Author |
Topic: Fritz's commentary (Read 13041 times) |
|
Fritzlein
Forum Guru
Arimaa player #706
Gender: 
Posts: 5928
|
 |
30g
« Reply #60 on: Jun 1st, 2010, 7:29am » |
 Quote  Modify
|
30g
The Mob has a credible option to defend the c6-trap with 30g ED>E< plus a fourth step. Since I would have to spend one step to save my eastern dog, I would not be able to frame the Mob’s rabbit, and would probably have to defend my b6-rabbit with another rabbit on c7, allowing the Mob to pull it into c6 on the next turn for an awkward-looking position for me in the west. Nevertheless, the Mob’s move 29g has convinced me once and for all that the Mob is constitutionally incapable of playing passively. They will always counter-attack rather than defending, and I just have to get used to that.
I don’t see any effective material counter-attack for the Mob, so all that remains is goal an eastern goal attack, despite my delaying rabbit step on 29s. I therefore expect 30g H^Rh2^ plus two more steps. For those two steps the Mob has many reasonable-looking options which seem to hinge on sharp tactics several moves into the future. Their f2-rabbit would be hanging, and their g-file is open, both of which can be remedied with Rf2>, but then they would have no defense of the f3-trap, which is important in some lines.
30g H^Rh2^ED< is optically strong for goal attacking, as it gets my dog on the wrong wing too, leaving only rabbits defending goal in the northeast, plus it incidentally threatens capture. It appears, however, that the simple 30s CR^*Rh6<Rg5> is then adequate to defense. A subtle point is that with the elephant off the fifth rank, my horse is able to come back to f5 in some lines.
30g H^Rh2^Rf>Re> protects the g-file and keeps my horse out of the east, but the lack of a capture threat allows 30s CR^*Rh6vE> as a defense.
30g H^Rh2^Re^^ precludes my elephant taking one eastern step on 31s, but perhaps then I could take two with 30s CR^*ERev.
I guess I will have to predict the shiny-looking 30g H^Rh2^ED<, which makes me feel good because I believe that will let me catch up materially and still defend goal.
|
|
 IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
30s
« Reply #61 on: Jun 1st, 2010, 7:30am » |
Quote Modify
|
30s
The Mob played 30g DfRvRf^^, and my heart sank when I saw it. I was so proud that my fourth step of 29s slowed down the Mob’s h-file rabbit that I didn’t even consider they would attack up the f-file instead. I was right that they would counter-attack rather than defend their threatened rabbit, and right that their counter-attack would need to be focused on goal rather than material. What I didn’t see was their excellent move to achieve these objectives.
I simply must spend two steps of turn to capture the Mob’s c5-rabbit in c6. If I delay, my position will only get worse, because the pieces holding the rabbit in place, particularly my horse, can’t otherwise participate in the game. That leaves me two steps to try to defend against the Mob’s powerful goal threats.
I would like to make use of my rabbit on g4 and the Mob’s thin goal defense by stationing my elephant on e3. Then if the Mob advances their f4 rabbit, it opens a four-step goal for my g4-rabbit through the f3-trap. They would be required to use one step on defense while attacking, which makes my own use of steps more efficient. Furthermore, it is good for my elephant to rest temporarily on e3 so that if the Mob attempts to take my horse hostage with 31g EHvv, I will have the resource of 31s HRvRH> to keep my horse at liberty. Finally, my elephant on e3 protects f3 so that I don’t lose my g4 rabbit, and thus remain behind only by cat for rabbit.
So what to do for my fourth step? Defending the f6-trap with Rh6< would be ideal for trap control, but unfortunately weakens my h-file too much. After 30s CR^*E>Rh6< 31g DgR6<Rf^Re>, the Mob’s goal threat is too strong and mine is too weak. Their threat of 32g R5>E> is so strong that if they can stop my goal threat with their other two steps, I just flat lose. In particular, my move 31s Rg<E>HE<*, which captures a horse, and creates two goal threats, loses to 32g R5>E>Rh2<. What a bummer.
