Author |
Topic: Move 2 (Read 11202 times) |
|
99of9
Forum Guru
Gnobby's creator (player #314)
Gender:
Posts: 1413
|
|
Re: Move 2
« Reply #15 on: Apr 20th, 2007, 7:21am » |
Quote Modify
|
OK, I've finally had a chance to look at it. I know you guys are all thinking far ahead about Jean attacking our horse or h-rabbit, but I'm a bit worried about this move of his: 3w Ee5w Ed5n Ed6n Hb2n From there he could: * Attack our camel * Attack our c-rabbit * Go behind our trap onto b7 threatening either our b-horse or a-rabbit So, unless you have a good response to that (E blockade maybe?)... ... I think we should consider the preventative move: 2b ed7s ed6s hg7s dd8s
|
|
IP Logged |
|
|
|
99of9
Forum Guru
Gnobby's creator (player #314)
Gender:
Posts: 1413
|
|
Re: Move 2
« Reply #16 on: Apr 20th, 2007, 7:57am » |
Quote Modify
|
Ok, just an extra point of information... In my postal against chessandgo, I played Karl's suggestion, and chessandgo replied as follows: 3w Ee5n Ee6w Hb2n Db1n How would you continue against that Karl? My reply sequence was clearly not good enough, as he was soon up by a solid horse hostage.
|
|
IP Logged |
|
|
|
jdb
Forum Guru
Arimaa player #214
Gender:
Posts: 682
|
|
Re: Move 2
« Reply #17 on: Apr 20th, 2007, 8:50am » |
Quote Modify
|
Quote:1." .. but if we put our own horse on g6 to hinder the EH attack, he can use his elephant to attack our horse instead. " Is this attack valid? A horse hostage on the opponent's side is a bad idea. And anyway, is a horse hostage worth the reduction of elephant's mobility when the attacker does not win the strongest free piece (as in a camel hostage)? If the attacker only wants to pull it out, that's not a very serious attack since he would be wasting his initiative doing slow moves that can be undone. 2. ".. if he decentralizes his elephant." That can be found in many posts. Why is this important? When I am afraid of my opponent (which by the way is not a good idea, you may suffer position "erosion") I always count piece by piece if my pieces are more or less in the same columns as his. If he decentralizes his elephant, so would I. But since "the mob" is maybe 2200 Elo decentralizing the phant against us should expose some weakness we can exploit. Is that what you mean? 3. ".. in this game with his rabbit dangling on d1." What's the problem with this? I would understand the setup where that rabbit was in b1 i.o. the dog more easily. I understand he did that to use the dog to control b3 when the horse in b2 has gone hunting. Is that correct? But returning to the rabbit, it will only take part in the game much later. Since there are 8 rabbits, is it a good idea if a lonely rabbit tries a central path i.o. the usual paths on the side? I can imagine pros for this idea: A single advance threat does not work unless it is very well backed up, but many simultaneous threats may work just because the defender cannot defend them all. Also, this rabbit can free a frozen piece even at the price of exposing itself too much. Even if its lost, I may have given enough worries to the opponent to be worth the small sacrifice. |
| Nice questions. I'll try and answer as best I can. Attacking the horse on g6 with gold's elephant is questionable because it does not provide enough compensation for decentralizing the elephant. Taking a horse hostage with the elephant is not worth too much. However, taking the horse hostage with a camel is valuable. In general terms it is not a good idea to decentralize the elephant. The elephant is the only piece that can protect a trap square on its own. When tactics happen, it is very important to be able to position the elephant next to any of the trap squares. Just count how many steps it takes the elephant to reach all the trap squares when its in the centre, compared to on the side. When the elephant is on the side, an alert opponent would be looking for tactical opportunities on the other side of the board. A rabbit on d1 is much easier to drag than a rabbit on c1. On d1, the elephant can start on d3, drag the rabbit and return to d3. On c1, the elephant ends its turn on d2, and if c2 is occupied it has to start its turn on d2 also. An elephant on d2 is vulnerable to a blockade, where on d3 it is not.
