PLL help please :)

[*LukeMayn*]

OK my time attack consists of 266 moves (excluding x/y/z) my best Time Attack is 1:19.63 (79.63) which is 3.34tps.... I know this is slow and is what is letting my solves down a bit. I have decided to improve my times (obviously) and get all sub 3.5, then 3, then, 2.5, then 2. This seems sensible and I think it should give my times a bit of a kick in the A$$ and make them about 1-2 secs faster.
Now this is where you come in, I want good people (or people who are quick on PLL) to help me with improving them fast. I want to know what to focus on more, flow, quick execution or recognition.
Flow: the consistancy of the alg, contiuous moves not trigger, gap, trigger, gap.
Execution: The time it takes to do quick triggers with small gaps.
Recognition: Time to recognise and AUF it. (This takes about 1-2 secs)
Thanks for the help I know know to practise too!!!

 

[brunson]

I'm not particularly fast in execution or recognition, but to help with the recognition part I wrote a program with a friend to drill me on OLLs and PLLs.

The program randomly selects a scramble created by combining an OLL an AUF and a PLL (optimized using optiqtm so you don't recognize the algs while scrambling). I solve the OLL (one look, of course), then I try to identify the PLL *without* rotating U, just by looking at the two sides facing me. Macky claims on his page that this recognition by two sides can be done for any PLL. I'm not sure if that's true or not, but I have definitely noticed an improvement in my recognition time, and there are definitely some angles in some PLLs that I can immediately id.

I definitely need to work of finger tricking my algs better, most of my PLLs are around 3.5 seconds, but this PLL recognition is definitely paying off.

 

[AvGalen]

After OLL, you only need to look at the 6 stickers on 2 (adjacent) faces to determine the PLL. I don't know full PLL, but I can do the same for the corner-cycles and edge-cycles seperately.

I don't understand what time-attack and recognition have to do with each other?

 

[*LukeMayn*]

After OLL, you only need to look at the 6 stickers on 2 (adjacent) faces to determine the PLL. I don't know full PLL, but I can do the same for the corner-cycles and edge-cycles seperately.

I don't understand what time-attack and recognition have to do with each other?

I was just showing how slow some were

 

[rjohnson_8ball]

...Macky claims on his page that this recognition by two sides can be done for any PLL. I'm not sure if that's true or not...

Yes, it's true for 3x3. Looking that the 5 pieces on the 2 adjacent sides, the only pieces you cannot identify are the 2 edge pieces in back and the corner in back. You can deduce the hidden corner piece from the 3 corners you can see. And you can deduce how the 2 hidden edge pieces must be arranged, due to parity.

Eventually you can become hard wired so that when you glance at the pattern for those 2 sides, you can initiate the PLL right away. (I have trouble distinguishing between R and G without adjusting the upper face, but it can be done. Actually, I have trouble distinguishing most of these quickly without adjusting the up face.) I would think that it would not hurt much to adjust the upper face before doing the PLL, because it saves the time to adjust it at the end.

 

[brunson]

I made a web version of my drill program and added a cube applet to see the scramble.

http://brunson.com/drillsergeant/

Since the scramble is composed of an OLL and a PLL, only the LL is affected.

Of course the inverse of the scramble can be used to solve the LL in one step, is anyone up for recognizing the 4704 unique positions and their algs to solve? It would mean a LL solution that averages 14 moves. ;-) Who wants to be the first to claim a true 1LLL?

 

[fanwuq]

That's nice, brunson!

 

[*LukeMayn*]

I made a web version of my drill program and added a cube applet to see the scramble.

http://brunson.com/drillsergeant/

Since the scramble is composed of an OLL and a PLL, only the LL is affected.

Of course the inverse of the scramble can be used to solve the LL in one step, is anyone up for recognizing the 4704 unique positions and their algs to solve? It would mean a LL solution that averages 14 moves. ;-) Who wants to be the first to claim a true 1LLL?

If anyone, Jason

 

[badmephisto]

I made a web version of my drill program and added a cube applet to see the scramble.

http://brunson.com/drillsergeant/

Since the scramble is composed of an OLL and a PLL, only the LL is affected.

Of course the inverse of the scramble can be used to solve the LL in one step, is anyone up for recognizing the 4704 unique positions and their algs to solve? It would mean a LL solution that averages 14 moves. ;-) Who wants to be the first to claim a true 1LLL?

I like how the icon to the page is the standard Apache icon

 

[Lucas Garron]

I made a web version of my drill program and added a cube applet to see the scramble.

http://brunson.com/drillsergeant/

Since the scramble is composed of an OLL and a PLL, only the LL is affected.

Of course the inverse of the scramble can be used to solve the LL in one step, is anyone up for recognizing the 4704 unique positions and their algs to solve? It would mean a LL solution that averages 14 moves. ;-) Who wants to be the first to claim a true 1LLL?

If anyone, Jason

Hey! I have a 1-look LL
Cross > F2L > PLL

 

[linkmaster03]

I made a web version of my drill program and added a cube applet to see the scramble.

http://brunson.com/drillsergeant/

Since the scramble is composed of an OLL and a PLL, only the LL is affected.

Of course the inverse of the scramble can be used to solve the LL in one step, is anyone up for recognizing the 4704 unique positions and their algs to solve? It would mean a LL solution that averages 14 moves. ;-) Who wants to be the first to claim a true 1LLL?

If anyone, Jason

I was thinking the exact same thing.

 

[yurivish]

What if you could simply recognize the positions? You could do the OLL and PLL together after figuring out which ones you need to do from the scramble. Maybe there are even some easy tricks to figuring out a lot of them, just like you can tell which PLL to do just from the sides of the cubies.

 

[brunson]

What if you could simply recognize the positions? You could do the OLL and PLL together after figuring out which ones you need to do from the scramble. Maybe there are even some easy tricks to figuring out a lot of them, just like you can tell which PLL to do just from the sides of the cubies.

Sébastien Felix has some interesting ideas on recognizing PLLs on his page. http://pagesperso-orange.fr/absolutemind/roll-angl.html

Since I've added the AnimCube I'm thinking of adding the ability to generate just a PLL, then I can practice recognizing them by the six stickers facing me without having to scramble a cube. Great time waster at work. ;-)

Edit: Mark that one done.