: Fritz's commentary (Read 13043 times) |
|
Fritzlein
Forum Guru
Arimaa player #706
Gender: 
Posts: 5928
|
 |
37s
« Reply #75 on: Jun 1st, 2010, 7:57am » |
 Quote  Modify
|
37s
The Mob played 37g EvDbC<Rh^, which was at least a move that I had considered. That’s a step up from my performance the move before. On the other hand, the fact that the Mob moved in only two days after having earlier reduced their reserve to a bare minimum suggests that they had anticipated my move and my analysis was flawed. I thought 37s Rb8>vRg>H^ would be an adequate counter. Obviously, that’s the first move I will examine in my present analysis.
…
My intended move turns out to suck. I believe it gets me a dog trade, but then the Mob has a vastly superior position, i.e. an advantage of more than just a cat. The situation is, however, extremely sharp. I can initiate a goal race with 37s Rg>H^ER^, and perhaps the only way the Mob can beat my goal attack in the southeast is by racing me in the northwest. That’s a situation where tactics rule, so I will have to roll up my sleeves and analyze.
…
After several brain-burning hours, my analysis tree is showing some lines where I win the opposite-corner goal race. It may be that the Mob must not take my dog in f6, racing for goal, but rather take my other dog in f3, leaving their elephant available for home defense. Could it possibly be that the Mob only calculated far enough to see material gain, and miscalculated the goal race? I shouldn’t even hope for a tactical blunder by the Mob, with their many eyeballs and computer assistance, but they did move very quickly, which could mean they were relying on analysis from their discussion of 36g. That would mean their analysis was two ply shallower from the perspective of 37g, so an error creeps into the realm of possibility. But when I work on these goal races, I severely miss computer help, and I am probably overlooking critical lines. The position is so sharp that one tiny oversight in one line can totally change the evaluation.
…
Further analysis leaves me in doubt as to whether the Mob has a winning dog capture if I play the southeast goal attack. If the race is double-edged, as opposed to being clearly in my favor, I should expect that the Mob will outplay me tactically, so I should avoid the line merely because it is unclear. On the other hand, I see no alternative other than jumping into the minefield. There is no longer any way for me to quietly play for control and hope to come back gradually.
The position is too complicated for me to analyze, so just maybe there is at least a tiny chance that the Mob will also have an analysis failure, and I will win by sheer luck. If you had asked me a few moves ago whether I would be willing to gamble everything on a super-sharp unclear goal race, of course I would have accepted it rather than a tortuous loss. I am not going to cry now for lack of another option. I will play 37s Rg>H^ER^ and be glad it gives even the appearance of hope.
|
|
 IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
38g
« Reply #76 on: Jun 1st, 2010, 7:57am » |
Quote Modify
|
38g
The obvious direct try for the Mob to win is 38g DbR<EDc^*. I can’t recapture a dog in f6 without losing more material or a goal in short order. But, although I am two pieces down, I can race with 38s EH<Rh4vRb>, and apparently win because my lonely d4-dog is providing just enough defense to slow down the c4-rabbit and just enough offense to help my own attack through in some lines. Yes, there is a high probability my tactical calculation was faulty, in which case the Mob will play this move for a forced win, but I can’t predict a move without seeing why it works, can I?
A less dangerous move for the Mob is 38g EDv<*, capturing in f3 to bring their elephant home on defense. This stops my goal attack, but leaves the Mob’s c4-rabbit, g5-rabbit, and g7-dog all in jeopardy. It appears that I can at least get a dog back, plus a rabbit, after tying the Mob’s elephant down on defense. On the other hand, the Mob could get a rollicking goal attack started up the middle while I was trying to catch up materially. From my analysis, I suspect I am objectively lost after this move. Given that, and given the way the Mob loves to attack, this is the move that I will predict.
