Difficult Crosses

[blgentry]

I've been doing some cross training recently, just trying to get my move count to 7 or maybe 8 for each one, and I've got to so say this is extremely difficult for many crosses. I always solve the cross on white, which I know gives a small disadvantage.

So today during cross practice, I decided to not be lazy, and really try to plan out the cross for 7 moves or under. After only 2 or 3, I landed (with 25 random cube turns) on a situation that I can *not* do in under 8 moves. In theory there are only about 100 of these situations out of 190,000, and I'm having a hard time believing that I have randomly found one so quickly.

So, presented for your solving, hard cross #1. Scramble with white on top, green front, and please solve cross on white:

B D2 F2 R2 D2 B' D2 R2 D2 F' U2 R' U' L U2 L2 B L U' F' R'

I've actually been thinking that I should record crosses that I have particular difficulty with and see how someone with more experience would solve them. Perhaps I should start putting more hard crosses in this thread.

Any input on the above situation is appreciated.

Brian.

 

[Inusagi]

X2 L F' B R' D' R2 L2 D2 = 8 moves

I can't find a another way, but if you are so crazy about under 8 moves (which is impossible with out midlle slices), then there's a oppertunity here: Between F' and B, and R2 and L2, which is 6 moves.

 

[Johannes91]

There are several 7-move solutions:

U L' U L2 F R U
U' L U F' L' F2 R
U' L U' L2 B' R' U2
F' D' L B' R' U L2
F' L U R' F' R L2
F' L U R' L2 F' R
F2 U' L U R F L'
F2 R U' L U F L'

I've actually been thinking that I should record crosses that I have particular difficulty with and see how someone with more experience would solve them.

If you're just interested in optimal solutions, why not use a computer? I wrote one solver a while ago, it's here. Paste the scramble to the input field, click [Send], and then either [Show depths] or [Show solutions], depending on what you want. It works on most proper browsers (like Firefox).

 

[blgentry]

^ I was wondering if you used a computer, or if you were just a machine yourself.

Thanks for the solutions and the links.

As to whether I'm "only interested in optimal solutions", certainly not. I want to solve the cross as fast as possible. I'm using the number of turns as a metric, as, in about 20% of crosses, I take 9 to 11 moves, which is obviously too many. Again, speed is the real important factor to me.

I guess I can use your solver to prove to myself there are ways with 8 moves or less, and perhaps to give me ideas on how to do combinations of cross pieces. Combos seem to be the key to a short cross solution. I hope they are the key to a *fast* cross as well.

BTW, I'm still interested in other people's responses, even your response is that you would have taken 8 moves to solve it. That's important to me, because there are world class cubers here and if they are using 8 moves for some crosses that could be 7 with a computer, then I need not waste my time going for 7 or under.

Thanks,

Brian.

 

[brunson]

I could only find an 8 move solution on white, but I still suck quite a bit at the harder crosses.

However, I did find a 5 move cross on Red: B2 U F2 B' R and a 5 move cross on yellow: R2 U F B' R

I think it was Lucas in another post about crosses pointed out that optimal cross solutions my not be "finger friendly", but until someone comes up with a quantifiable metric for "finger friendliness" FTM is a pretty easy one to comprehend.

It works on most proper browsers (like Firefox).

As a developer of web applications, I feel the need to high five Johannes for that statement.

 

[FU]

x2 L F' R' B D' R L2 D2

L F' R' - Insert blue and green pieces
B - Match orange piece with red center so it is in the correct relative position to the red piece
D - Adjust bottom to insert red and orange pieces
R L2 - Insert red and orange pieces
D2 - Align the whole cross

 

[Johannes91]

I think it was Lucas in another post about crosses pointed out that optimal cross solutions my not be "finger friendly", but until someone comes up with a quantifiable metric for "finger friendliness" FTM is a pretty easy one to comprehend.

And another point I just remembered is that some top speedcubers have said that you shouldn't be turning full speed at cross. So using optimal solution instead of 1-2 longer but easier ones could be a good idea, because looking ahead is easier.

FYI, there are 481 different solutions with FTM less than equal to 8, and 14451 with FTM less than or equal to 9. There are also 14 solutions with FTM = QTM = 8:

[size=-2]U R' F L F U R F
U' F' R' F' L U R F
U' R' F' D' L B' U F
F U' F L U F R L'
F' U' F' L U R F L'
F' L U R' L F' R L
L D' F' B' R' B' U L'
L D' B' R' F' B' U L'
L F' U L' F' R' F' R
L F' D' B' R' B' U L'
L' U L F R' L U F'
L' D' B U' L U' B L
L' B' U R' L F L U
L' B' U L F R' L U[/size]

 

[Lucas Garron]

B D2 F2 R2 D2 B' D2 R2 D2 F' U2 R' U' L U2 L2 B L U' F' R'
x' L Us' l' U' M2' u2'

Slice turns and double-layers can be okay.