• Welcome to SpeedSolving.com — the world’s largest puzzle community!

    You’re currently browsing as a guest, which means you have limited access to discussions, resources, member profiles, and community features.

    Join 50,000+ cubers and puzzle enthusiasts from around the world to ask questions, share solves, improve faster, and be part of the community.

    Registration is free, fast, and easy — create your account today.

    Already a member? Log in here to hide this message and start participating.

2x2x2, X-cross and 3x2x2 move count?

JL58

JL58

Regular
Joined
Dec 15, 2008
Messages
127
I am getting bored with regular Fridrich on 3^3. I am starting to play around with block building and X-cross. I found it very interesting. My move count, however, is erratic at best. I was wondering if anyone has computed the average and upper bound move counts for building:

- an X-cross
- a 2x2x2 block
- a 3x2x2 block

I am looking for this under 2 conditions: color agnostic solve and using a pre-defined color for the cross or the blocks. So I am asking for 12 numbers in total.
 
cuBerBruce said:
I don't know if values are known for when choosing among all possible 2x2x2 blocks
Click to expand...
I solved 3 million scrambles a couple of years ago (link), and the average was 4.870 moves. If I remember correcrly, I used 25 move scrambles, and the move count was very slightly higher for longer or perfect scrambles.

Worst case is 8 moves (link).

For color-fixed XCross, average is 7.976 and worst case 10 moves (link). Note that this is using the same cross *and* same ce-pair every time.

I'll do some averages for color agnostic XCross and 2x2x3 and post here later.
 
Last edited:
Johannes91 said:
cuBerBruce said:
I don't know if values are known for when choosing among all possible 2x2x2 blocks
Click to expand...
I solved 3 million scrambles a few years ago (link), and the average was 4.870 moves. If I remember correcrly, I used 25 move scrambles, and the move count was very slightly higher for longer or perfect scrambles.

Worst case is 8 moves.

For color-fixed XCross, average is 7.976 and worst case 10 moves (link). Note that this is using the same cross *and* same ce-pair every time.

I'll do some averages for color agnostic XCross and 2x2x3 and post here later.
Click to expand...

What's a perfect scramble?
 
Robert-Y said:
What's a perfect scramble?
Click to expand...
One that gives every position an equal probability. Doing a finite number of random moves doesn't, generating a random position does.

No need to quote the whole post when you're only replying to a small part.
 
Last edited:
Johannes91 said:
Robert-Y said:
What's a perfect scramble?
Click to expand...
One that gives every position an equal probability. Doing a finite number of random moves doesn't, generating a random position does.
Click to expand...

Hmmm.... How many perfect scrambles are there? Also, can you give me an example of a perfect scramble, just to make sure I understand what you mean please?
 
I should've said "perfect scrambling method" instead of "perfect scramble".

Scrambling method #1: Start with a solved cube and make N random moves.
Some positions are more probable than others. The differences get smaller as N gets larger.

Scrambling method #2: Generate a random position.
Assuming the random number generator is good, this gives every position an equal probability.

Scrambling has been discussed many times here and in yahoo groups, search if you're interested.

Yay, 6 off-topic posts in a row.
 
Last edited:
Code: 
                      | Average | Worst  | Scrambles
----------------------+---------+--------+-----------
XCross color neutral  | ~6.536  | [9,10] | 2^26
XCross opposite cross | ~6.986  |   10   | 2^28
XCross fixed cross    | ~7.354  |   10   | 2^29
XCross fixed          |  7.976  |   10   | n/a
----------------------+---------+--------+-----------
2x2x3 color neutral   | ~7.991  |   11   | 2^21
2x2x3 fixed           |  9.116  |   12   | n/a
----------------------+---------+--------+-----------
2x2x2 color neutral   | ~4.916  |    8   | 2^28
2x2x2 fixed           |  6.034  |    8   | n/a
----------------------+---------+--------+-----------
1x2x3 color neutral   | ~5.144  |  [8,9] | 2^26
1x2x3 LD|RD|LU|RU     | ~5.943  |    9   | 2^27
1x2x3 LD|RD           | ~6.306  |    9   | 2^28
1x2x3 fixed           |  6.701  |    9   | n/a
----------------------+---------+--------+-----------
1x2x2 color neutral   | ~2.671  |    5   | 2^25
1x2x2 fixed           |  4.360  |    5   | n/a
----------------------+---------+--------+-----------
EOLine color neutral  | ~5.006  |    9   | 2^25
EOLine fixed          |  6.127  |    9   | n/a

- All move counts are in HTM/FTM.
- I used random state scrambling.
- Color neutral numbers are only approximations.