What if I instead use my fourth step on something innocuous like R8>? That doesn’t weaken my h-file, but it does give the Mob one extra step on attack because their dog doesn’t have to push to g6. 30s CR^*E>R8> 31g Dg^Rf^>Re> sets up a goal-in one threat, so I again don’t have time to take the gold horse. Moreover, the threat is made without moving the gold elephant, so my horse is still shut out of the action. Having achieved material gain by a lopsided position, I would gladly restore piece balance to stop the Mob’s goal threat, but it seems they need not give me that option.
I hate to contemplate that I might not get the efficiency of having my elephant on e3, but I might be forced to play 30s E^HR^*E>. My horse would rather chase after the invading gold dog, but if it doesn’t run away from the gold elephant, it will simply be taken hostage. My elephant is the wrong piece for defending against the goal attack, but do I have a choice?
Perhaps in this position I will, for a change, do some deep tactical analysis to decide between my primary choices, rather than relying solely on my strategic instincts as I have for most of the game. I built my reserve up to fifteen days, and this seems like a fine move to spend it on.
…
I confess, this is the first position of the game in which I miss having a computer to help me analyze. The goal threats on both sides mean that I have to spend time sorting out moves that merely look deadly from moves that actually are deadly. A computer could do that instantly and accurately, but I must suffer through the analysis myself, and probably come out with the wrong answer anyway. Yet I don’t regret my resolution not to use computer assistance. My goal attack and defense are still quite weak, so I need all the practice analyzing in my head that I can get.
…
My reserve is almost drained, and I haven’t done that deep analysis that I promised myself. I would simply submit 30s E^HR^*E> on instinct and on the shallow analysis I have done, except that it occurred to me that 30s CR^*RE> deprives the Mob of the d3 square for forking my horse, so that in some lines my elephant can leave e3 while my horse is still standing on d5. So before I decline to take the rabbit in that fashion, I would rather be very sure the Mob has a crushing attack to deter me. If my defenses hold, for example 30s CR^*RE> 31g Dg^Rf^^E> 31s RDe^DRv, then leaving my horse in the center looks strongest.
…
All right, I think I am going to take the plunge with 30s CR^*RE>, hoping that there is no forced goal that I have overlooked. The gold rabbit on d3 matters in the critical continuation 31g Dg^Rf^>Re> 31s E^>^Rg< 32g Rf^H<RH>* 32s E^Dev*Rf>, where Gold can’t push the horse to d3, forking and winning it. Alternatively 32g Rd>H<RH>* 32s E^Dev*Rf> 33g EHvv 33s Db>vDC> means there is no good square for the Mob to hold my horse hostage on, allowing me at least to make a race of it. A losing race, perhaps, but at least a legitimate fight.
In short, I have several intuitions about the position, i.e. that I must take the Mob’s rabbit, that my horse wants to be on d5, and that my elephant wants to be on e3. If that configuration doesn’t lose me the game by goal, it should gain me time in the trap control fight, and leave me just down a cat for a rabbit in a tense position. I will still be losing, of course, but given how desperate I have been since giving up my camel hostage in the opening, such an ending is all I can ask for even if I ultimately lose it.
…
Aargh! Just as I was going to enter my move I saw that 30s CR^*RE>31g Dg^Rf^>Re> 31s E^>^Rg< is crushed by 32g EvHvDg^Rg>. It makes a world of difference that the rabbit is not pinned in the trap, so that the Mob has a scattering option available. I tried to save my d5-horse, e3-elephant configuration in a variety of ways, but I just couldn’t do it. It turns out the Mob’s f4 rabbit is simply to powerful to permit to advance.