|
|
IP Logged |
|
|
|
NIC1138
Forum Guru
Arimaa player #65536
Gender:
Posts: 149
|
|
Re: Move 2
« Reply #18 on: Apr 20th, 2007, 11:04am » |
Quote Modify
|
on Apr 20th, 2007, 8:50am, jdb wrote:Taking a horse hostage with the elephant is not worth too much. However, taking the horse hostage with a camel is valuable. |
| ...And how do you see my proposal of taking the camel to g6? Could it be valuable enough?
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 2
« Reply #19 on: Apr 20th, 2007, 11:32am » |
Quote Modify
|
So many posts, so little time! The one I can't resist answering immediately: on Apr 20th, 2007, 5:34am, JacquesB wrote: 1." .. but if we put our own horse on g6 to hinder the EH attack, he can use his elephant to attack our horse instead. " Is this attack valid? A horse hostage on the opponent's side is a bad idea. |
| To my understanding, the most important open question of Arimaa theory is when and whether it is an advantage to take an opposing horse hostage with one's elephant. This is NOT settled. Let's suppose for a second that it is always a bad idea to take an opposing horse hostage with your elephant. The consequence would be the extinction of lone-elephant attackers like myself. If I tried to attack with my elephant alone, the other player would always launch a more forceful elephant-horse attack, so I would have to come home with my elephant, and I would get stuck holding his horse hostage with my elephant, a disadvantage to me. To avoid this, I would have to try to launch an elehpant-horse attack myself, in order to get in the first shot. But there are two ways in which the elephant holding the horse hostage can make it into an advantage. Either the horse hostage can be framed, or it can be passed off to a friendly camel, freeing the elephant to do damage elsewhere. An example of the horse frame happened in the ongoing postal game Brendan vs. chessandgo. Brendan gave chessandgo a free horse hostage, maybe erroneously thinking it was not a disadvantage. Chessandgo framed that horse for a clearly superior position, which has converted to a material advantage of two horses for a camel. An example of camel-holding-horse is the ongoing postal game arimaa_master vs. Fritzlein. I went after his attacking horse with my elephant, and eventually got that horse as a hostage of my camel. This was clearly worth more than the rabbit pull he got. Indeed, it is possible (and we used to all think it was true) that taking a horse hostage with your elephant is always a good idea. If this is the reality, then it isn't the lone-elephant attackers that will die out, it will be the elephant-horse attackers. They will get into trouble every time they expose a horse, and eventually they will get tired of always being at a disadvantage, and will switch to more conservative play. My current belief is that the truth is somewhere in between. Sometimes an elephant holding a horse hostage is good, and sometimes it is bad. This is closely related to my belief that sometimes an elephant-horse attack is sound, and sometimes it is unsound. Chessandgo's current game against Brendan should be a caution to those who think that chessandgo will be thrown off balance by having to defend against an elephant-horse attack instead of being able to launch his own elephant-horse attack. On the contrary, he can play just fine on either side of the situation. Quote:2. ".. if he decentralizes his elephant." That can be found in many posts. Why is this important? |
| A centralized elephant can attack or defend at any of the four traps. A decentralized elephant can attack or defend only at two. Sometimes the difference matters and sometimes it doesn't. After seanick's move and the response I feared, the more important consideration is not whether he can operate on the other wing, so much as whether can win the race to pull something. An additional subtle point that if one elephant is decentralized, the opposing camel gains additional freedom of movement. That doesn't seem to matter in the present opening, though. Quote:3. ".. in this game with his rabbit dangling on d1." What's the problem with this? |
