Learning Petrus - Having Trouble

[fleeting_penguin]

What is the maximum number of moves at each step, being color neutral?

 

[Johannes91]

Step 1: 8 moves (6)
I've found positions where each corner requires 8 moves, but chances to get one after random scrambling are purely theoretical. The chance that all corners require at least 7 moves is very low as well, less than 1/10000. So in practice, you can trust that 6 moves is enough.

Step 2: 9 moves (7)
It's possible that all 3 options require 9 moves, the probability is 1/21946982400. And 8 moves is rare, too, I believe it's less than 1/100000 (but I haven't checked yet). So again, in reality, you won't need more than 7 moves.

Step 3: 8 moves
Color neutrality doesn't matter here, worst case is 8 moves. With practise, though, you'll be able to look ahead during the pevious step and you often get a cancellation.

Step 4a: 10 moves (8)
Probability for 10 moves is 1/146966400 and 9 is rare, too.

Step 4b: 9 moves
Color neutrality doesn't help here, either; worst case is 9 moves.

LL
Move count is not the most important thing here in speedcubing, and there are many different ways to solve LL. But using optimal ZBLL, the worst case is 16.

 

[AvGalen]

So 6+7+7+9+9 = 36 moves in total or are there steps after 4b?

36 would be pretty good for Fewest Moves

 

[Johannes91]

Sorry, that 7 for Step 3 was incorrect. I edited it to 8. There's LL after 4b, but remember that those are worst cases and averages are (naturally) lower.

 

[AvGalen]

I have actually became pretty interested in Petrus. I think I will look into it after the Polish Open. I have tried Fridrich, but it will take a lot of time before I can beat my keyhole times with it.

Keyhole: I don't think I will get much faster with keyhole. My F2L is about 14 seconds and that is not going down anymore.
CFOP: I need to do a couple of thousand solves before I will be as fast as with keyhole. It might be the most logical step for me, but if I am going to switch I might as well look into the alternatives
Petrus: Sounds interesting and can also be used for fewest moves
Roux: Might be fast for others, but probably not for me because I am but with slice moves. I will at least learn the method though.
Corners First: Doesn't appeal to me for some reason.

I think I will change my 3x3x3 goals to "be able to do sub-30 on average with keyhole, cfop, Petrus and Roux by the end of 2007". Speed doesn't really interest me, I just like to learn.

 

[deadalnix]

Step 3: 8 moves
Color neutrality doesn't matter here, worst case is 8 moves. With practise, though, you'll be able to look ahead during the pevious step and you often get a cancellation.

You say that worst case is 8 moves. I suppose that this case is flipping (I can say flipping ?) all 8 edges ? How do you do this in 8 moves ?

 

[Johannes91]

You say that worst case is 8 moves. I suppose that this case is flipping (I can say flipping ?) all 8 edges ?

Yes you can.

We have only 7 edges here, not 8. The hardest case is when 6 are bad and the only good one is at UF, where U and F are the free layers. One solution is U R L' U L R2 F' R.

 

[deadalnix]

Thank, realy much better than a previously do.