Therefore I’m back to the uninspiring 30s E^HR^*E> in order to slow down the Mob’s rabbit. It doesn’t give me half the chances at counterplay of the other move does, but at least it will lose relatively slowly. Oh, well. I guess the Mob could always make a mistake later.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
31g
« Reply #62 on: Jun 1st, 2010, 7:30am » |
Quote Modify
|
31g
The Mob has me over a barrel with their elephant-dog attack. Given that MHx is gone from both sides, the ED attack is the new EH attack. My elephant will be obliged to defend the f6 trap, and my horse had to leave the center and relinquish its crossing prospects. The standard strategy would be for Gold to swarm on the wing where my elephant is tied down, in order to tie me down even more and release their own elephant.
However, they can’t be in too great a hurry with 31g ED<Dg^, because then my horse would triumphantly cross to the east after all, temporarily able to slide through e5 due to my elephant on e4. The gold elephant has to stay put for now.
On the other hand, I am threatening to capture the f4-rabbit in f6. There is no way the Mob wants to let go of their excellent rabbit in exchange for capturing my sickly g4 rabbit with 31g RHvHR<*. Therefore the Mob must play at least Dg^ to secure shared control of f6.
The Mob’s other three steps are less clear to me. Coming after my horse with the gold elephant would be a clear waste of time. First they must tie down my elephant, and only then use the fact that their elephant is more mobile than mine, making my horse less free to roam.
The all-out attack 31g Dg^HR^Rf^ is amusing, but can be slowed by 31s H^>De>H> with a counter-threat to capture the attacking rabbit in f3. I would have time to secure my position with the positional improvement that my endangered g4-rabbit would be safer on g5.
The Mob’s three steps don’t suffice to unfreeze the f4-rabbit with anything other than the horse, so I suspect that their best play is to put the goal attack on hold and think strategically about the control game. One option is 31g Dg^Ra2>RCav to threaten two rabbits at once. That’s a good plan to keep for future reference, but at the moment f6 is not yet secure, so 31s ERvE^Rg< threatens a dog and goal and gives me time to consolidate and try to save everything. If I managed to get my horse and dog going against c3, that rabbit on a3 would be an asset to me rather than a liability. First the Mob should make sure I can’t contest the c3 trap.
And that brings me to my actual move prediction. 31g Dg^Rh2^Re^^ shores up the Mob’s f3 trap and starts a swarm consistent with the general plan of an ED attack. For a positional move like this, I am unlikely to predict exactly correctly, but the Mob will love to advance rabbits with three pieces off each side, so I doubt I will be far wrong. Rd^ is also tempting because it overprotects c3 in case I have any ideas of counterplay there, but I rather expect the Mob to be focused on its own attack rather than on slowing down my plans.
31g Dg^Re>^^ might be optically strong, but appears over-eager to me. After 31s REvER<, the Mob might regret having filled their f3-trap, granting my g4 rabbit a move to breathe.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
31s
« Reply #63 on: Jun 1st, 2010, 7:52am » |
Quote Modify
|
31s
I was shocked by the Mob’s 31g Dg^ED<Re^. In my analysis, I had seen my possible reply 31s H>>>Ev and looked no further. I reasoned that the Mob surely couldn’t allow their advanced dog to be given as a hostage to my horse. Alas, after the fact I see that the Mob could pull my f7 rabbit into the f6-trap, after which I would have no way to save it. My instincts are not attuned to a situation where I have so few defenders around my home traps that I don’t have an extra one to bring up behind it.
This turn of events is terribly demoralizing. I had begun to see lines where I retained a rabbit as compensation for my lost cat, but now it appears I am forced to return the rabbit immediately. I don’t see anything better than 31s ER^<, trying to endanger the Mob’s advanced rabbit in the c6-trap.
This late in the game it is probably suicidal to bring an enemy rabbit forward, but how else should I even complicate the game? Since it turns out that my horse is an inadequate defender of the northeast, my elephant must park there if I am not to lose in short order. The northwest is the only other place I have pieces that can threaten anything.