| As jdb says, it saves us a few steps in the dual-lone-elephant rabbit-pulling race. In fact, it might tip the balance of that race enough that chessandgo can't afford to let the game become a dual-lone-elephant rabbit-pulling race. There are some heretics that don't mind advanced rabbits in the opening, most notably blue22. However, the large majority of top players would consider any rabbit advance this early to be a disadvantage. The advantages in terms of unfreezing pieces don't matter until some pieces (besides elephants and horses) are exposed. The advantages in terms of goal threats don't occur until much later. Quote:4. .. Predictions from 2w to 5b: I'm really impressed! |
| Don't worry, our opening theory will be exhausted after one more move at most, and then we will have no idea what chessandgo will do.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 2
« Reply #20 on: Apr 20th, 2007, 11:45am » |
Quote Modify
|
on Apr 20th, 2007, 11:04am, NIC1138 wrote: ...And how do you see my proposal of taking the camel to g6? Could it be valuable enough? |
| That camel on g6 wouldn't get a chance to take a horse hostage, because chessandgo wouldn't advance his g3 horse into the teeth of our camel. Instead chessandgo could advance his other horse up the b-file with no fear, because our camel is so far away. This consideration (not danger to our camel) is the primary reason I would recommend against your proposed move. Chessandgo once told me that one of his main objectives in the opening is to make the opposing camel commit to one wing, so that he can safely launch an elephant-horse attack on the other wing. In that context I would advise against voluntarily decentralizing our camel, unless it leads to some attacking possibilities.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 2
« Reply #21 on: Apr 20th, 2007, 1:11pm » |
Quote Modify
|
on Apr 20th, 2007, 1:48am, NIC1138 wrote:Just charge with the phant already: 2b ed7s ed6s ed5s ed4s. |
| The way I calculate it the results of the bloodbath is: 2b Silver elephant attack 3w Gold elephant-horse attack 3b Silver captures a cat 4w Gold captures a dog 4b Silver captures a horse 5w Gold captures a camel Seems to me we get the short end of the deal, even though we get to make the first capture.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 2
« Reply #22 on: Apr 20th, 2007, 1:51pm » |
Quote Modify
|
on Apr 20th, 2007, 7:57am, 99of9 wrote:Ok, just an extra point of information... In my postal against chessandgo, I played Karl's suggestion, and chessandgo replied as follows: 3w Ee5n Ee6w Hb2n Db1n How would you continue against that Karl? My reply sequence was clearly not good enough, as he was soon up by a solid horse hostage. |
| If he puts his elephant between our traps as he did against you 2b ed7s ed6s ed5s hg7s 3w Ee5n Ee6w Hb2n Db1n then I don't think we should try to win a race to pull something. I would probably play the defensive 3b ed4n hb7s xxxx xxxx and I would NOT decentralize the camel until I had to. With our elephant near home and no central rabbits, we can probably prevent his lone elephant from getting anything, even a rabbit pull. If he attacks with a horse as well, we are in a better position to respond in a way that hurts him than if our own elephant is off hunting rabbits. If he just diddles around, we can find ways to slowly improve our position and eventually threaten an attack of our own. on Apr 20th, 2007, 7:21am, 99of9 wrote:I'm a bit worried about this move of his: 3w Ee5w Ed5n Ed6n Hb2n |
