|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Title: Arimaa Strategy: Camel Hostage Post by tharkun on Mar 8th, 2012, 6:46pm Greetings! I've been looking to hone my strategic skills and decided to start by tackling camel hostage situations. I gathered quite a few positions and wrote down my impressions about the plans for both sides and a tenative assessment of the position. I would love to learn the feelings/thoughts of other Arimaa players, especially when it differs from my own, so that I can correct my misconceptions and improve my intuition in these kind of positions. I may do something similar for other strategic concepts in the near future, so feedback on my concept is very much welcome as well. Do you think something like this (a large number of positions with a specific strategic theme) can be a useful resource in general, possibly as an extension of the wikibook?
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Title: GoldRe: Arimaa Strategy: Camel Hostage Post by clyring on Mar 8th, 2012, 10:01pm I'll offer my opinions on these positions as well. #1: While the hostage situation itself is fairly solid, gold to move can shut the silver horse out of g3 with 15g Hf2n Rf1n Dg3s Hf3e and silver will have a hard time both stopping gold's swarm and preventing gold from dragging minor pieces to f3. The game is far from over. #2: Silver's dog and rabbit are ill-prepared to fight against a horse and the silver camel is not very useful on f7. Further, silver's western clump of pieces is far enough advanced to make a nice target if the gold elephant switches wings, but not far enough forward to protect one another effectively. However, at the same time, gold's horses are far away and more gold pieces need to be advanced to support the attack. Neither player has any fast options and it seems the game will take a slow manoeuvring course. The position's value is very unclear; I don't honestly have a clue who should win here. #3: Gold's position has worsened considerably despite winning the rabbit. Gold's only free horse simply cannot offer enough power to handle silver's attack and too much material for both sides is concentrated around f6 for either abandoning it or attacking to be feasible. Silver should have a significant advantage here. #4: Silver's lack of a seconds horse would significantly reduce the value of the hostage, but silver's own advanced pieces in the east are in the way of any attempts by gold to make progress by pulling out small pieces. Both of gold's dogs are poorly placed and gold's goal defense is dangerously thin across the board. The rabbit on g4 is vulnerable, but perfectly positioned to delay a silver attack. If silver can get an attack going in the west fast enough, they should have an appreciable advantage. However, gold will always have the threat of leaving f6 with their elephant and silver can do little to prevent gold from eventually switching the elephant out without losses, making the position rather double-edged. #5: In my opinion, gold has more than sufficient positional compensation for the lost rabbit. Taking the gold camel hostage here seems unlikely to yield good results with the silver horse buried and an encroaching mass of pieces. Silver's best hope is probably to play for profit elsewhere, as gold simply has too many pieces there for silver to make progress. #6: Silver is clearly down materially, h for R. Gold is also not far from switching his elephant out, but since c3 is temporarily safe, it may be possible for silver to intervene with the process using her camel. There is also the possibility that gold, in manoeuvring around c6 will create a long enough window during which the gold elephant is tied down further by the pieces to replace it but not yet able to leave wherein silver may be able to attack f3. I consider gold to be better off, but the position is admittedly rather unclear to me. (tbc) |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Title: Re: Arimaa Strategy: Camel Hostage Post by chessandgo on Mar 12th, 2012, 7:26am Interesting Tharkun! As a meta-remark, this is a lot of material, and I fear discussing so many positions might dilute the discussion quality. Maybe another idea would be to have slightly fewer positions, but regrouped by a theme (for example a camel hostage for a rabbit), and compare position #12 with some other position with different strategic features. All positions where silver [edit: by silver I mean the hostaging side, although sometimes it's gold] has a good camel hostage, material is equal and army advancement is about par should be massively favourable to silver (1,2,4 [assuming H~=rr],5 to some extent,9 to some extent,10,11,13 etc). Normally silver should have had to work to establish a good camel hostage position, which means Gold's development should be way ahead of Silver's. In all these positions it feels like Gold voluntarily gave up the caMel hostage, so that the number of effective development steps that each army has taken is about the same, and so silver should have a very clear advantage. In almost all these positions, I think pulling a rabbit would be a bad plan, especially the positions where captures have occured already. Advancing rabbits, and advancing pieces towards the enmy trap when that looks plausible is generally a better plan as far as I can tell (for example #16 the Rh4 is better on h4 than on g2 for Gold, silver pulling it towards f6 would just let Gold move his horse to g6 through g4 and get a strong attack. Likewise, #11 silver shouldn't drag a Rabbit to f6, but put a horse on g3, camel on f4 (possibly dog on e3), with rabbits following and attack around f3, that would be crushing.) #6 is fun, I think trading the M for hr on the next move should be best for Gold. md for HH is pretty balanced material. #7 is interesting as well, I'm not positive that the two hostages in the east are better than just one (I would actually guess it's worse). Silver should probably avoid the horse flip to g2 you mention, after