I have looked at a few tries for the Mob to gain more than simply capturing my g4-rabbit, and I don’t see any worse trouble than that. Then I would be a cat down once again, and every trade makes my cat disadvantage worse, but I would at least have some minimal compensation from threatening their advanced rabbit. I could be wrong, though. I almost hope that the Mob has a devastating followup to put me out of my misery.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
32g
« Reply #64 on: Jun 1st, 2010, 7:53am » |
Quote Modify
|
32g
The Mob has some flashy moves that might look tempting given the fragile nature of my position. 32g EDvv threatens my dog, my horse, and goal in one. But 32s E^DE<*Db> trades dogs with a probable rabbit trade to follow. The Mob can do better than trade DR for DR.
Going straight for a goal would be the Mob’s style, but 32g ED<EvRe^, threatening my horse and threatening goal in two steps appears comfortably met by 32 Dc>H^DRv. Other direct goal tries also appear to let me get a piece in front of the threatened rabbit.
According to my analysis, 32g RHvHR<* is simplest and best, so that’s my prediction. The Mob’s horse would not be hanging, since taking it would allow goal, so I would play something like 32s H>RHvDb> to get a threat against the Mob’s rabbit before it could be supported. My threat would be poor compensation for the Mob’s extra cat, which is precisely why the Mob should allow it for the privilege of getting my rabbit off the board.
Given my track record, though, my analysis is not to be trusted. Probably the Mob will come up with some devastating plan that hasn’t even occurred to me.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
32s
« Reply #65 on: Jun 1st, 2010, 7:53am » |
Quote Modify
|
32s
As expected, the Mob made the simple and strong move of capturing my g4 rabbit. That makes 8 of 32 moves I have correctly predicted.
...
After a week of vacation, I return to the game with fresh eyes only to see how hopeless my position is. Being down by a cat in an endgame is in itself quite bad, but my depleted eastern flank and the Mob's well-placed dog on g6 put me at a positional disadvantage as well. I wanted a race, but now that the position has become extremely imbalanced, it is obvious that the Mob can generate a goal in the east faster than I could in the west. My only hope is for the Mob to blunder, which they haven't done all game.
I see nothing better than continuing with my plan to harass the Mob's advanced rabbit with 32s H>RHvDb>. That incidentally creates a threat to capture the Mob's hanging horse as well. My hope is that the Mob will make the strategic mistake of attempting to defend their rabbit, but throughout the game they haven't shown the slightest inclination to defense. Of course they will ignore my current threat as well and counter-attack instead. Earlier, I had briefly hoped to win back a rabbit to be down only slightly in material, but at present I don't imagine that I can win a rabbit without giving up either a rabbit or a goal in return.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
33g
« Reply #66 on: Jun 1st, 2010, 7:53am » |
Quote Modify
|
33g
I don't see a way for the Mob to advance its d5-rabbit or to save it without allowing its elephant to be awkwardly pinned.
The Mob could pull my f7 rabbit into f6, which would have won a rabbit outright in another line, but here is too slow. Among other things, it would open the possibility of my horse switching wings along the fourth rank.
It would be typical for the Mob to ignore material and advance a rabbit as quickly as possible in the east, say with 33g H>Rh2<^^, but then 33s H>>HRv is fine for me. They may have a safer way to advance a rabbit, though, as I have overlooked their attacking possibilities in the past.
The Mob should at least consider trying to capture my dog and/or horse in their home traps, but I don't see how to pull it off. Trading dogs would be silly for them given the current power of their g6-dog.
I'm going to predict 33g H>RH<Re^. A rabbit trade is great for the Mob, as it would weaken my east even further while not hurting them positionally. I would be forced to launch a doomed elephant-rabbit attack in the southeast. The Mob, with its prodigious tactical calculation ability, will not fear a dangerous-looking line that they can prove will peter out. They will instead force matters to a head, which will be a mercy to me relative to drawing out my misery even further.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
33s
« Reply #67 on: Jun 1st, 2010, 7:54am » |
Quote Modify
|
33s
The Mob thought for the full week and played the surprising 33g ED<R5>E>. I would have thought they had something better to do than reverse my dragging their rabbit from e5 to d5. True, because of the way the repetition rule works, I can’t directly undo their move, but I can effectively undo it with 33s Dc>H>HR<. Then it will have to be the Mob that deviates, not me. I frankly don’t see what they gain by having forced my horse to from d4 to e5.