| Actually that move looks more critical to me too, but it turns out the threat to our camel is illusory. For example, after 2b ed7s ed6s ed5s hg7s 3w Ee5w Ed5n Ed6n Hb2n 3b ed4n ed5n hb7s cc7w He can't pull our rabbit without getting blockaded, so... 4w me7s Ed7e me6s Ee7s 5b ed6s me5e ed5e xxx And suddenly his elephant needs to scramble to prevent our camel from getting his g3 horse hostage. I learned that 5b maneuver from chessandgo himself... In any event, your prophylactic move 2b ed7s ed6s hg7s dd8s doesn't look bad, I just don't think it is necessary, and I think our elephant on d4 gives us better options against his lone-elephant-attacks-our-flank strategies.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 2
« Reply #23 on: Apr 20th, 2007, 2:06pm » |
Quote Modify
|
on Apr 19th, 2007, 3:45pm, jdb wrote:What if gold plays 4w Me2n De1n Rd1e Hb2n ? It threatens to bury, but not quite blockade the silver elephant. |
| My general experience has been that pieces which try to threaten to blockade an opposing elephant without the help of the friendly elephant will just get themselves in trouble. After 2b ed7s ed6s ed5s hg7s 3w Hg3n Hg4n Hg5e Hh5n 3b rh7w ed4s Md2e ed3s 4w Me2n De1n Rd1e Hb2n My instinct says there must be some way to take advantage of the exposed gold pieces. What about 4b ed2n Me3n ed3e hb7s threating to capture a piece or harass the camel? Note that if his elephant leaves e5, our camel can zip to g6 for a free horse hostage. There are lots of possible lines in a wild opening like that, but my gut feel is that we stand no worse and probably better. That's not even counting our other options on 3b. I think 3w in this line is dubious and basically throws away the first-move advantage, so I don't expect to see it.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 2
« Reply #24 on: Apr 20th, 2007, 2:28pm » |
Quote Modify
|
Incidentally, let me give my justification of 2b ed7s ed6s ed5s hg7s, rather than just poking at other moves. I suggest the move as a defense against 3w Ee5e Ef5e Eg5w hg6s. If our elephant is on d4, we can play 3w ed4e ee4e ef4e hg5n, and Gold has lost the advantage of the opening move. Therefore 2b ed7s ed6s ed5s hg7s prevents 3w Ee5e Ef5e Eg5w hg6s. Other moves such as 2b ed7s ed6s hg7s hb7s (somewhat common in the postal tournament) 2b ed7s ed6s hg7s dd8s (99of9) 2b ed7s hg7s df7e me7e (seanick) all permit 3w Ee5e Ef5e Eg5w hg6s I can hear the collective cries of, "So what?" Who cares if he pulls our horse? Well, in my (rapidly-shrinking) world of defensive play, this horse hostage is one that gives a slight advantage to the hostage-taker. I don't want chessandgo to drag our horse home to frame it and/or pass it off to his camel. Some may say that, even if grabbing the horse would be a slight advantage to chessandgo, we needn't worry about it, because there is no way he will take a horse hostage rather than starting an elephant-horse attack. I'm not sure that is true, and in any case it is bad practice to count on your opponent not to make the best move. If we permit chessandgo to play 3w Ee5e Ef5e Eg5w hg6s, then it should be because we believe it is not dangerous. Another objection could be, "Why should we listen to you if chessandgo is beating you in the postal tournament?" Er... well... um... maybe I should spend some time on that game instead of posting endlessly here.
|
|
IP Logged |
|
|
|
NIC1138
Forum Guru
Arimaa player #65536
Gender:
Posts: 149
|
|
Re: Move 2
« Reply #25 on: Apr 20th, 2007, 9:52pm » |
Quote Modify
|
on Apr 20th, 2007, 2:28pm, Fritzlein wrote:(...)it is bad practice to count on your opponent not to make the best move. If we permit chessandgo to play 3w Ee5e Ef5e Eg5w hg6s |
| Why is 3w Ee5e Ef5e Eg5w hg6s much better than ...Eg5s hg6s or Eg5e hg6s (my personal favourite)?... Is it the central phant thing?... Is this why I'm not a 1700RU player yet?