Apart from taking one of my dogs, the Mob can’t really improve its western position; two steps would be useless. Therefore their other alternative as far as I can see is to use all four steps on defense, say with 38g Rf>H>Rd>>, to stop my attack before it gets rolling. But this and similar moves allow 38s EvRh6vRa7>>, leaving only my d4-dog for capture, after which I get at least two rabbits and a good position in return, if not a dog and a rabbit outright. Therefore I expect the Mob to be forced into a sharp race whether they like it or not.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
38s
« Reply #77 on: Jun 1st, 2010, 7:58am » |
Quote Modify
|
38s
38g DbR<EDc^* has hit the board. For the second move in a row, the Mob quickly played a move that I had considered but declared unsound. This time I saw JDB post at least twice to the discussion thread, so I know the Mob has had computer help. All signs point to the conclusion that my analysis was flawed, I am dead meat, and all that remains is for me to figure out why 38s EH<Rh4vRb>, which I considered a refutation of the Mob’s move, is actually a forced win for them.
…
Now I see it. I had been focused on the Mob directly racing with its c4-rabbit, or on coming home with its elephant to defend, or on trying to defend without its elephant. None of these plans work as far as I can see. But I failed to account for the power of the Mob’s g5-rabbit. With 39g H^>Rg< plus another step, the Mob sacrifices the rabbit, but buys enough time that they will win the goal race instead of me.
It is in no way surprising that it would be me rather than the Mob overlooking a tiny but decisive tactical point. That is their strength. Oh, well, I hope the game was at least interesting to them.
I puzzled over the Mob’s move for my own curiosity, not because it can help me now. I committed myself two moves ago. Now I am down a cat and a dog with no hope of recovery. Since I have no good moves any more, continuing the goal race as I originally planned is no longer a last, desperate hope, but rather a graceful way of resigning.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
39g
« Reply #78 on: Jun 1st, 2010, 7:58am » |
Quote Modify
|
39g
If the Mob begins its move with 39g H^>Rg<, a lot of different fourth steps win the game. For the sake of completeness in making a prediction every move, I will predict 39g H^>Rg<E>.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
39s
« Reply #79 on: Jun 1st, 2010, 7:58am » |
Quote Modify
|
39s
The Mob quickly played 39g H^>Rf>E>. I had the idea, but missed the execution by one step. Since they didn’t set up a goal in one, I have a chance to for a last hurrah, capturing their horse and setting up my own goal in one. Yay! Of course all of moves are losing. For the benefit of posterity, I will let the Mob demonstrate how the most obvious of my moves loses. Therefore I will play 39g EH<v*.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
40g
« Reply #80 on: Jun 1st, 2010, 7:59am » |
Quote Modify
|
40g
The Mob must spend at least one step to defend my goal threat with Rg2^. The three steps that get them closest to goal are Cc^Rc^^C*. Therefore I predict 40g Rg2^Cc^Rc^^C*. There may be other winning moves for the Mob, but I don’t need to find them, do I?
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
40s
« Reply #81 on: Jun 1st, 2010, 7:59am » |
Quote Modify
|
40s
A correct prediction of 40g Rg2^Cc^Rc^^C* leaves me 9 of 40 predicting for the game. I can mercifully no longer stop the Mob’s goal threat even if I want to, so there is no harm in ending the game with 40s E>ER<*Ev for a material advantage and an unstoppable goal threat. Good game, Mob!
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 2
« Reply #82 on: Jun 1st, 2010, 8:47am » |
Quote Modify
|
2g
on Dec 19th, 2008, 7:51am, Adanac wrote:| He was joking -- the move that we played is most likely the one that he anticipated. If he was expecting (fearing?) E->e5 if was only because he lost to Chessandgo with that move last week. |
|
Yep, I was joking with my trash talk, but really did feel at the time that E->e5 was best and the Mob's actual move was slightly inaccurate. I didn't properly account for the vulnerability of my a7 rabbit to being pulled.
There's a serious irony here, because I took 2g E^^^Hb^ as a sign that chessandgo was dictating affairs to the Mob, when in fact chessandgo had been persuaded away from his original suggestion and was advocating 2g E^^^>.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 18
« Reply #83 on: Jun 1st, 2010, 11:50am » |
Quote Modify
|
on Aug 31st, 2009, 10:05am, RonWeasley wrote:| Is there a consensus on the rabbit pull and Dd1w or is the consensus still D->b3? |
|
Rats, it looks like the Mob was on the verge of a snap consensus for the much weaker move. Time pressure almost worked in my favor there, but then you guys voted for the stronger move and still gained reserve. Maybe I had slightly more chance in this game than I realized.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 14
« Reply #84 on: Jun 1st, 2010, 11:55am » |
Quote Modify
|
14g
on Jun 11th, 2009, 11:34am, Simon wrote:| My impression now is that we are losing. |
|
Well, I managed to scare Simon and arimaa_master at least. I am happy that you guys at least felt some pressure, although I never thought I was winning at any point after 10s. Perhaps my over-confident trash talk served some purpose of keeping you in doubt. 