I am going to make this move with perhaps the least analysis of any move so far in the game. Mostly this is a byproduct of despair; I think I am strategically forced to hunt down the Mob’s advanced rabbit even though it is a losing strategy. If there is some tactical point that I am overlooking which will speed my demise even more relative to the doom I already face, then so be it.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
34g
« Reply #68 on: Jun 1st, 2010, 7:54am » |
Quote Modify
|
34g
I am trying to understand why the Mob wanted my horse on e5 rather than on d4. They can fork it with 34g EHvv, but this seems to lose material after 34s E^DEv*Ev. More promising would be a single push while protecting their exposed horse and dog with, say, 34g EHvRe>Dg^, but then I can hold everything together starting with Rh6< to protect the f6 trap.
I think I will predict the same move I did for 33g, namely H>RH<Re^. A rabbit trade is as fruitless for me now as it was before. The Mob must realize this and not expect me to trade rabbits, because if they did expect a trade then their move 33g results in the same position after the trade, except one move later, delaying their ultimate victory by two weeks.
Instead I am guessing that my horse on e5 is a liability in this line, frozen if my elephant leaves first, and soon in danger of capture in f6. I think this is the answer to the riddle of 33g: The Mob already knew that a rabbit trade was hopeless for me, so they maneuvered to make my last-ditch ER charge hopeless as well.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
34s
« Reply #69 on: Jun 1st, 2010, 7:54am » |
Quote Modify
|
34s
The Mob confounded my prediction again, playing 34g EHvRe>Dg^ as I considered but dismissed. I certainly didn’t do any deep analysis to justify my intuition last time, so the obvious conclusion is that the move I did predict works less well than I thought it would, or that the move the Mob actually played works better than I thought it would. It seems improbable that the Mob simply miscalculated.
On the other hand, I noticed several people posted to the forum thread for the Mob’s move after their move had been played on the board. (I don’t read the Mob’s discussion, but since I constantly hang out in the forum, I can’t help but see when people post.) My recollection from the 2007 Mob game is that we seldom discussed anything after sending the move unless there was a sharp split in the voting and someone thought we had made the wrong choice. Knowing that the Mob’s discussion continued raises my estimate of the probability of a Mob mistake from miniscule to merely minor.
One possibility for me is 34s Dc>H>>^, chasing the Mob’s exposed dog. This seems promising because my horse can’t easily be persecuted by the Mob’s elephant, and my exposed dog can’t be chased by the Mob’s horse because of the weakness of their eastern wing. Whether my horse takes the Mob’s dog hostage, or their dog runs away, my elephant is suddenly freed from defending the Mob’s urgent goal and material threats in the east, while their central rabbit remains an liability to them. Strategically this looks promising.
Tactically, however, the Mob could apparently force the trade of DR for DR with 35g EvEDvDg^ 35s H^^DHv 36g ED<*CaR>. After the pieces come off, I am still behind by a cat, which weighs more as the board become emptier. Futhermore, my rebalancing of pieces would leave horse facing horse in the east, which would tend to deaden the position and make the creation of an endgame race less likely.
Therefore I am drawn to the sharp move 34s Rh6<EvDdR<. Yes, that permits my horse to be forked and/or taken hostage between the Mob’s traps. On the other hand, eliminating the Mob’s only advanced rabbit, even at the potential cost of a rabbit in the f6-trap, would give my small western pieces the freedom to advance, potentially enabling a double-trap swarm by me, with my elephant contesting the Mob’s eastern trap while my little guys fight for the southwest.
My pieces are not badly aligned for a swarm; it would be a bonus that my elephant would be attacking the trap with the Mob’s deputy. Also my advanced rabbits, which are a liability in a tame and balanced position, become an asset in a swarm. Finally, the possible dissension within the Mob gives me courage to play sharply, even though the Mob should by rights be better at calculating tactics than I am, and therefore should have foreseen and refuted my move before allowing me to play it.