|
|
IP Logged |
|
|
|
The_Jeh
Forum Guru
Arimaa player #634
Gender:
Posts: 460
|
|
Re: Move 2
« Reply #26 on: Apr 20th, 2007, 10:32pm » |
Quote Modify
|
Quote:Why is 3w Ee5e Ef5e Eg5w hg6s much better than ...Eg5s hg6s or Eg5e hg6s |
| Eg5w hg6s puts Gold in a position to push the horse to the f3 trap from behind (hg5s Ef5e), making saving it much harder for Silver because the horse itself will be blocked from retreating straight north, and rescuing pieces will be blocked from reaching the horse. Eg5e hg6s threatens to do the same thing, but the Gold elephant is taken far from the action. It needs to be nearer the center so it can have influence over all traps, namely f6, and not be in danger of losing maneuverability near the side of the board. Using Eg5s hg6s takes away some of his anterior offensive pressure, and it is less ergonomical for the Gold elephant to keep ahold of the horse.
|
« Last Edit: Apr 20th, 2007, 10:41pm by The_Jeh » |
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: Move 2
« Reply #27 on: Apr 21st, 2007, 9:00am » |
Quote Modify
|
on Apr 20th, 2007, 9:52pm, NIC1138 wrote: Why is 3w Ee5e Ef5e Eg5w hg6s much better than ...Eg5s hg6s or Eg5e hg6s (my personal favourite)?... Is it the central phant thing?... Is this why I'm not a 1700RU player yet? |
| 3w Ee5e Ef5e Eg5s hg6s seems rather passive. We can unfreeze and retreat our horse and then what does Gold have? 3w Ee5e Ef5e Eg5e hg6s, your favorite, is definitely a move to worry about. Yes, we can easily unfreeze and retreat our horse (more easily than after 3w Ee5e Ef5e Eg5w hg6s, because our elephant can approach from the side), but then the Gold elephant just pulls our h7 rabbit, for a lead in the rabbit-pulling race. I'm not sure what to do then.
|
|
IP Logged |
|
|
|
jdb
Forum Guru
Arimaa player #214
Gender:
Posts: 682
|
|
Re: Move 2
« Reply #28 on: Apr 21st, 2007, 9:39am » |
Quote Modify
|
For the sake of discussion, lets assume Gold is hoping to play this move. (He might not, but I can only think about so much at once...) 3w Ee5n Ee6w Hb2n Db1n So if silver plays E forward three on his second move: 2b ed7s ed6s ed5s hg7s 3w Ee5n Ee6w Hb2n Db1n 3b ed4n hb7s xxxx xxxx I'm not aware of a term for this manouver, so I'll call it "capping the elephant" Quote:... I think we should consider the preventative move: 2b ed7s ed6s hg7s dd8s |
| 2b ed7s ed6s hg7s dd8s 3w Ee5n Ee6w Hb2n Db1n This gets us to the same sort of position, using fewer steps. Another possibility is 2b ed7s ed6s hg7s hb7s This allows silver's camel a little more breathing room.
|
« Last Edit: Apr 21st, 2007, 9:39am by jdb » |
IP Logged |
|
|
|
NIC1138
Forum Guru
Arimaa player #65536
Gender:
Posts: 149
|
|
Re: Move 2
« Reply #29 on: Apr 21st, 2007, 4:51pm » |
Quote Modify
|
I know you are still not confident on my powers to dream up awkward and naļve moves , so I came up with another lone-elephant candidate... We can avoid the blood bath of 2b ed7s ed6s ed5s ed4s by simply playing 2b ed7s ed6s ed5s ed4e. In this position, the phant threatens the lonesome cat at his eastern trap if he gets down for the E-H attack. But it doesn't mean that we would do this. Instead, in case he comes down (or better, north) with 3w Hg3n Hg4n Hg5n Ee5n, we can simply answer with 3b ee4n ee5e ... . Perhaps 3b ee4n ee5e hb7s dd8s. Now our trap becomes safe, his horse is in a bad position (can't just run back south), and the gold elephant is even perhaps half-capped!... It's a very good lure, IMHO!... How do you like it? It's a mix of Fritz's and jdb's defensive elephant coming back, and my own joyous experience in kidnapping Jean's horse in that game I told you all...
|
« Last Edit: Apr 21st, 2007, 4:52pm by NIC1138 » |
IP Logged |
|
|
|
|