which Gold would replace the Mg3 with a Horse and probably be fine. So moving a dog to g4 first looks good. If silver can keep all three EMH sealed in around f3, I'm not even sure Gold is ahead. #8 If Gold were to play and get his Horse out (to g4 for example) he'd be clearly ahead, so silver has to put pieces on h5 and g4 on this move. Then I agree the position is unclear. Like clyring I can't make it though the whole list in one pass, I'll try to come back to it later) PS: as for #5, there is no rabbit deficit, is there? |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Title: Re: Arimaa Strategy: Camel Hostage Post by tharkun on Mar 12th, 2012, 12:33pm Thanks for the comments! I think that both of you are right that I posted too many positions at once. On itself knowledge of strategic features like hostages or frames seems very useful but ultimately incomplete. As mentioned in the Wikibook, judging when they are favourable and determining how to proceed is the main area where humans can make use of their strategic intuition. That's why I thought such a discussion would be interesting. However, the length of my post may well have been too imposing to facilitate a high quality discussion. What would be a suitable format in general? About 8 positions at maximum, or is that still too imposing? I think it is beneficial to be able to compare positions, so too few of them isn't good either. Maybe I could post a set of positions and observe the run of the discussion for 1 or 2 weeks before possibly moving on to the next theme? First I want to make sure that the idea is useful for others and I'm not just clogging the forum ;D I have collected positions on 8 different strategic features so far, but I expect that I'll have to split up some of them to limit the number of positions. What shall I do with this thread? Leave it as it is? I am certainly going to try and group the positions according to subtheme as chessandgo suggested, so I could also leave just those positions belonging to a single subtheme in this thread and repost the others in a different thread later. (Of course, in that case I would copy and repost your comments as well wherever applicable). My first suggestion for dividing the positions into subthemes:
A small problem is that I doubt that I'm sufficiently capable to identify the salient strategic aspects in most positions apart from the glaringly obvious. In this case I borrowed my reasoning from your insights. Thus, I'm not really sure I will be able to make a sensible division into subthemes in general. I guess we will find out in due course ;) |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Title: Re: Arimaa Strategy: Camel Hostage Post by chessandgo on Mar 12th, 2012, 1:48pm Continuing: 14: has Gold lost a Cat in the meantime? 15: moving a dog east looks good indeed, if both silver dogs are used in the center / west, Gold's Cat will push into h5 and follow with the h2 Rabbit for a very strong goal attack. Both sides want to advance rabbits on both wings whenever possible. |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Title: Re: Arimaa Strategy: Camel Hostage Post by chessandgo on Mar 17th, 2012, 9:09am 16: I guess I'd take Gold, but the position is very complicated. The Hf2 must remain in the east : HH vs h is not such a decisive advantage, and the two silver dgos would be able to control f3 and replace the elephant without a godl Horse to counter. To me the question is whether Gold wants to advanace his Horse to g6 through g4, and threaten a goal attack if silver's pieces move towards f3, or stay around f3 to prevent a rotation. In both cases Gold needs to advance rabbits (west and center), whenever possible. 18: I like silver too. Although Gold should be able to avoid quick captures by relocating his Horse towards f6 if silver threatens a capture there, and using the camel to avoid captures in the west. So Gold has some time. I'd start by pusing the hg3 back to g4, so that the Ch3 be unfrozen, and then push the rh2 back. When the rabbit has been pushed back once or twice, Gold should have more space to unearth his pieces (Hg1, Cg2) and get a more decent hostage than the current (terrible) position. I guess silver should try to prevent that (the de6 would obviously be able to seize f4 whenever it wants, and threaten pieces on f6 if that seems good). 19: again, the ra4 is better for silver than on a8. I guess retaking control over f6 is the most pressing, and then silver can h+r attack on either wing. I'm not sure Gold can replace the elephant by small pieces around c6, especially needing to defend against a goal attack, maybe trading the camel for a horse when the horse reaches b3 would be an option (R for d is not too big a disadvatange, although bigger now that on a full board since dogs are becoming stronger) 20: Silver actually has a dog for two Rabbits if I read the position correctly. Re3 and Rd3 are fine, silver won't be able to threaten a capture. It's going to be dificult to rotate the Elephant out of d6 though, and silver wil quickly be able to threaten captures on f6. If Gold can advance pieces quickly to c5 and d6 he'll be fine, but I think I'd take silver and try to get a quick capture on f6. 21: Silver ahead. Gold should push the he6 back to e7 and advance pieces to f5 and e6 as quicly as possible I guess, but silver's western attack is going to be painful. 22 I guess I like Gold too. Silver should try to attack f3 though, pulling Rh2 would be too long. h->g4 and m->g5 on the next move threaten a strong attack on the next move. 23 two rabbits should be more than enough, I like silver. Even if Gold somehow manages to capture the rh5 while silver activates his pieces (which hopefully won't happen), the position would be still about equal or slightly better for silver. Also Gold must not lose a Horse for the camel right away, as HRR > M. |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Arimaa Forum » Powered by YaBB 1 Gold - SP 1.3.1! YaBB © 2000-2003. All Rights Reserved. |