Yes, the Mob will have a pair of powerful strategic assets in the form of holding my horse hostage and their dog playing sherrif of the northeast, which combined with their extra cat still give them the upper hand. However, according to my calculations, I don’t immediately lose material and I do get a tense, unbalanced position. Given that my other option is trading down into a sterile endgame behind by a cat, I will not be sad to lose via a last-ditch swarm instead.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
35g
« Reply #70 on: Jun 1st, 2010, 7:55am » |
Quote Modify
|
35g
The Mob’s c5-rabbit can be saved only by its elephant, which has more important things to do. Passive play is no longer an option for them. Indeed, they must not only play actively, but quickly. My threat is to share control of all four traps, making it impossible for them to ever make a capture for the remainder of the game.
The two steps EHv, forking my horse, seem mandatory. If the Mob has no capture threat in the c3-trap, I have too many possibilities for capturing the c5-rabbit and consolidating with the other two steps. Pushing the horse a second time, however, can only weaken the threat by either giving me a second defender of the c3-trap, or by allowing my dog to retreat from the c3-trap.
On the other hand, my horse standing next to the f3-trap will give me the bodacious goal threat of 35s Rh4vEvEH^. This may have been overlooked by the Mob’s computer checking, given that 36g HE<EDv* appears to win a horse for a rabbit, when in fact 36s HvvHR^* appears to force goal or win the Mob’s horse in another few moves.
Therefore, if the Mob intends to fork my horse (and I think they must), it seems they also need the defensive step Rh2^. That is hardly pure defense, since my own eastern defense is so weak. My goal line is almost h4, given the gold dog waiting to help any gold rabbit that gets so far. Nevertheless, that leaves the Mob only one free step. I don’t see anything they can do with that step to deter 35s CR^*Ddvv, winning a rabbit for nothing. Indeed, very few single steps improve the Mob’s position at all. Therefore I predict 35g EhvRh2^Rh1^.
Am I hallucinating? Probably, because the Mob never blunders material. But they do get a horse hostage for it, so perhaps it is according to their plan, and I will shortly feel the pain.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
35s
« Reply #71 on: Jun 1st, 2010, 7:55am » |
Quote Modify
|
35s
The Mob played 35g Rh2^Db^E^<, defying my glib assertion that their elephant could not defend passively, and had more important things to do. I forgot to adequately account for the fact that defending an advanced rabbit is inherently also threatening to bring that rabbit to goal in some degree. The Mob’s dog advance prevents me from simply framing the advanced rabbit, as I otherwise intended to do (based on remembered calculations from when I had a dog on b4, blocking the Mob’s dog). Indeed, it seems that I must now keep the Mob’s dog out of b6 at all costs. Furthermore, the Mob’s step Rh2^, which was necessary for defense, also has a latent attacking threat because my defense of the h-file is paper-thin.
Another way in which my predictions have been confounded is that I confidently asserted that the Mob would never again be in time trouble. Indeed I stopped keeping track of how they were using their time because it seemed like a distraction. The temptation to time my own moves differently in order to influence the Mob’s time usage was simply a pitfall for me to put pressure on myself. But somehow over the past several moves they have slowly drained their reserve to just one day. I can only guess that they perceive their position as getting worse.
A brief analysis suggests that 35s Ra7^^Rb8^< is a viable way to protect the c6-trap and keep the Mob’s dog at bay. I will have to look closer at the tactics, but strategically it would feel good to be using three of four steps on rabbit advancement in a position nearing the endgame, in which I was just last turn calculating whether I was sufficiently advanced to start a swarm.
If the rabbit advance is not viable, it seems that I would have to keep the b4-dog out of b6 by pulling the Mob’s advanced rabbit to b5 while plugging c5 with a dog. This would be a strategic concession and net loss of time, as I make my capture threat less pressing while advancing nothing. But if that’s the only way to prevent an invasion without losing material, it may be forced on me.
…
Further analysis shows that 35s Ra7^^Rb8^< is refuted by 36g ED>EvRd^. In previous analysis I had dodged this sort of maneuver by sliding my c4-dog back to b4, but now that the Mob’s dog is on b4, I have no retreat. Thus the Mob’s move 35g was strong for more reasons than I realized.
My alternatives appear to be 35s RC<Dc^E> and 35s Dc>D5RvE>. I would prefer to keep the Mob’s advanced rabbit on b5 rather than c4, because on b5 it is more threatened and it blocks the b4-dog from b6. On the other hand, the rabbit would remain more of a threat on offense from b5, and I may need the option of pushing the rabbit into the c3-trap in order for my swarm not to lose a piece.
The elephant step is both an attempt to race and a recognition that my elephant has nothing else to do. It can’t leave the east without permitting immediate catastrophe anyway, so my elephant might as well try to generate a threat on offense instead of remaining passive while the rest of my pieces fight for their lives.
I have three threats in this position: capturing the Mob’s advanced rabbit; forcing goal in the southeast; and chasing their g7 dog with my horse. The latter is so slow as to be incidental to the main action, so I would temporarily prefer to leave my horse in the middle where it can choose to help with the goal attack or help me not lose material in the c3-trap.
But, all positional thinking aside, it all comes down to tactical analysis. Knowing that the Mob has a refutation to my strategically preferred move makes me wonder whether both of my present two candidates are tactically flawed as well.
…
Yep, they’re both flawed.
35s RC<Dc^E> fails to 36g EDvv. My d3-dog could then only be saved by my c5-dog, but mobbing the latter opens up both the threat of the Mob’s b5-rabbit goaling up the center, and the threat of my b6-rabbit being captured in the c6-trap.
35s Dc>D5RvE> also appears inadequate, although it prolongs the struggle. 36g EvEdvRh1^ 36s EH<DcRv 37g Re2>^EDv gives the Mob enough defense of the southeast while regaining control of the c3-trap to win more material.
Therefore I am now leaning towards 35s Dc>D5RvHv. I hate for my horse to abandon the hope of ever coming back to the f6-trap to menace the Mob’s dog and provide some defense to free up my elephant. Also I hate to voluntarily move my horse closer to a position where the Mob’s elephant can fruitfully hold it hostage. Unfortunately, the strategically poor horse step seems tactically necessary to help me contest the c3-trap just a little bit longer than I otherwise could, buying me another few moves in which the Mob could blunder. I know the Mob isn’t going to crumble, but I have to keep hope alive as long as I can. If I need to resign, I can always do so later.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
36g
« Reply #72 on: Jun 1st, 2010, 7:56am » |
Quote Modify
|
36g
With its goal threat pushed back, the Mob has to switch gears again, and not continue its attack with 36g DbC<DR<. I could defend the c6-trap for a little while as I progressed in the southwest, and ultimately reach goal faster.
36g DbC<EvRh1^ is more critical, as I can’t defend both eastern traps at the same time, and therefore must lose a dog. Apparently, however, I can attack instead of defending and have just enough tempo to come out on top after 36s E>EH<R. Neither dog capture seems to work. 37g DbR<ED^* 37s EvRcE^H*C> or 37g EDv<* 37s EvER<*. I would not be surprised if I have miscalculated this race, because this is exactly the type of position where the computers and many eyeballs of the Mob should enable them to outplay me, but beggars can’t be choosers. Since I am a cat down, racing is my only hope.
One should keep in mind that Mob’s dog is poised to make a capture threat in the f6-trap at any time, but at the moment it would be awkward with my elephant still in striking distance and none of my heavy pieces (yet) held hostage by the Mob’s elephant.
