« back Browsing Levels: #1804 "Botkoban 2" | |
prev: #1803 "All A Round" | next: #1805 "IM IN POLAND" |
"Botkoban 2" by Tom 7 + Computer |
|
|
Comments (turn spoilers off) | ||||
15702 | Botkoban 2 | Jim (2369) | Tue 26 Jul 2011 18:26 | SPOILER |
I finally solved this one, only 3 left in triage now for me (Botkoban 3, Land of 10000 Bots, and of course block cipher which I never intend on trying). | ||||
15688 | Botkoban 2 | radiant (1469) | Tue 12 Jul 2011 16:24 | SPOILER |
That position can actually be reached in 38, but does bot 4 really need to be blocked off like that? | ||||
15687 | Botkoban 2 | Jim (2369) | Sat 09 Jul 2011 21:38 | SPOILER |
I lined up the bots #4, #2, #3 from the top to the bottom in the 3rd column from the left, where I am to the left of bot #3. I will leave it to you to figure out how, but I was able to do this in 43 moves. That is the only way I've been able to get the #1 bot out. This still does not lead to a solution, I have yet to solve this level. | ||||
15269 | Botkoban 2 | N7DOT 3 (2490) | Sun 20 Mar 2011 18:10 | SPOILER |
I'm at noname's stage 2 and I've been able to manipulate the bots but I can't figure out what shape/position I need to have the bots in to get bot #1. Can anyone offer a hint as to where they need to be? Nothing that gives away too much, though. I still want to have to figure it out. | ||||
14892 | Botkoban 2 | Dave (2400) | Thu 25 Nov 2010 16:52 | SPOILER |
Hats off to anyone who solved this by hand. I have to confess that I gave up and solved it by computer having tested the program out on Botkoban 3 first. As an optimisation exercise I found it nowhere near as difficult as Revenge 5 which took about 3 hours of runtime rather than a few seconds. In both cases I did use a hash table to record visited positions with minimum step count. You could get probably away without it here as there were just over 1000 entries when I re-ran it with a 71 move limit imposed. That would still require quite a lot of paper if working this out by hand! I may try it by hand using noname's tip but that's for another day when I have a lot of time on my hands... |
||||
9757 | Botkoban 2 | noname (559) | Sun 30 Dec 2007 18:13 | SPOILER |
wowei, my step 2 probably can be divided furthermore, but I don't really see good subgoals for step 2. When I tried to solve by hand, I did have to write on a piece of paper the bot order of some of the otherwise identical positions, just so that when I actually visit an apparently identical position, I know whether the bot order has changed. As for my program, it's written in Python and so I can use dictionaries instead of worrying about the details of hashing. And my program is written specifically to solve this one and Botkoban 3, so I don't have a move generator for every possible Escape position -- my program only deals with bots, walls, floor and player. Also, starting positions are hardcoded into my program by hand and I didn't write any interface for reading and interpreting esx-files. |
||||
9703 | Botkoban 2 | wowei (685) | Thu 27 Dec 2007 16:06 | SPOILER |
noname, plenty of dead-ends can be reached here, i dont't know about two major branches. i didnt organize the solving really carefully, but without spoiling too much here i think your step 2 needs to be divided further more: there are bots to swap places in otherways identical positions. i guess, your program uses a hashing-mechanism to keep track of already reached positions. btw i still am too lazy to study Tom's sources in depth, did you write some interface for reading and interpreting esx-files, and - more important - a move-generator for any position? |
||||
9616 | Botkoban 2 | noname (559) | Sat 22 Dec 2007 06:15 | SPOILER |
Congratulations, wowei! Anyway, I just tried to solve by hand again and successfully solved it. Of course, since I've seen the computer-generated solution, I probably have some memory about it, so it may not count. One problem with this level is that there is a place where it branches into two branches, where one of them goes quite deep but is fruitless, while the other one is not so deep but leads to a solution. I gave up after deciding that the first branch is dead. Another problem is that there are many repeating positions, and I have to mark down some of the positions which keep repeating. What's wrong when I first tried to solve this is that I marked down ALL intermediate positions. I should have just marked down SOME of the positions. I think I'll divide this level into three stages: 1. Get out of the beginning positions; 2. Get bot 1 out; 3. Get all the bots into positions. I think that stage 2 is the most difficult. Once I'm done with stage 2, it isn't too difficult to finish the level. But I still don't find anything intelligent from this level. This level is still mostly trial and error for me, instead of having to think and plan. |
||||
9600 | Botkoban 2 | wowei (685) | Fri 21 Dec 2007 17:52 | SPOILER |
You can searate it into a few different phases, that are to accomplish. And that helps a lot. That is in opposite to botkoban #3, where i haven't found some discrete stages yet. | ||||
9599 | Botkoban 2 | wowei (685) | Fri 21 Dec 2007 17:45 | SPOILER |
Added solution "Untitledonly ": only a little longer than the original, which i guess is otimal. Aaaand without computer-assistance! No brute-force or fuzzylogic did help here - only Blood Sweat and Tears ;) |
||||
9588 | Botkoban 2 | wowei (685) | Fri 21 Dec 2007 10:53 | |
your danggg server ;) doesnt let me upload a solution | ||||
8895 | Botkoban 2 | noname (559) | Tue 02 Oct 2007 23:20 | SPOILER |
I gave up solving by hand and end up writing a program to solve this. Any person who solved this by hand must be a genius (or have much more patience than me). | ||||
8886 | Botkoban 2 | Tom 7 (1) | Mon 01 Oct 2007 00:55 | |
'Botkoban 2' uploaded by Tom 7: I think this one is substantially harder than the last. |
||||