I just had a BLD epiphany

[Jason Baum]

Okay, first of all I'm sorry if this is common knowledge and I am just wasting space, but I have been blindfold solving for over two years now and didn't know about this until a few minutes ago. At the US Open in Chicago, my second BLD solve was off by two mis-oriented corners. I was upset about this because the time was 3:44.xx, which is pretty good for me, and I was positive that I didn't make any mistakes during my solve. I was so confident that I had done everything right and couldn't believe it when I took the blindfold off and saw those two corners. I tried to think of how I possibly could have messed up CO and I was positive that I didn't. Well I just figured out what I did wrong. I had a parity error on that solve, the last thing I did was fix it. Going by Macky's numbering system, the parity error was (1 9) (2 6), or (FU FD) (FUR FDR). So, to solve this at the US Open, I did x (F permutation). Well, that was my mistake right there. I realized just now that doing that PLL on the F face also twists two corners. This confused me a lot because I had a similar situation happen to me at WC2005. One of my BLD solves there ended with the parity error (1 9) (1 6), or (FU FD) (FUL FDR). I very stupidly wasted about two minutes coming up with a ridiculously long setup move to bring all the pieces to the U face and doing a T perm. I ended up successfully completing the solve, but I was mad that I wasted so much time fixing that parity. I was talking with Leyan Lo and Tyson Mao about it afterward, and they told me that I could have simply done an N perm on the F face and it would have fixed it. From then on I just assumed that it was okay to use PLLs to fix parity errors on any face. Well after some experimentation I figured out that it is okay to use PLLs that swap corners diagonally. It is NOT okay to use PLLs that swap corners straight across! This means that if you are fixing the parity error on any face except U or D, the ONLY PLLs you can use are the Y perm, V perm, or the N perm! Any other PLL will mess up corner orientation.

Again, I am sure this is known by a lot of people, and I am sorry if this is common knowledge. It just blew me away when I figured this out, because like I said, I have been blindfold solving for over two years now and have never known this. I'm also glad to have figured out what I did wrong on that BLD solve in Chicago. I knew I didn't mess up

 

[AlexandertheGreat]

wow, i never knew that! I've been doing BLD for a few months, and I've always been under the impression that i had to fix a parity on the U or D face. I'll definitely keep that in mind. So i guess this concept can also be applied to CP, not just parity? For example (1 6)(2 5) for CP, I guess you can just do x(H-permutation)U2x'

This should help me improve my BLD times, thnx for posting

 

[pjk]

Hmm... interesting. I don't understand why that is though, since the PLLs aren't affecting the orientation. The only thing I could think of is if your setups are only 90 degree turns, a swap straight across would mess the orientation up, but if a diagonal swap was done, it wouldn't. but what if your setup had a 180 degree turn, would it still mess it up?

 

[Jack]

A parity fix wouldn't work on the front face because PLLs preserve orientation, but only on the face they are solved on. If you have all pieces on the front face, then you do a PLL on it, the PLL will preserve orientation on the front face only. So if you do an F perm that moves FRD to FUR, this is the same as doing a front turn that will put it FRD in FUR except of course it doesn't affect the other pieces. And if you do a front turn, the corner orientation will not be preserved.

 

[pjk]

Ah, yeah, my bad, that is obvious. Thanks Jack.

 

[cuBerBruce]

People who solve orientation first and recognize corner orientation patterns on different faces besides U/D are well aware of how orientation patterns change when you try to recognize them on the side faces.

Also, beware that V-perm and Y-Perm do not necessarily preserve edge orientation when applied to side faces. This is because they effectively move two edges 90 degrees in the face that the algorithm is applied with respect to. So if you do a V-Perm or Y-Perm on a face that you are not allowed to make a quarter-turn on, edge orientation will not be preserved for the two moved edges. So V-Perm and Y-Perm only work on two of the side faces, not all four. In summary, beware when trying to use PLL algorithms on side faces.

 

[pjk]

This is something that I been working on:
If I have a parity and I know it directly after solving the corners, I will go ahead and fix it right then and there, and then go on to EP. Have you guys ever tried it? It sometimes works well, other times it doesn't.

 

[Mike Hughey]

This is something that I been working on:
If I have a parity and I know it directly after solving the corners, I will go ahead and fix it right then and there, and then go on to EP. Have you guys ever tried it? It sometimes works well, other times it doesn't.

I will do this if it's REALLY obvious. Otherwise I'll wait (probably 95% of the time I won't do it). But I've also been trying to improve on my recognition of this - I'm hoping I can get where more of them are really obvious.