Stefan
Legend
Was hoping you'd actually answer the question so we could give better answers. But it's ok if you don't want that.
AvGalen
Premium Member
The topic starter already gave 2 answers:
and
His curiosity is probably satisfied by now and maybe he already learned an 11 move Blind-T-Perm, or the usual 14 move Speed-PLL-T-Perm, or maybe even a FMC 10 move-PLL-T-Perm (cancel the AUF)
cuBerBruce
Veteran
Just to be clear, there is no 10-move (face-turn metric) T-Perm that changes the parity. An AUF is needed to make a 10-move T-Perm into the conventional T-Perm case that swaps two corners and swaps two edges (and thus, changes parity). So Stefan is correct in saying that an 11-move T-Perm is needed (for the context of the question that was asked).
AvGalen
Premium Member
I completely agree. That is why I added the context (optimal blind, speed, fmc) for those T-Perms (11, 14, 10).
The T-Perm and the T-Perm with U2 need 11 moves. (These are the ones that change the parity)
The T-Perm with U and the T-Perm with U' can be done in 10 moves. (These are the ones that don't change the parity, or rather: These are the ones that change the parity with the T-Perm, but then it gets changed again with the quarterturn U/U')
Last edited:
cmhardw
Premium Member
Ok here's my average. To be perfectly honest tonight was my roommate's birthday and we went out drinking tonight. I'm still buzzing a bit when I took this average. I found a couple things when doing this. This was harder than I thought it would be, and I think maybe sub-4 is a bit faster than I thought at first. I can see getting sub-5 with practice, but sub-4 seems like it might take a lot of training.
07.03, 05.20, 06.05, (08.30), 06.62, 05.97, 07.78, 05.80, 05.22, 04.76, 05.17, (04.72) = 5.96
Some things to point out:
1) These times represent how long it took me to determine whether or not there was parity.
2) I made no attempt to memorize the cycles, I simply traced the permutations of the pieces visually and determined if it would take an odd or even number of transpositions to solve.
3) I think I can do faster when at my peak, but I think sub-4 would take ridiculous practice to achieve. Sub-5 should be easy with a couple weeks practice, but sub-4 might be fairly difficult to achieve.
--edit--
After looking back in this thread I feel I should mention that I did not pre-orient the cube before tracing the cycles. This is because a quarter turn cube rotation is an even permutation if viewed as a permutation on corners. I did start with the cube in a random position each time, but traced it as if it was already in my solved orientation.
07.03, 05.20, 06.05, (08.30), 06.62, 05.97, 07.78, 05.80, 05.22, 04.76, 05.17, (04.72) = 5.96
Some things to point out:
1) These times represent how long it took me to determine whether or not there was parity.
2) I made no attempt to memorize the cycles, I simply traced the permutations of the pieces visually and determined if it would take an odd or even number of transpositions to solve.
3) I think I can do faster when at my peak, but I think sub-4 would take ridiculous practice to achieve. Sub-5 should be easy with a couple weeks practice, but sub-4 might be fairly difficult to achieve.
--edit--
After looking back in this thread I feel I should mention that I did not pre-orient the cube before tracing the cycles. This is because a quarter turn cube rotation is an even permutation if viewed as a permutation on corners. I did start with the cube in a random position each time, but traced it as if it was already in my solved orientation.
Last edited:
- Joined
- Jun 7, 2007
- Messages
- 11,923
- Location
- Indianapolis
- WCA ID
- 2007HUGH01
- YouTube
- Visit Channel
Yeah. That's more fair. What I did wasn't really fair to the idea, because I did waste some time preorienting the cube. My times would probably have been several seconds faster if I hadn't done that.
Anyway, those are some pretty nice times, Chris. It would take me a lot of practice to get as good as you at it.
mrCage
Veteran
- Joined
- Jun 17, 2006
- Messages
- 655
Hi 
I dont even trace parity for fewest moves. There i could analyse parity by looking at the scrambling algorithm (number of quarterturns).
And indeed, why trace the parity, it comes out from under the carpet eventually ;-)
Per
I dont even trace parity for fewest moves. There i could analyse parity by looking at the scrambling algorithm (number of quarterturns).
And indeed, why trace the parity, it comes out from under the carpet eventually ;-)
Per
Stefan
Legend
Ha. Just learned something new, thanks. I knew the cornerParity=edgeParity mentioned earlier in this thread, but I wasn't aware you could ignore the centers and just look for cornerParity like you did there. That doesn't work for edges, though, but of course you wouldn't want to check those anyway. Still... at some point you have to care about the centers so this is a neat trick for parity detection but not for overall solving.
E
Eaudevian
Guest
ho no !
WCA rule for fewest moves :
"- The solution of the competitor must not be in any way related to the scrambling algorithm. Penalty: disqualification of the solve."
