Welcome to the Speedsolving.com, home of the web's largest puzzle community! You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

Old Pochmann is funny for move count. It's surprising how fast one can solve with Old Pochmann, considering how many moves there are in a typical solve.

cycle edges:
A -> V -> D : y [D R2 D', M2] y'
A -> B -> T : U' L E2 L' U' L E2 L' U2
A -> I -> F : [M, U L' U']
A -> G -> M : S R2 (U perm: L U' L U L U L U' L' U' L2) R2 S' Obviously not optimal, but easy to execute. Optimal would be: x' [E2, R U' R'] x
A -> R -> W : L [M' D2 M, U] L' This is not optimal. Optimal would be: D L' D M2 D' L D M2 D2
A -> U -> H : y L2 U2 L E2 L' U2 L E2 L y' This is not optimal. Optimal would be: y' x [E2, R U2 R'] x' y

cycle corners:
A -> C -> B : A perm (R2 B2 R F R' B2 R F' R)
A -> W -> S : R2 U2 R' D2 R U2 R' D2 R'
A -> G -> D : [L D L', U']

@Mike: Now that I'm reconstructing solves, I see that I don't use BH edges on 3x3x3 every cycle. There do seem to be some cases where I just do what feels fast at the time, even though I know that it's sub-optimal. For 3x3x3 I'm starting to see that this is probably the better choice. Of course I try to use mostly optimal stuff on 5x5x5 and 7x7x7, but after looking at my solve reconstructions I definitely don't on some cycles. The Speffz central edge cycle A -> U -> H I would definitely do sub-optimally as a 9 mover even on the larger cubes. I feel the 9 mover is more "comfortable" with fewer cube rotations I guess. Again with the foot in the mouth for me.

@Ralph: No worries on the movecount for Old Pochmann. Trust me that most of us are already convinced that it is a fast BLD method. I would say not to count your overall movecount, but maybe your setup moves movecount, just to make sure you're always using the most efficient setup moves? We all know the premise behind the method, so that repeated use of the T perm or whatever PLL you're using (R, etc.) shouldn't drag down your results here. Try a movecount without counting any of the T perms, then tell us how many cycles you did. That would give us an idea of your setup turns per cycle, which could be useful information to know for Old Pochmann.

--edit--

Ok Aron, Mike, Ralph, those reconstructing solves here's one more. This solve was REALLY fast for me, like holy crap!

Scramble: L U2 D B' U' F2 B D R' F L' F U' B' L' D' U' L' F' L2 U2 L' R2 B F'
Time: 1:11.92
Total turns: 90

flip edges: L' y' E' R E' R E' R2 E R E R E R2 y L

cycle edges:
D -> U -> R : y' R2 U' R' E R U R' E' R' y optimal is [L, U M2 U']
D -> C -> T : [L E2 L', U]
D -> L -> J : x [U2, R' E2 R] x'
D -> B -> S : y [M' D M, U2] y'
D -> O -> M : [R2, U' M' U]

cycle corners:
A -> C -> G : [U2, R' D R]
A -> B -> I : y' x' [D', L' U' L] x y
A -> W -> S : R2 U2 R' D2 R U2 R' D2 R'
A -> K -> N : R [L' D L, U2] R' Optimal would have been: y [U L' U', R'] y'

Got 40.76 on this scramble.
Moves: 92
Was good scramble with easy cycles. But it looks like I'm behind you guys in move count

Spoiler

R' U2 R' D' R U2 R' D R2
x R U R' D2 R U' R' D2 x'
y2 D' R2 U R2 U' R2 D R2 U R2 U' R2 y2

U R' U' M U R U' M'
L2 R U' R U R U R U' R' U' R2 L2 (though I actually wanna switch to l2 U M U2 M' U l2 for this)
M' U L U' M U L' U'
y' x' R2 U R U R U' R' U' R' U' R' x y
U R U R' U' R' U' R' U R
y' L2 R U' R U R U R U' R' U' R2 L2 y (again, planning to switch to y r2 U M' U2 M U r2 y')

Chris, I'm doing OK as far as set up moves go, never more than 2 moves for edges as I am using four targets.

