WANTED: New Dedge Flip Algorithm!

[Christopher Mowla]

Kenneth,

What particular case are you saying that can be theoretically done in 14f? Is it pure? If so, is it doubleparity, only OLL parity, or could be either?

reThinker,
Well, you can take out the "x2" rotation and have turns about D, (just like your "DoubleTrouble" has. But if you don't like it,well...

 

[Kenneth]

It is only OLL parity, the double looks much the same but is diffrent, like if you add a H-PLL on the 3x3x3 corner case.

It has got more solutions in 14 turns also:

Searching depth 14

U R2 B2 D' L F2 L' D B2 R2 U R' U R (14f*)
U F U F' U F2 L2 D B' R2 B D' L2 F2 (14f*)
U' F' U' F U' F2 R2 D' B L2 B' D R2 F2 (14f*)
U' L2 B2 D R' F2 R D' B2 L2 U' L U' L' (14f*)
R U R' U R2 F2 D L' B2 L D' F2 R2 U (14f*)
R U2 R' U2 F2 L' F' L F' D2 B R' B' D2 (14f*)
R U2 R' U2 D2 B R B' D2 F L' F L F2 (14f*)
F L F' D2 B R' B R B2 U2 L U2 L' D2 (14f*)
F' R' F D2 B' L B' L' B2 U2 R' U2 R D2 (14f*)
D2 R U2 R' U2 F2 L' F' L F' D2 B R' B' (14f*)
D2 F L F' D2 B R' B R B2 U2 L U2 L' (14f*)
D2 F' R' F D2 B' L B' L' B2 U2 R' U2 R (14f*)
D2 L' U2 L U2 F2 R F R' F D2 B' L B (14f*)
L' U2 L U2 F2 R F R' F D2 B' L B D2 (14f*)
L' U2 L U2 D2 B' L' B D2 F' R F' R' F2 (14f*)
L' U' L U' L2 F2 D' R B2 R' D F2 L2 U' (14f*)
B2 R2 D F' L2 F D' R2 B2 U' B U' B' U' (14f*)
B2 R' B' R B' D2 F L' F' U2 D2 L U2 L' (14f*)
B2 L B L' B D2 F' R F U2 D2 R' U2 R (14f*)
B2 L2 D' F R2 F' D L2 B2 U B' U B U (14f*)
Searching depth 15

 

[Stefan]

I have completely lost the overview. Could you guys create a page in the wiki listing the lengths of the shortest known algorithms along with one example algorithm each? I see three dimensions:

- Which parity (orientation, permutation, both)
- Which freedom (pure, LL, LL-oneF2Lslot, justKeepCentersAndDedgesIntact)
- Which metric (WCAqtm, WCAhtm, SiGNqtm, SiGNhtm)

These are just my suggestions, particularly the metric names, although I do believe they make it both clear and easy to count/check.

 

[Christopher Mowla]

I am sorry to ask again, but I am assuming you mean a 14f pure OLL alg, right?

Thanks for the 3X3X3 algs.

 

[Kenneth]

Na, that are the same cases as the big cube paritys, the cases we setup when we look for algs for them using computers but restricting to turns that don't destroy edges on the big ones like the 14 turn algs does. The shortest in that group are 15 turns.

And for 3x3x3 that are not OLL's but ZBLL algs... or it is what I use them for on that cube (but the 15 turn versions I also use for big cubes).

Edit: look at the corners as the where wing edges on a 5x5x5, then it is pretty clear it's the same.

 

[reThinking the Cube]

I have completely lost the overview. Could you guys create a page in the wiki listing the lengths of the shortest known algorithms along with one example algorithm each? I see three dimensions:

- Which parity (orientation, permutation, both)
- Which freedom (pure, LL, LL-oneF2Lslot, justKeepCentersAndDedgesIntact)
- Which metric (WCAqtm, WCAhtm, SiGNqtm, SiGNhtm)

These are just my suggestions, particularly the metric names, although I do believe they make it both clear and easy to count/check.

Good suggestions. If I had a little more time, the wiki listing would already be there. It's a to-do that needs to get done. I would like to add though, that my main interest in parity algs, is more practical , and not overly obsessed with the arbitrary metric lengths of the algs. I am much more concerned about how easy they are to learn, and how fast they can be executed. So adding the dimensions (merits) of symmetry, and speed would be useful. The freedom dimensions you gave, should allow for the categorization of algs for use with different solving methods.

Stefan: maybe you (or someone else so qualified) could get it started - setting up the proper format so that the different parity algs can be categorized and compared, and then the rest will follow.

 

[reThinking the Cube]

I took DoubleTrouble™ - r U2 r' U2 r' D2 r D2 x' r' U2 r U2 r' U2 2s2

and reworked it (backwards+rotated) to see if I could make it easier to speed turn. I also wanted to tailor it to my solve method, which prefers the flipped dedge to start out at FR rather than FU.

The result: DoubleXTrouble™ R2' u' R2 u R2' u' y R2 u R2' u' y2 R2 u' R2' u m2

The whole sequence flows really well. Very fast and easy to learn. Though Eastsheen™ cubes are not optimal - sometimes locking/catching up on (u,u') triggers. For cubes like that, the original alg (DoubleTrouble™) works better.

Suggestion: start out grip with left thumb spanning all 6 pieces lower left (DFL-rd), and LH middle/ring holding the 6 pieces on DB, so cube is pinched in the left hand without interfering with (Uu) turns. Right thumb loosely on (uRF-dRF). Turn R2', and use RHmiddle finger at an angle to do the "u'", then R2, followed by using the RH index trigger for the "u", then R2' and RHmiddle "u'". So that is (R2' u' R2 u R2' u'). Do the whole alg like that - with flip-flop triggering. I settled on the [y2] rotation as the best compromise for the last part, but it is ok since the RH thumb can get into position for the last sequence (R2 u' R2' u) and then the "m2" can be done without adding another cube rotation. Maybe an even better way to finish this without the regrip [y2] is: R2' u' R2 u R2' u' y R2 u R2' u' 3r2 d' 3r2' u m2
For this variation the last "u" is done by pushing with the LHthumb. Smokin'.

Notation is SiGN.

 

[reThinking the Cube]

I was playing around with the starting sequence (r U2 l' U2) which turns even faster for me than (r U2 r' U2). Also (r' B2 l B2) = x' (r' U2 l U2) another fast turner. After some thinkin' , and keeping with the 8-cycle of edges idea, I came up with this:

reParity™ - r U2 l' U2 x' r' U2 l U2 r' U2 l U2 l' U2 y' 2m2 y'

It's gonna be real hard to improve on this one.

EDIT: reParity™ - "(re)THINKER's (re)VENGE Parity Algorithm."

 

[deadalnix]

These trademark signs become ridiculus xD

Can I see regstration paper or are you just kiddies playing grown up persons ?

 

[masterofthebass]

If you look, its only the two people who make this thread pointless in the first place. Normal people aren't that stupid.

 

[blade740]

I'd like to point out that Lars Petrus has done that for years:
http://lar5.com/cube/fas6.html

 

[deadalnix]

It's not because someone do something the wrong way that everybody has to do the same.

And the trademark is just the « extra » sign that make the speach perfect.

 

[qqwref]

I'd like to point out that quite a few of Lars's names have become standard terminology

 

[reThinking the Cube]

If you look, its only the two people who make this thread pointless in the first place. Normal people aren't that stupid.

How is this thread pointless?

The new (non-pure)parity algs that I have found, and then shared with this forum, IMO - are both easier to learn, and faster to execute than all others. I think it is quite amusing, that the so-called experts, have not picked up on this yet. FYI - I am not cmowla, and neither are you.

 

[masterofthebass]

If you haven't noticed, there's plenty of "experts" who have posted in this thread, and they all agree the algs you think are good are crap (myself included. I do have the 4x4 single WR). Nothing in this entire thread has been useful for speedsolving and really only useful for some fewest moves stuff, which for 4x4 is pretty dumb to do anyway since you wouldn't do a solve with OLL parity anyway.

 

[reThinking the Cube]

I'd like to point out that quite a few of Lars's names have become standard terminology

That's right. I mentioned earlier in this thread, the naming™ by Petrus, and I also suggested giving names like that to any new algs found. Names for algs are awesome! Somebody should start a thread to get some commonly accepted names for more of them.

 

[DavidWoner]

which for 4x4 is pretty dumb to do anyway since you wouldn't do a solve with OLL parity anyway.

And if you did, you would want HTM optimal non-pure algs, not QTM optimal pure ones.

 

[reThinking the Cube]

If you haven't noticed, there's plenty of "experts" who have posted in this thread, and they all agree the algs you think are good are crap (myself included. I do have the 4x4 single WR). Nothing in this entire thread has been useful for speedsolving and really only useful for some fewest moves stuff, which for 4x4 is pretty dumb to do anyway since you wouldn't do a solve with OLL parity anyway.

reParity™ - r U2 l' U2 x' r' U2 l U2 r' U2 l U2 l' U2 y' 2m2 y'

I don't care about fewest moves. I think you have been getting me confused with someone else. Actually do this alg, (SiGN) and then tell me your opinion.
I want to give you a chance to redeem yourself.

 

[deadalnix]

That's perfectly unusable compared to :

r'U2rU2'r'F2r2U2'rU2'r'U2xU2r2U2' or something similar.

 

[qqwref]

I think it is quite amusing, that the so-called experts, have not picked up on this yet. FYI - I am not cmowla, and neither are you.

I think it is quite amusing that you feel you are an expert when people who are far faster/smarter than you are not

That said, a few of the algs from this thread seem to be decent enough to be usable for speedsolves, or at least interesting for fewest moves purposes. So, reThinking and cmowla, don't think you are suddenly gods of the cubing world, but I would like to thank you guys for the new algs.