Thanks to Bruce for the fixed color 2x2x3 numbers.

Edit 2009-06-04: Added 1x2x3 block a.k.a. Roux block.
Edit 2009-07-06: FYI, the distributions are at the end of this post.
Edit 2009-08-01: Added two partially neutral 1x2x3 blocks and 1x2x2 block.
Edit 2009-08-09: Added EOLine.

Code: 
(total positions, mean, list of (depth, positions at depth) pairs)

XCross
(67108864,6.535903126001358,[(0,20),(1,315),(2,3691),(3,40391),(4,429582),(5,3973663),(6,24007367),(7,36373364),(8,2280468),(9,3)])
(268435456,6.986002378165722,[(0,27),(1,440),(2,4895),(3,56666),(4,620074),(5,6130185),(6,47023196),(7,157020549),(8,57518550),(9,60874)])
(536870912,7.353692278265953,[(0,20),(1,441),(2,5044),(3,56557),(4,624256),(5,6206267),(6,50307192),(7,232710047),(8,239188357),(9,7772726),(10,5)])
(72990720,7.975721804086875,[(0,1),(1,15),(2,172),(3,1950),(4,21535),(5,220368),(6,1989591),(7,13431990),(8,40963892),(9,16325184),(10,36022)])
2x2x3
(2097152,7.990720748901367,[(1,2),(3,36),(4,358),(5,3838),(6,40971),(7,350772),(8,1274997),(9,425962),(10,216)])
(1532805120,9.115899440628173,[(0,1),(1,12),(2,141),(3,1746),(4,20935),(5,243092),(6,2698935),(7,27258179),(8,216204042),(9,830686751),(10,453825501),(11,1865784),(12,1)])
2x2x2
(268435456,4.915635708719492,[(0,8271),(1,74990),(2,757290),(3,7061775),(4,52689991),(5,161090773),(6,46724373),(7,27993)])
(253440,6.033834438131313,[(0,1),(1,9),(2,90),(3,852),(4,7169),(5,44182),(6,131636),(7,68940),(8,561)])
1x2x3
(67108864,5.144310295581818,[(0,1192),(1,10717),(2,111293),(3,1126409),(4,8975643),(5,35710034),(6,21051484),(7,122092)])
(134217728,5.9427317678928375,[(0,400),(1,3575),(2,37944),(3,399141),(4,3678904),(5,25824316),(6,77456888),(7,26767934),(8,48626)])
(268435456,6.306151166558266,[(0,365),(1,3679),(2,37788),(3,399662),(4,3733678),(5,28185912),(6,120408288),(7,112126360),(8,3539713),(9,11)])
(5322240,6.700925174362674,[(0,4),(1,36),(2,376),(3,3976),(4,37140),(5,289044),(6,1495720),(7,2885332),(8,609716),(9,896)])
1x2x2
(33554432,2.6706851422786713,[(0,248598),(1,1443471),(2,9779999),(3,19719644),(4,2362678),(5,42)])
(12672,4.360164141414141,[(0,4),(1,24),(2,208),(3,1388),(4,5104),(5,5432),(6,512)])
EOLine
(33554432,5.006114274263382,[(0,1496),(1,12273),(2,112415),(3,967997),(4,6226405),(5,17516431),(6,8673453),(7,43962)])
(270336,6.126535126657197,[(0,1),(1,9),(2,91),(3,851),(4,6831),(5,41703),(6,130239),(7,88683),(8,1927),(9,1)])
 
Last edited:
you shouldnt completely devote yourself to one single method and just keep your options open and just use X-cross when its convenient

expecially if your not color netural.

it would help though it you could to opposite color crosses, when you could build the 2x2x2 block anywhere and just either finish the cross one of the 2 sides, rather then all 6 sides
 
Johannes, what are the numbers for solving X-cross on either of two opposite colors, with preference of color being always toward the easiest of the two, and not fixing the slot? I have a theory that this is when using Xcross is the most practical, and this is how I use X-cross. I'd be curious to see the data behind this particular approach.