The Mob’s elephant is under-utilized at the moment. It would like to either take my wayward horse hostage or capture one of my dogs in the c3-trap. Using three steps on 36g EvEDvRh1^ is a bid to keep both options open. I would need five steps to occupy both c4 and g3, so instead I would have to defend with something like 36s DcRvHvCb>, offering my horse hostage on e2 to prevent my dog from getting pushed to d2, which would lose control of the c3-trap. Oh, and the Mob’s rabbit step stifles my counter attack again. Ugh. That’s why I predict this move.
Ironically, if the Mob directly attempts to take my horse hostage with 36g E>vvRh1^, it works less well because I have the option of 36s DcRvDdvH>.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
36s
« Reply #73 on: Jun 1st, 2010, 7:56am » |
Quote Modify
|
36s
The Mob played 36g HvRh3<^H^. I didn’t even consider this move, so I am presumably dead meat.
At first glance I would say that I can no longer contemplate swarming, because the Mob’s g4-rabbit is now too threatening, but I am too far advanced to pull out of my swarm without suffering losses. First attention will go to 36s H^^>>, with a plan to give up on the f3-trap, use my elephant to defend the c3-trap, and (very optimistically) win something in the f6-trap with my horse or (somewhat optimistically) keep pace with what I am losing to the Mob in their home traps. The only thing the Mob’s move gives me strategically is the opportunity to target the rabbit they just pushed and the dog they left offside. Tactically, though, I expect my horse retreat to lose material because of my own offside rabbit and dog.
…
My defensive mindset is betrayed by the fact that I didn’t even notice the possibility of attacking with 36s EvEHvRhv until later. Of course, this will have been the first counter the Mob looked at. My attack is apparently strong enough to force the gold elephant home with 37g E>v>v, but it is cold comfort to have forced the Mob into a winning position. After 37s HvHR^Rhv 38g Rh<HR^Rg^, my elephant is the only thing that can stop the Mob’s rabbit, so I have to defend on 38s. Whether I use my elephant on defense then or on 37s, it would be strategic resignation on my part. I must defend the northeast with my horse to keep my elephant free to keep fighting.
I’m not even going to use my whole week on this move, because no move other than 36s H^^>> looks remotely viable.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
37g
« Reply #74 on: Jun 1st, 2010, 7:57am » |
Quote Modify
|
37g
Although I made a defensive move, the position remains as tense as before if not moreso. The Mob can make capture threats in both c3 and c6, as well as having some goal pressure from their two fourth-rank rabbits. Meanwhile I am threatening an immediate capture in c6, threatening to take a hostage around f6, and threatening a goal attack of my own around f3.
An optically devastating tactic for the Mob is 37g EvDbC<, threatening captures in both c3 and c6, which I can’t defend in only four steps. However, the Mob must use its fourth step for defense with Rh^, or else my goal attack with 37s RhvER> is too powerful for the Mob to execute either of its capture threats. And 37g EvDbC<Rh^ itself permits me to play 37s Rb8>vRg>H^, initiating a dog trade that at least keeps me hoping for a few more moves. That’s better than the immediate loss I feared when their 36g hit the screen.
The direct attack on c6 with 37g DbC<DR< is tempting because it is awkward for me to defend, but in that case I can initiate a strong counter-attack instead. In that line I would be helped by having a slightly more advanced rabbit than the Mob has.
I wish it were hard to find a crushing move for the Mob, but 37g EvEDdvDg^ is a very straightforward attempt to win a dog, for which I have no adequate counter. I talked before about using my elephant to defend the c3-trap, but analysis of this line convinces me that my center is too open for me to commit my elephant to the southwest. So instead of switching flanks with my elephant, I can perhaps try 37s DcRvC>Rhv, trying to keep the tension in the position for another move or two, but after the Mob takes my central dog hostage on d2 on move 38g, I can’t keep my position from collapsing for very long.
I’m afraid the move 37g EvEDdvDg^ is only too easy to correctly predict. Sigh.
|
|
IP Logged |
|
|
|
|
Home
Help
Search
Members
Login
Register
Reply
Notify of replies
Send Topic
Print