|
||
Title: Game 184376: Belteshazzar vs. bot_Marwin2010CC Post by Belteshazzar on May 18th, 2011, 8:36am The game (http://arimaa.com/arimaa/gameroom/opengamewin.cgi?client=1&gameid=184376&role=v&side=w). What am I doing wrong? This is the last CC bot I have left to beat (until the 2011 ones are added to the ladder), but I still often feel like I don't know what I'm doing in the opening. In this case, I came out of the opening OK, but probably could have done better. After that I don't really see a clear mistake on my part, but I must have made some. |
||
Title: Re: Game 184376: Belteshazzar vs. bot_Marwin2010CC Post by Fritzlein on May 18th, 2011, 11:53am I don't see a clear mistake either until 18g. You can't let that silver dog get into c2, and having let it in, you should realize immediately that your horse's mission in life is to hunt down that dog as a priority over everything else; even the one step forward with your horse on 19g looks misguided strategically. Losing control of the f3 trap put you at a long-term strategic disadvantage which persisted for over twenty moves until the end of the game. You played fine tactically after that, making powerful threats and skillfully responding to marwin's threats, but minus a home trap you were fighting with one hand tied behind your back. There may be a specific point at which you could have done better tactically, but it doesn't jump out at me the way Silver's control of f3 does. There is a chess saying, "Tactics flow from a superior position," and I think that is what ultimately happened here. If we step back and think why you lost control of f3, it wasn't just that you neglected to chase the invading dog with your horse immediately. Your elephant could have kept control of f3 if it had had more help than just rabbits. In particular, you needed only a dog helping, but your two dogs were stuck on d1 and e1, where they had started the game twenty moves before. When you traded horses on move 13 (a nice trade for you, by the way, which should have meant that you were winning) you should have realized (A) your dogs have been promoted; only the enemy elephant and camel can attack them. (B) your defense of the g3 corner (your horse) has disappeared, so you might want to shore it up some time. Putting two and two together, moving your e1-dog to g3 has become strategically desirable already on move 13. The exactly correct time to execute this shift of forces isn't clear to me, but you were always so focused on the tactic of the moment that you never found time to activate your dogs until it was too late. I hope that this helps. Usually when you have a feeling of having lost without making any bad moves, it is because there is a strategic marker that you missed. In this case I think the missing strategy was activating your dog(s) and shoring up the weak corner on g3. |
||
Title: Re: Game 184376: Belteshazzar vs. bot_Marwin2010CC Post by ocmiente on May 18th, 2011, 12:00pm I still feel like I don't know what I'm doing in the opening many times either, but I'll give it a shot. 3g: Bringing back the elephant feels like losing momentum. Either don't bring the elephant forward so far at the beginning, or find something better to do once it's there. 5g: I do appreciate pushing the rabbit forward, but it seems like the best players have steered away from that strategy. The two main exceptions to this I've seen are when either nothing else seems to be working and a rabbit pull is available, or if the rabbit pull is basically handed to them to take. In this case, it seems like you are going out of your way to push that rabbit. I recommend only doing that if it isn't out of your way, and it doesn't support the opponent's strategy. This time, I think it takes too much time. 7g: This seems like the most significant problem move to me. This puts the camel in a position to be taken hostage, or taken - as it was in this game. I don't know what the best move would have been here, but moving the camel into the f3 trap rather than to g4 might have been better than the move made, just because the camel would have been safer. However, even if your camel had been taken hostage it wouldn't necessarily have been the end of the game. Since silver's camel is also on the east, a swarm to bury silver's elephant after taking the camel hostage would still be viable. I don't think I would have accepted the horse and cat for camel trade - but other people would. On 9g I might have pushed the silver horse into f6 and then started swarming. The horse needs to be pushed back because leaving the horse behind the elephant with the cat there might lead to other problems. |
||
Title: Re: Game 184376: Belteshazzar vs. bot_Marwin2010CC Post by Belteshazzar on May 18th, 2011, 1:30pm The points about the dogs make me think that I would have been better off starting with dogs behind the traps and cats back. I have been switching back and forth in that regard, but I suppose with my style of play it would be better to start with dogs behind the traps, at least in a long-thinking game. |
||
Title: Re: Game 184376: Belteshazzar vs. bot_Marwin2010CC Post by ocmiente on May 18th, 2011, 2:31pm on 05/18/11 at 13:30:15, Belteshazzar wrote:
I've wondered about that also. My current theory is that if a player tends to bring pieces to his own side of the board and frame or hostage them, then dogs are better behind the traps because it means he can frame dogs easier than if there was a cat behind the trap. On the other hand, if a player tends to take control of an opponent's trap, even at the expense of his own traps, then a cat is better because the player is more likely to lose the piece behind the trap. Better to lose a cat than a dog. Of course, if the player is really aggressive, maybe a rabbit behind the trap is a better idea. Using an unbalanced setup, the player can put a dog behind the trap on the side where he expects a frame of one of his opponent's pieces might be more likely, and a rabbit on the other side. |
||
Title: Re: Game 184376: Belteshazzar vs. bot_Marwin2010CC Post by lightvector on May 18th, 2011, 2:55pm Also, my feeling is that for attacking, having a piece like a dog on b1 or g1 can be helpful. The dog can very naturally advance up the b or g file to support an attack, or can help hold b3 or g3 while the horse or other pieces advance. In asymmetrical setups, doubling the horses on one side also can do this very well. |
||
Title: Re: Game 184376: Belteshazzar vs. bot_Marwin2010CC Post by Belteshazzar on May 19th, 2011, 3:04am Ouch (http://arimaa.com/arimaa/gameroom/opengamewin.cgi?client=1&gameid=184495&role=v&side=w). I guess I should have made use of my remaining horse, and not worried about it being taken hostage. |
||
Title: Re: Game 184376: Belteshazzar vs. bot_Marwin2010CC Post by ocmiente on May 19th, 2011, 12:34pm I think you had the game well on the way to a win at move 13g if you had moved mc5e Ec4n Mb5s Mb4e, firming up the horse hostage, and moving the silver camel away from the west side so that it can't break the frame easily. I would forget about the rabbit on h6. A good horse frame is worth more than a rabbit. If the frame were solidified, the game would probably be much slower and less exciting than the way this game went, but gold's advantage seems significant to me going this route. Generally, you have to adjust your style of play to what is on the board. Once a horse frame is established, you need to strongly consider going for more of a control type game rather than a game in which you focus on creating greater threats than your opponent's threats. |
||
Title: Re: Game 184376: Belteshazzar vs. bot_Marwin2010CC Post by Belteshazzar on May 19th, 2011, 3:44pm on 05/19/11 at 12:34:48, ocmiente wrote:
I had not considered that, but it makes perfect sense, and I probably would have done that had I seen that option. |
||
Title: Re: Game 184376: Belteshazzar vs. bot_Marwin2010CC Post by Belteshazzar on May 20th, 2011, 10:03am I finally beat it (http://arimaa.com/arimaa/gameroom/opengamewin.cgi?client=1&gameid=184663&role=v&side=w). How could I have won faster? |
||
Title: Re: Game 184376: Belteshazzar vs. bot_Marwin2010CC Post by ocmiente on May 20th, 2011, 4:58pm Congratulations! Well done. Letting the camel be taken hostage is not something I usually go for, but when I do, swarming the opponent's elephant to blockade it, and then rotating out my own elephant is what I look for first. With silvers camel on d7 (move 10g), this seems especially good because usually the biggest threat to that kind of swarm is the opponent's camel doing damage on the other side of the board. Also, with silvers weak cat and rabbit ahead of the hostaged camel, it's likely that they could be used as part of the blockade, saving more of gold's own pieces for attacking the other parts of the board. However, I can't say that that approach would be faster than what you did. For a faster game, you could have begun advancing rabbits earlier and threatening goal sooner, rather than making easy captures. I have a hard time doing that myself, but I'll throw out this possible different move: 39g Hc7e Hd7w rd6n Ra2n Taking the horse on 39g is good, but the horse is not going anywhere, and not capturing it leaves silver's camel and two horses frozen. There are enough pieces in the west to swarm the elephant, so maybe that's a little faster. |
||
Arimaa Forum » Powered by YaBB 1 Gold - SP 1.3.1! YaBB © 2000-2003. All Rights Reserved. |