COLL recognition

[WTSnacks]

Hi.

I'm at the beginning at learning COLL, and I'm already stuck at recognition.
At first, I look at U-color stickers in U-face and recognize one of the 7 group cases. Then, I suppose, I should look for something in F/R/L (and maybe B) faces, but for what exactly? I can, of course, look at ALL corner stickers and determine what goes where, but it will take several seconds, and I suppose people don't recognize COLL that way.

It appears there are two description systems for COLL: "normal" and hyperorientation. I somewhat understand hyperorientation, where I should look only for F/B-colored stickers. But I totally don't get the "normal" system. Can anyone help me?

Example: let's get a solved cube, yellow top, blue front, standard color scheme, and scramble it with Anti-Sune and T-perm (RU2R'U'RU'R', RUR'U'R'FR2U'R'U'RUR'F'). How do you recognize a COLL case from this point? First, you automatically recognize it as Sune-group case by yellow stickers. What are the EXACT corners and stickers you look at next?

 

[Cride5]

Dan's cube station has good info on COLL here.

All cases can be uniquely identified by looking at the 4 non-U colours on the upper and front stickers in the U-slice. It turns out, that for every case (apart from a completely oriented U face) there will be exactly 4 non-U colours visible in the top and front of the U-face (provided the U face is rotated correctly). The link has a diagram showing the 6 corner stickers you need to look at in white.

Which exact stickers you look at in the top or front of the U face depends on the case. For example, with a sune, you have 3 non-U colours on the U face, and if you position the sune so that the only U-face colour is in the top-left then the fourth non-U colour is in the front-left U-slice.

.. I hope this makes sense!

 

[masterofthebass]

I use this method. That page uses CMLL algorithms, but you can look up the COLL algorithms elsewhere. I added this recognition system to the speedsolving wiki, so you can also use it there.

The advantage of this method over the other method, is that you are looking at positions of opposite stickers, instead of looking at the colors of certain positions. It's just a small difference that I thought worked well.

 

[WTSnacks]

All cases can be uniquely identified by looking at the 4 non-U colours on the upper and front stickers in the U-slice. It turns out, that for every case (apart from a completely oriented U face) there will be exactly 4 non-U colours visible in the top and front of the U-face (provided the U face is rotated correctly). The link has a diagram showing the 6 corner stickers you need to look at in white.

Thanks! It totally makes sense now.

The advantage of this method over the other method, is that you are looking at positions of opposite stickers, instead of looking at the colors of certain positions. It's just a small difference that I thought worked well.

Yes, I understand hyperorientation (and find it a bit easier), it's just all the sites with algorithms in the internet are using pictures from standard method, which I didn't understand until now.

 

[Sune]

Since this is a COLL thread, I might as well ask here.

In which order do you recommend learning COLL cases for Petrus? My thoughts:

I might do one of the 7 basic OLL/COLL algs on yet unknown cases and then PLL, but I don't know any PLLs besides J/Y and don't plan on learning them in the near future.

S/As/L can be converted to H and U/T can be converted to Pi by a basic Sune/Antisune alg.
Moreover, H and Pi are best for recognition, H is 4 algs only and Pi cases = combinations of Sunes/Antisunes/Niklases. So I think of learning H+Pi first, then U+T, then S+As+L. Is this optimal?

 

[Athefre]

In which order do you recommend learning COLL cases

Learn the hardest to recognize first.