SwitchTile - A New Puzzle for your Touchscreen Device [Android]

5x5 (touchscreen)

number of times: 50/50
best time: 10.30
worst time: 22.95

current mo3: 15.45 (σ = 2.75)
best mo3: 12.34 (σ = 1.11)

current avg5: 14.11 (σ = 0.29)
best avg5: 12.80 (σ = 0.41)

current avg12: 14.54 (σ = 1.89)
best avg12: 13.90 (σ = 1.97)

session avg: 15.19 (σ = 1.86)
session mean: 15.25

I've also got an 8.94 single :)
 
Current pb's (will keep updated):

3x3 - 0.000 single (1 move lol), or 0.134 single (2 moves = 14.93 tps), 0.829 avg5 [(0.664) (1.466) 0.726 0.796 0.964], 1.316 avg12, 1.884 avg100
4x4 - 3.338 single, 4.884 avg5, 6.009 avg12, 7.270 avg100
5x5 - 10.116 single, 12.292 avg5, 14.108 avg12, 15.943 avg100
6x6 - 19.567 single, 23.524 avg5, 26.121 avg12
7x7 - 37.586 single, 39.038 avg5, 43.127 avg12
8x8 - 53.644 single, 57.327 avg5
9x9 - 1:14.217 single, 1:24.332 avg5
10x10 - 1:31.733 single, 1:43.166 avg5
12x12 - 2:49.320
14x14 - 4:16.957
20x20 - 9:30.639
30x30 - 26:20.364
99x99 - 1823:21.123
132x132 - 2997:51.582
3x3 marathon - 2:27.554
3-4-5 relay - 23.661 single, 25.618 avg5, 27.011 avg12
3-4-5-6-7 relay - 1:29.958 single, 1:39.161 avg5
 
Last edited:
Method idea for large puzzles: http://i.imgur.com/wspn23r.png

Basically, instead of solving the puzzle directly, you'd build the left shape from the bottom up, leaving one column open for speed. Instead of solving green/blue and black/yellow directly you can just put any green/blue pieces in that place and then solve the rectangular groups later. Then you do the transformations in the image to actually solve it. I haven't really tried this yet but it was an interesting idea I had.

EDIT: 12x12 in 2:31.735, 590 moves. I left one column open, solved blue/green directly, filled in the column, and then finished off black/yellow as a block at the end.
 
Last edited:
VcZvOox.png
So uh.

That was fun.

Total solve time was approximately 6 hours, 15 minutes, and 30.548 seconds, using my own method.
 
qqwref said:
Wow, that's really fast. What IS your method?
Click to expand...

Solve a center block with edges about 1/3 the size of the puzzle. Next, solve everything directly below that block, leaving an upside-down U unsolved. Then, fill in either the left or right side. Finally, finish the last two "slices" one layer at a time, alternating between them. A little hard to understand, perhaps some pictures will help.

http://imgur.com/a/PucfM/all

I didn't realize Cale already had UWR for size, so I feel like I wasted a lot of time now. XD Maybe I'll go for 128x128 next. I tried that once but Firefox crashed before I finished. :(
 
Last edited:
qqwref said:
That's a pretty interesting method. I like how you solve entire blocks of color at once. And yeah, too bad you only tied the size UWR, but the size allows 3 digits now :D
Click to expand...

Yeah. I kinda used a completely random method for a 69x69 solve I did, but afterwards I made it more structured and it turned into the method I use now. I still haven't been able to beat your 12x12 record though. :P My method works best for bigger puzzles.
 
Deathranger999 said:
Solve a center block with edges about 1/3 the size of the puzzle. Next, solve everything directly below that block, leaving an upside-down U unsolved. Then, fill in either the left or right side. Finally, finish the last two "slices" one layer at a time, alternating between them. A little hard to understand, perhaps some pictures will help.

http://imgur.com/a/PucfM/all

I didn't realize Cale already had UWR for size, so I feel like I wasted a lot of time now. XD Maybe I'll go for 128x128 next. I tried that once but Firefox crashed before I finished. :(
Click to expand...
That's a cool method, I started using it for the 132x132: http://i.imgur.com/DDX3q54.png
Edit: Second step finished http://i.imgur.com/vIGGeTn.png 44x88 block solved (3,872 pieces), 2/9 done
 
Last edited:
Deathranger999 said:
Dammit, I should've known this would happen! Ugh I guess I'll have to go even higher now. :/ THANKS CALE.
Click to expand...
If you do anything higher than 132, I won't even try to beat it.

Current 132x132 progress:
http://i.imgur.com/6GFp3Ik.png
about 22 hours have passed since I started
lower-left 88x96 block is solved
8,448 pieces solved out of 17,424 (not counting random pieces that happen to be solved)
48.48% done
 
Cale S said:
If you do anything higher than 132, I won't even try to beat it.

Current 132x132 progress:
http://i.imgur.com/6GFp3Ik.png
about 22 hours have passed since I started
lower-left 88x96 block is solved
8,448 pieces solved out of 17,424 (not counting random pieces that happen to be solved)
48.48% done
Click to expand...

K. 150x150 it is then.

Also, I hope you notice this, because you're doing this step wrong. You alternate between doing a column on the side and doing a row on the top. It looks like you're doing all the columns on the side first, which I think will end up being slower.
 
Deathranger999 said:
K. 150x150 it is then.

Also, I hope you notice this, because you're doing this step wrong. You alternate between doing a column on the side and doing a row on the top. It looks like you're doing all the columns on the side first, which I think will end up being slower.
Click to expand...

That happened because I started solving the right block instead of the left, but I realized that all the yellow pieces were on the left side because of the way I did the second step, and then I switched to the left block. To balance out the columns I have already solved, I'm going to start by adding 8 rows on top, then alternating between side columns and top rows.
 
You must log in or register to reply here.

Similar threads

MineKB
69x69 SwitchTile Stream
Replies
8
Views
2,070
Deathranger999
D
Jteddy
Speedsliding 2026
Replies
8
Views
162
pi=3.14
pi=3.14
Puzzle Maniac
4-Pad Results Thread
Replies
3
Views
289
acrystal
acrystal
Puzzle Maniac
[OcR] Andrew Feng 34.53 5x5 Single
Replies
11
Views
515
Puzzle Maniac
Puzzle Maniac
C
Cubelelo Daily Calendar Puzzle - September solutions now up
Replies
1
Views
193
abunickabhi
abunickabhi
Back
Top