| « back Browsing Levels: #2634 "Quintuple File (v3)" | |
| |
| |
| prev: #2633 "Tutorial Package 1/3" | next: #2635 " liberez le coeur" |
"Quintuple File (v3)" by radiant (originals by Max and noname) |
|
| ||||||||||||||||||||||
| Comments (turn spoilers off) | ||||
| ||||
| 20129 | Quintuple File (v3) | noname (559) | Tue 18 Jul 2017 05:00 | SPOILER |
| In case anyone is still wondering, my updated solver, which solved this level from the beginning, says that 40 is indeed optimal. | ||||
| 15509 | Quintuple File (v3) | Dave (2400) | Sat 11 Jun 2011 15:41 | SPOILER |
| I agree, it does look unlikely, but... The 18 split seemed optimal until 17 was stumbled on .-) Also, 4 of the 17 moves in stage 1 and 8 of the 23 moves in stage 2 are either net zero or net negative, the rest being made up for in boosts, this gives quite a bit of manoeuvring room. If an improvement existed, I would expect it more of the form 20:19 where part of stage 1 is used to prepare stage 2. To me, the style points in this level come from the counter-intuitiveness of it. I now see that the reason I never solved it before was because it is so hard to tell a winning position from a loser. What at first sight seems the best split position turns out to be the worst and vice versa. I don't think I've seen that to such an extent in any other bot-pushing level. |
||||
| 15508 | Quintuple File (v3) | radiant (1469) | Fri 10 Jun 2011 19:38 | SPOILER |
| A faster split time to part 1 seems unlikely, since the key square is already 16 spaces from start (and even if you're not the one to trigger it, you can't get a bot to go there without going at least as far out of the way anyway). When I stumbled upon the original solution that allowed me to upload the level (18+38; improvable to 17+36), I particularly liked it because it took full advantage of the grid dimensions that had been standard ever since Single File. That aspect is lost in the optimized route, which likely costs the level some style points in the long run. (So does the poor laser mechanism which adds a 19th column to the grid, thereby shrinking the thumbnail.) |
||||
| 15485 | Quintuple File (v3) | Dave (2400) | Tue 07 Jun 2011 01:49 | SPOILER |
| hats off radiant, to come up with this by hand must have taken some serious effort! I gave up in the end and resorted to computer. Given that I am not a programmer, that, for me, really is the last resort. Surprisingly, I found the problem space to be a magnitude smaller than I had allowed for. The 40 speedrun uses the optimal solution for stage 2 if this particular solution for stage 1 is taken as a starting point. It begs the question 'How many stage 1 solutions are there which could potentially lead to a faster overall run?' The speedrun length is split 17:23. So far the best alternatives I have managed to come up with are 18:31 and 19:28. |
||||
| 14899 | Quintuple File (v3) | radiant (1469) | Thu 25 Nov 2010 22:25 | SPOILER |
| Added speedrun: 40 moves (old: 41). | ||||
| 13155 | Quintuple File (v3) | noname (559) | Wed 11 Nov 2009 21:25 | SPOILER |
| Added speedrun: 41 moves (old: 54). Found by computer. Turns out my program had some bugs. After I removed the bugs and improved the strategy, the program found this solution in less than a minute. |
||||
| 13071 | Quintuple File (v3) | noname (559) | Tue 03 Nov 2009 22:32 | SPOILER |
| I tried, but decided to kill the process after a few minutes. In contrast, the brokoban levels were solved in a few seconds. Of course, since I had to include behavior for electricity and trap, I had to rewrite some code, and I don't know whether the problem is with search space or with bugs. | ||||
| 13059 | Quintuple File (v3) | radiant (1469) | Sun 01 Nov 2009 19:36 | SPOILER |
| Has noname tried his Botkoban 2 solver on this? The levels are really quite similar, and I'm sure he's familiar enough with the first half of the puzzle... I guess the search space is a good deal larger, which might cause it to choke. Who knows. |
||||
| 12873 | Quintuple File (v3) | devin (2311) | Wed 30 Sep 2009 07:28 | |
| i cant stay in quintiple file :( | ||||
| 12871 | Quintuple File (v3) | radiant (1469) | Wed 30 Sep 2009 04:17 | SPOILER |
| Added speedrun: 54 moves (old: 56). Always with the 2-move optimizations... |
||||
| 12870 | Quintuple File (v3) | radiant (1469) | Wed 30 Sep 2009 03:53 | |
| 'Quintuple File (v3)' uploaded by radiant: Hmm...two excess wall tiles, then one excess wall tile...I wonder what comes next? |
||||