As described on Joel's page:
Name Shoots to Algorithm
T1 [shoot to UL] R U R' U' R' F R2 U' R' U' R U R' F'
T2 [shoot to LU] x' R2 U' R' U x R' F' U' F R U R' U'
J1 [shoot to UF] R U R' F' R U R' U' R' F R2 U' R' U'
J2 [shoot to UB] R' U2 R U R' U2 L U' R U L'

I really like T2 since shooting to LU with the regular T-perm can be confusing for undoing set up moves.

For corners I use two targets: LFD and DFR but I find that I have a strong preference for LFD.
The actual DFR alg feels nice but since it starts with a cube rotation I find that doesn't flow as nice into set-up moves.

I use UBL for corners. I use DF for edges; it transitioned nicely from M2. A number of cases for M2 are valid BH algorithms, and many other valid BH algorithms can be constructed by optimizing M2 a little.

Here's my list of corner orientation algorithms. There will be a wordfile online in a couple of hours, I'll edit this post for url.

Spoiler

Dia = R U R’ U R U2 R U2 R’ U’ R U’ R’ U2 R’ U2
Dia’ = U2 R U2 R U R’ U R U2 R’ U2 R’ U’ R U’ R’
Bowtie = R U' L' U R' U' R U R' U' L U R U' R'
H-Sune: R U R’ U R U2 R’
V-Sune: L’ U’ L U’ L’ U2 L
H-Anti-Sune: R U2 R’ U’ R U’ R’
V-Anti-Sune: L’ U2 L U L’ U L

The sticker indicated should go to the U/D-layer, because that’s the way I recognize them.

I use UBL for corners. I use DF for edges; it transitioned nicely from M2. A number of cases for M2 are valid BH algorithms, and many other valid BH algorithms can be constructed by optimizing M2 a little.

(Is that optimal? I've never tried to be sure it's optimal, but it's pretty fast and easy for me and I can see it easily.)

There are certainly a few cases that are awkward, but it seemed to me that was true no matter what you pick for your buffer. Different buffers give different awkward cases, but there are always some that are awkward.

Some of my least favorite cases are actually U perms with all of the pieces on the bottom, such as DF-DR-DB. I do the 2-gen U perm instead of BH (like Chris), but it's kind of a pain reorienting the cube for those - I always feel like it takes longer than it should because of the reorient.

Roman rooms is when you walk into a room and place objects at locations you make, such as a bed, drawer, light switch etc, and have something like 10 locations in the room, then when you solve you "walk" into the room and just remember what's in each location. I like this style much then typical journey.

Roman rooms is when you walk into a room and place objects at locations you make, such as a bed, drawer, light switch etc, and have something like 10 locations in the room, then when you solve you "walk" into the room and just remember what's in each location. I like this style much then typical journey.

Thats pretty interesting,but how actually do you mix journey and roman rooms in multi BLD? Do you do Roman rooms for edges and journey for corners,something like that?

Some of my least favorite cases are actually U perms with all of the pieces on the bottom, such as DF-DR-DB. I do the 2-gen U perm instead of BH (like Chris), but it's kind of a pain reorienting the cube for those - I always feel like it takes longer than it should because of the reorient.

Actually, as I was typing this, it occurred to me that the logical thing was probably to do that. And that's actually how I do it on a 5x5x5 or a 7x7x7, although I generally have to think about it too much (I think of the L2 R2 U' as setup moves, and usually don't do the cancel).

I break down the D-face U perms into two types: one that have DB and ones that don't. If they don't, the reorientation is just z2 which is easy. If they do, I'd do something like R2 u M u2 M u R2. The worst U perms for me are the ones like DF->RF->UF.