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 19
« Reply #85 on: Jun 1st, 2010, 12:00pm » |
Quote Modify
|
19g
on Sep 9th, 2009, 10:06am, jdb wrote:| One question, why did fritz move the horse east? |
|
In hindsight my 18s was a tactical miscalculation. If I had seen in advance how badly the horse trade would work out for me, I would have played differently. However, the invasion of the Mob's western horse was putting me in a serious tempo shortage, so I am not sure I had a good move available. I don't feel nearly as bad about this tactical inaccuracy as I do about 10s.
The Mob's discussion on 20g and its eventual move really showcase the Mob's tactical prowess, and show why it is so hard for the lone player to compete. Even if there has been a long sequence of relatively quiet moves, cashing in on a strategic advantage always requires some sharp bloodletting eventually, and that is when any tactical deficiency can more than undo any strategic gains.
|
| « Last Edit: Jun 1st, 2010, 12:08pm by Fritzlein » |
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 28
« Reply #86 on: Jun 1st, 2010, 12:30pm » |
Quote Modify
|
28g
on Jan 16th, 2010, 5:01pm, RonWeasley wrote:7 mobsters voted, but only one move could win. Actually, there was a tie (3-3). The winners:
Cc3s Cb3w Db2n Hf3e and Cb3w Db2n Hf3e Df4s.
I flipped a galleon, as mobsters do on TV, see? The winner was |
|
Gosh, given the incessant coin flipping, it's no wonder I had trouble predicting the Mob's moves.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 35
« Reply #87 on: Jun 1st, 2010, 12:57pm » |
Quote Modify
|
35g
on May 4th, 2010, 4:12am, RonWeasley wrote:The winner, with only 5 votes cast, Rh2n Db3n Ee5n Ee6w.
It took three tries for me to get the candidate moves right. And the few remaining mobsters were not very confident in this move. Perhaps this will confuse Fritz as much as us. |
|
Argh! How could the Mob be so chaotic and still crush me with ease? That's rubbing salt in the wound. 
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Fritz's commentary
« Reply #88 on: Jun 1st, 2010, 1:21pm » |
Quote Modify
|
I have finished skimming all the Mob discussion. I now have a very different impression than I did before. The Mob is not nearly as infallible as I was giving it credit for. In particular, after 38g I was convinced that the Mob had analyzed to a forced goal when it had done no such thing. I think I was just psychologically broken down from having tried so hard to come back from the early camel hostage, to no avail.
This game has taught me above all how flawed my own play is, and how much room there is to play better than I do, particularly since there seems to be plenty of room to play better than the Mob which already plays far above me. There is still much uncertainty and confusion about Arimaa strategy, and great possibility to refine our tactical training. More clearly than before, I anticipate the possibility of a 3000-level player emerging. The future of Arimaa is bright.
|
| « Last Edit: Jun 1st, 2010, 1:22pm by Fritzlein » |
IP Logged |
|
|
|
Hippo
Forum Guru
Arimaa player #4450
Gender:
Posts: 883
|
|
Re: Move 14
« Reply #89 on: Jun 2nd, 2010, 12:13am » |
Quote Modify
|
on Jun 1st, 2010, 11:55am, Fritzlein wrote:14g
Well, I managed to scare Simon and arimaa_master at least. I am happy that you guys at least felt some pressure, although I never thought I was winning at any point after 10s. Perhaps my over-confident trash talk served some purpose of keeping you in doubt. |
|
I were not in arimaa comunity at that time ... but ex post ... was the move 14g ending with Ef3 instead Eg4 winning fast?
BTW: Thanks for the comments
|
| « Last Edit: Jun 2nd, 2010, 3:05am by Hippo » |
IP Logged |
|
|
|
|
Home
Help
Search
Members
Login
Register
Reply
Notify of replies
Send Topic
Print