Chris

--edit--
Using this approach I feel I should mention that I start by solving an X-cross between 40%-50% of my solves. I feel such a strategy allows you to use Xcross frequently (decently high probability to have a somewhat easy case) without having to learn cases or study patterns via memorization. I still do my Xcross solving intuitively and by block building.
--edit--
 
cmhardw said:
Johannes, what are the numbers for solving X-cross on either of two opposite colors, with preference of color being always toward the easiest of the two, and not fixing the slot?
Click to expand...
Worst case is 10 moves, for example superflip with swaps (UF UB) and (DF DB). Average seems to be a bit below 7.
 
Some more data than just average and worst case:
OMG I'm sorry for bumping a 2-week-old topic! I don't even have a question!
XCross. If you could easily find 6-movers but rarely 7-movers, with full color neutrality you could use XCross in ~42% of your solves; with opposite cross, ~20%; and with fixed cross, only ~11%.
XCross.png


2x2x2 block.
2x2x2.png


2x2x3 block.
2x2x3.png


First Roux block a.k.a. 1x2x3 block. The difference is quite big because there are 24 options a color neutral cuber can choose from. AFAIK, many cubers are partially neutral when using Roux.
1x2x3.png
 
Johannes91 said:
Some more data than just average and worst case:
OMG I'm sorry for bumping a 2-week-old topic! I don't even have a question!
XCross. If you could easily find 6-movers but rarely 7-movers, with full color neutrality you could use XCross in ~42% of your solves; with opposite cross, ~20%; and with fixed cross, only ~11%.
[images]
Click to expand...
I am not quite understanding the data. Shouldn't color neutrality always give you the highest percentage for each number of moves? For example, for the XCross plot, at 7 moves, opposite color has a higher percentage than color neutral. This doesn't quite make sense because color neutral includes opposite color, so the percentage of color neutral should be at least the percentage of opposite color.

Edit: Ah, it is based on percentage of all moves, and not just each move individually. Obvious now.
 
Last edited:
pjk said:
Johannes91 said:
Some more data than just average and worst case:
OMG I'm sorry for bumping a 2-week-old topic! I don't even have a question!
XCross. If you could easily find 6-movers but rarely 7-movers, with full color neutrality you could use XCross in ~42% of your solves; with opposite cross, ~20%; and with fixed cross, only ~11%.
[images]
Click to expand...
I am not quite understanding the data. Shouldn't color neutrality always give you the highest percentage for each number of moves? For example, for the XCross plot, at 7 moves, opposite color has a higher percentage than color neutral. This doesn't quite make sense because color neutral includes opposite color, so the percentage of color neutral should be at least the percentage of opposite color.
Click to expand...

However, if you look at 6 and 5 moves, there are many, many more possibilities for fully color neutral. The percentage of remaining cases for color neutral at 7 moves is higher, but not the gross, because there are simply very few 5 and 6 solves for opposite cross.

If that graph is to scale, then it seems that over 90% of color neutral x-crosses can be done in 7 moves or less, but a considerably smaller percentage can be done with opposite cross in 7 moves or less.

Of course, the graph could be inaccurate, but I think that makes sense.
 
Last edited:
Johannes91, Kian and all, I want to thank you for your responses. Very informative. I loved the opposite cross data - a favorite (it seems) of Erik the half God.

I am still baffled by the results. I wonder how many of the CFOP disciples are close enough to find the optimum Xcross within normal inspection time.

This data is encouraging. Working on Xcross would save 25% of the pair insertion time with almost no move cost and is therefore worth a lot.

In general I find CFOP (like probably all other methods) very compelling as progress relies on shifting the automatic moves to the more creative steps (algorithmic pair insertion to intuitive, 4 pair insertions to Xcross, etc.) - obvious to many of you, a learning for the rest of us.

Thanks again for your support.
 
You must log in or register to reply here.

Similar threads

DYGH.Tjen
3x3x3 cross and 2x2x2 first LAYER (Layer, not face)
Replies
2
Views
1,912
y235
y235
B
2x2x2 Domino Reduction example solve thread
Replies
7
Views
598
acrystal
acrystal
D
  • Locked
New Method or this Method exist? Block 2x2x2 in 3x3
Replies
5
Views
554
Deleted member 27249
D
PhanPhuocan
Is there a cube explorer for 2x2x2 or 4x4x4 puzzles+?
Replies
3
Views
753
John Williams
J
abunickabhi
2x2x2 football
Replies
0
Views
376
abunickabhi
abunickabhi
Back
Top