AvGalen
Premium Member
Please define related. Sometimes my solution starts/ends with the same move(s) as the scramble by accident.
E
Eaudevian
Guest
you are right, i don't know what does mean "related" in that context ...
(http://www.worldcubeassociation.org/regulations/#fewestmovessolving)
anyway, analysing parity by looking at the scrambling algorithm doesn't sound to me like an "accident" ! :rolleyes:
Stefan
Legend
Wait... you weren't kidding about that rules violation?
AvGalen
Premium Member
Apparently he wasn't
First: Why would that be cheating
Second: How could anyone prove I "cheated"
Third: For such a weird topic, some very interesting things were mentioned. Half of this topic could be moved to the "cube theory" subforum
cmhardw
Premium Member
Mike, to be fair remember that I used to practice this on 4x4x4. A few years ago I was convinced that counting cycles during inspection was a viable way to avoid OLL parity when solving the 4x4x4 for speed. I used to practice trying to count the parity of the wings within the 15 seconds. I never made it of course, but I feel that was useful practice for trying to count the corner parity using essentially the same method.
And if anyone is wondering I don't think counting cycles is a viable way to avoid parity on 4x4, unless you truly dedicated years to learning how to count cycles quickly.
Chris
fanwuq
Legend
I got 6.53 for a best RA of 12 in a session of 17 that was 7.21. Best single was 5.24. Not bad for over a month without practice. I wouldn't be surprised if Ville, Rowe, or anyone else who can already memo sub-20 do this in sub-4.
I think this method would be usable. The tough part for me would be memorizing from a turned position.
I think this method would be usable. The tough part for me would be memorizing from a turned position.
Average: 4.79
5.71, 4.75, (3.91), (6.20), 5.02, 4.29, 5.65, 4.29, 4.56, 4.20, 5.19, 4.21
I haven't done a single BLD solve since my last (unofficial) competition, which was more than a month ago. (Unless you count the 2/3 multiBLD that I did last week for a friend, which took something like 20 minutes
) Conclusion: Ville/Rowe can sub-3 with a bit of practice.
5.71, 4.75, (3.91), (6.20), 5.02, 4.29, 5.65, 4.29, 4.56, 4.20, 5.19, 4.21
I haven't done a single BLD solve since my last (unofficial) competition, which was more than a month ago. (Unless you count the 2/3 multiBLD that I did last week for a friend, which took something like 20 minutes
) Conclusion: Ville/Rowe can sub-3 with a bit of practice.
Last edited:
E
Eaudevian
Guest
what do you think about this for solving parity problems :
for memorization :
1. memorize the corners normally (and thus, detecting parity)
2. in case of parity, mentally turn the U face
3. memorize the edges (according to the new position engendered by the U-turn)
for solving :
1. solve all corners (except 2 of them in case of parity)
2. turn the U face so that 1 corner is solved and 3 remaining corners can be solved with a 3-cycle
3. solve the edges
(unhappy this is a simple case, where a U-turn can solve a corner while leading to a 3-cycle position . )
for memorization :
1. memorize the corners normally (and thus, detecting parity)
2. in case of parity, mentally turn the U face
3. memorize the edges (according to the new position engendered by the U-turn)
for solving :
1. solve all corners (except 2 of them in case of parity)
2. turn the U face so that 1 corner is solved and 3 remaining corners can be solved with a 3-cycle
3. solve the edges
(unhappy this is a simple case, where a U-turn can solve a corner while leading to a 3-cycle position .
)
fanwuq
Legend
what do you think about this for solving parity problems :
for memorization :
1. memorize the corners normally (and thus, detecting parity)
2. in case of parity, mentally turn the U face
3. memorize the edges (according to the new position engendered by the U-turn)
for solving :
1. solve all corners (except 2 of them in case of parity)
2. turn the U face so that 1 corner is solved and 3 remaining corners can be solved with a 3-cycle
3. solve the edges
(unhappy this is a simple case, where a U-turn can solve a corner while leading to a 3-cycle position .
No.
Memo:
1. count corner cycles
2. Mentally picture U turn
3. memo edges
4. memo corners
Solve:
1. U turn
2. solve corners
3. Solve edges
The way I currently do is
Memo:
1. memo edges
2. memo corners
Solve:
1. solve all even corners
2. solve all edges
3. parity fix alg
4. solve last corner if it is odd.
I'm now liking this count cycles method, I'm going to try it out. Solving last corner + parity alg probably takes me longer than 6 seconds, so I should try this AUF method instead.
However, odd parity only happens 1/2 the time and this is a waste on good parity cases. Only singles count for BLD, so perhaps this is worth it. If they start counting averages, I would consider this.
Last edited:
