cage/cage variants sub minute average potential?

rachmaninovian said:
damn i can't run ksolve still...

This is what I've tried..
C:\Documents and Settings\Nicholas Ho\Desktop\cube\ksolve\ksolve 2x2.def 2x2test.txt

Then I get an error saying Pruning tables not found on file, computing. Can't open scramble file!
I guess i'm really dumb. xD

anyway..These few days will just be spent finding better algs for edge cases...i'm pretty much giving up on the centres system; for I see no future in it. After all it'll only save me 1 second or less per step..I rather work on lookahead..and clever 3cycling..and following the advice of blah...optimal commies will probably take too long to master...and wouldn't make too much of a difference xD
Click to expand...

In the short run, you will probably get more satisfaction out of what you are currently doing.

In the long run, solving the remaining centers in exactly 3 algorithms every solve (e.g., solve one center, solve opposite center, solve remaining two centers) will probably be required for significant drops in time.

Solving in exactly 4 algorithms might also be workable (get one center filled with its own color cubies or opposite center cubies, do the same with the opposite center, solve one pair of opposite faces, solve the other pair of opposite faces)

There might not be much point in exploring these possiblities without the use of a search program.
 
Solving in exactly 4 algorithms might also be workable (get one center filled with its own color cubies or opposite center cubies, do the same with the opposite center, solve one pair of opposite faces, solve the other pair of opposite faces)
Click to expand...

Now that I think about it....

This would be a reasonable number of algorithms to find/learn (already known). You wouldn't need very many algorithms for step 1 or 2, since you would not need algorithms which take cubies from the opposite face (saved for step 3 and 4), just algorithms involving the 2 adjacent faces.

So....

The Algorithms for Steps 1 and 2:

0 center cubies solved/sorted

a. 3 on one face, 1 on the other

r2 U2 F' r E2 r' E2 F U2 r2

b. 2 adjacent on one, 2 adjacent on the other

r2 U2 r' E2 r U2 r2 E2

c. 2 adjacent on one, 2 diagonal on the other
d. 2 diagonal on one, 2 diagonal on the other

u2 r2 U2 r' E2 r U2 r2 d2

e. 4 on one face

r' E R2 E' r l E R2 E' l'

1 center cubie solved/sorted

a. 3 cubies on one face

r' E R2 E' r l E R E' l' E R E'

b. 2 adjacent on one face, 1 on other face

r2 U r' E2 r U' r2 E2

c. 2 diagonal on one face, 1 on other face

2 adjacent center cubies solved/sorted

a. 2 adjacent on one face

r' E R2 E' r E R2 E'

b. 2 diagonal on one face
c. 1 on each face

r2 U2 r' u2 r U2 r2 u2

2 diagonal center cubies solved/sorted

a. 2 adjacent on one face
b. 2 diagonal on one face
c. 1 on each face

r2 U r' u2 r U' r2 u2

3 center cubies solved/sorted

a. 1 on adjacent face

r U l' U' r' U l U'

The Algorithms for Steps 3 and 4:

Solve 2 opposite faces

a. 3 solved on each face

r2 U' l2 U r2 U' l2 U

b. 2 adjacent solved, 2 adjacent solved

u2 r2 u2 r2

c. 2 adjacent solved, 2 diagonal solved
d. 2 diagonal solved, 2 diagonal solved
e. 1 solved on each face
f. none solved (full center swap)

r2 E2 r2 E2

21 algorithms to find/already know, maximum of 4 steps to solve all centers, guaranteed, every time, no special cases to worry about.

I am not even sure if some of the 0 cubies solved/sorted cases are really possible/necessary (i.e., if one center had 0 solved/sorted, and the opposite face had any unsolved/unsorted, that would leave a minimum of 3 on at least one of the adjacent faces).

Toss in a few spare algorithms which affect all four centers, and you would have a very high probability of being able to solve in 3 steps or less.
 
Last edited:
mmm..with this muscle injury as of hand I can't cube...so I generated more algs for bad cases for the bottom 2 edges..=D the average number of moves should be at max 9, with 4 moves being the minimum and 12 moves being the maximum (excluding solved state duh). the median should be about 8 moves.

the average number of moves for the last 2 edges is 10.54...so that gives me a sub 20 move average which can yield sub 10s times...I believe that a 2.5-3 tps is still achievable so the limit of my system for middle edges would be about 6-7s. or lesser because the minimum number of moves for unlucky cases is 9 xD
 
rachmaninovian:

I have tried to access your site from two different computers now, and both freeze up whenever I try to see any of the pages with java cubes on them.

I can see the alg you use to swap a single center cubie between adjacent faces (the 3 cycle), but can't scroll past it, as this is when my computer freezes. I like your alg better than mine, and am curious about how the other algs look.

Would you be able to cut and paste them into this thread?

Cubacca
 
r U l' U' r' U l U'
r U' l' U r' U' l U
l' U r U' l U r' U'
l' U' r U l U' r' U
similar but you only need to AUF instead of redundant set up moves involving the F face.
r U l U' r' U l' U' 3 cycling between 3 faces....reverse the alg to see what it does.
l' (u' d) l U2 l' (u d') l U2 cycling 3 blocks..this is the fastest angle for me.
d M2 d' L2 B2 u' M2 u M2 B2 R2 x2 for entire adjacent centres

I have already included the algs for the opposite algs somewhere here.
 
rachmaninovian said:
r U l' U' r' U l U'
r U' l' U r' U' l U
l' U r U' l U r' U'
l' U' r U l U' r' U
similar but you only need to AUF instead of redundant set up moves involving the F face.
r U l U' r' U l' U' 3 cycling between 3 faces....reverse the alg to see what it does.
l' (u' d) l U2 l' (u d') l U2 cycling 3 blocks..this is the fastest angle for me.
d M2 d' L2 B2 u' M2 u M2 B2 R2 x2 for entire adjacent centres

I have already included the algs for the opposite algs somewhere here.
Click to expand...

Thanks

The 3 cycles in particular are very slick in comparison to mine. Too bad there isn't an algorithm set (yet) that allows you to do all the swaps without needing to regrip/turn the whole cube.
 
lol after reviewing all my video solves, the max number of algs I use is 5 for the centres, and they are fast algs that can be executed very very fast.
Therefore I believe that freestyle centres would still be the best for the centres...for me :D of course the average would suffer a bit...so I gonna learn some more useful centres algs and use them freestyle :) just like freestyle 3x3 BLD edges...

anyway just got my first sub 50, 48.66 and it was non lucky, technically. 3 algs to finish centres. (my average alg count should be around 4.5 anyway)
 
rachmaninovian said:
lol after reviewing all my video solves, the max number of algs I use is 5 for the centres, and they are fast algs that can be executed very very fast.
Therefore I believe that freestyle centres would still be the best for the centres...for me :D of course the average would suffer a bit...so I gonna learn some more useful centres algs and use them freestyle :) just like freestyle 3x3 BLD edges...

anyway just got my first sub 50, 48.66 and it was non lucky, technically. 3 algs to finish centres. (my average alg count should be around 4.5 anyway)
Click to expand...

Nice work! As your times get low consistently, it will be interesting to see how much of a factor the total number of center solve algorithms required is.

I am working on the 21 algorithms for the 4 step centers. I am trying to find algorithms which keep cube rotations to a minimum. Once I find them all, I will post them. With any luck, some of them might be helpful to you.

In regards to the 4 step system, my initial thought was to just hammer through whatever configuration you wind up with once the cage is finished (i.e., stuff green/blue cubies in the blue center, then the rest of the blue/green cubies into the green center).

In the event of having all 4 blue center cubies avialable for sorting (all 4 in either the blue center or its adjacent centers-none in the green center), direct solving the blue center would ensure a 3 step maximum to solve the centers.
 
lol, just post whatever you find at the moment. =D
i'm more and more convinced that freestyle centres is the way to go - because if you are going to rigidly do it in say 2-3 algs the algs would be SUPER long and hard to execute fast, thus defeating its purpose...and most of my algs only takes 2s to execute, 3s at most...except some of the more stupid cases like the adjacent centre switch...
 
i do the first one as l' E l U2 l' E' l U2 or the mirror.
i know a 15/16 mover for the 2nd one but i keep forgetting it; mine is just 3 niklas' with cancellations so its more fingertricky then your alg =D
thanks for your help anyway :D
 
rachmaninovian said:
lol, just post whatever you find at the moment. =D
i'm more and more convinced that freestyle centres is the way to go - because if you are going to rigidly do it in say 2-3 algs the algs would be SUPER long and hard to execute fast, thus defeating its purpose...and most of my algs only takes 2s to execute, 3s at most...except some of the more stupid cases like the adjacent centre switch...
Click to expand...

Given your current rate of progress, you should keep doing what you're doing. You might as well push your system to its limits and see where it takes you.

The stuff I am working on is all too long (for now), as I am confined to finding algorithms manually. As you might imagine, lots of the algorithms are simply shorter algorithms welded together with set up moves. This is where a computer search would come in really handy.

So far, the only mildly interesting algorithms I found/made are as follows:

0 solved/sorted

2 adjacent cubies on one adjacent face, 2 adjacent on the other adjacent face

r2 U2 r (u2 d2) r' U2 r2 (u2 d2)

2 diagonal cubies on one adjacent face, 2 diagonal on the other adjacent face

u2 r2 U2 r (u2 d2) r' U2 r2 d2

2 adjacent solved/sorted

2 adjacent cubies on adjacent face

r' (u' d) (l' r) U2 (l r') (u d') (l' r) U2 l
(longer than the usual, I know, but no cube rotation required)

1 cubie on each adjacent face

u2 r' (u' d) (l' r) U2 (l r') (u d') (l' r) U2 l u2

not too much else to report, as the algorithms start getting ugly:

d2 F' r2 U2 r (u2 d2) r' U2 r2 (u2 d2) F d2
 
Just a few algorithms I found:

3 cycle centers (F centers to U, U centers to B centers to F)

r2 F2 D2 l E2 l' (E2 D2) F2 r2

or if you really want to just swap U and F centers:

r2 F2 D2 l E2 l' (E2 D2) F2 E2 r2 E2

In the 4 step centers method I am trying to develop, the 3 cycle algorithm would be acceptable, as the first two steps would be for sorting only.

Lets say that the orange and red faces are solved, and the cage is built, and we are solving/sorting the blue center in step one. The only thing that needs to be accomplished is to get 1 to 4 blue or green center cubies from the adjacent yellow or white centers, without disturbing the green center. For this method, it does not matter where the yellow or white center cubies which started in the blue center end up at the end of step 1. They will by default get sorted in the yellow and white faces. This means that the optimal (shortest) algorithms for steps 1 and 2 would be shorter than if we cared where exactly the yellow and white center cubies end up.

Man, I really would love to make or get my hands on an algorithm search program. sigh.
 
wells you see the problem with the 2/3/4 alg system: they have long and perhaps non-fingertrick-friendly algs xD
i'd just learn the most useful algs and stuff like that seriously..and niklas is very fast anyway? :D
 
rachmaninovian said:
wells you see the problem with the 2/3/4 alg system: they have long and perhaps non-fingertrick-friendly algs xD
i'd just learn the most useful algs and stuff like that seriously..and niklas is very fast anyway? :D
Click to expand...

In a word, yes.

Since I am searching manually, the chances of finding more fingertrick-friendly algorithms is fairly low. I'd love to be able to pick from ten to twenty algorithms for each case, which would only be possible with a search program. Even that might get tricky, when E and M moves come in to play. Since they each are turns of two layers, there are going to be cases where a search program will have to search very deeply for algorithms which might technically be longer in move count, but faster because of M and E moves.
 
wells I took an average of 20 for the number of algs needed...right now it's at 4.25 algs. not too bad for short and fast algs eh? i'd probably learn a few more useful ones and stuff like that..
i'd do an average of 400 to make sure...happy waiting...xD
 
You must log in or register to reply here.

Similar threads

abunickabhi
[Unofficial] 5x5 Cage Method UWR 1:28.18 by Hill Pong
Replies
18
Views
2,254
abunickabhi
abunickabhi
Boneless
Sandwich/Cage Method Algorithms
Replies
4
Views
2,130
Boneless
Boneless
E
6x6 cage method
Replies
4
Views
2,197
martinss
M
E
Cage Method for the 5x5x5
Replies
5
Views
2,411
EvanCahill
E
MennoniteCuber1
Cage method 4x4x4
Replies
2
Views
1,471
MennoniteCuber1
MennoniteCuber1
Back
Top