• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 40,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

Two-sided PLL recognition guide

mark49152

Premium Member
Joined
Oct 29, 2012
Messages
4,719
Location
UK
WCA
2015RIVE05
YouTube
Visit Channel
I'm having trouble differentiating between R perm and T perm as well as A perm from G perm from this angle.
I mean, it takes a long time to check whether you can see all 4 colors or not.
At first it does, but after a while you just see it without thinking or checking. It's hard to describe, but when solving I see different "weights" of colours between those cases, rather than specifically thinking about which colour is where. So A and T have a single strong colour but G and R have a second strong colour that also stands out.
 
Joined
Jun 26, 2018
Messages
1
But then how do you differentiate between Ja/Jb, Ga/Gc, and Gb/Gd ones? Having them grouped together like that isn't much help when having to pick out just 1 of them.

@mark49152
 

Tom Joad

Member
Joined
Jun 29, 2016
Messages
137
But then how do you differentiate between Ja/Jb, Ga/Gc, and Gb/Gd ones? Having them grouped together like that isn't much help when having to pick out just 1 of them.

@mark49152

The way to learn two sided pll recognition is to make a 4 by 21 table on one piece of paper and write YOUR OWN description of each pll from each side in each box.

First you make the empty table then you do slow solves and when you get to pll you write the description of what you see in the appropriate box.

You’re right, having them in groups doesn’t help. Also the writing the description of what you see in words doesn’t get much press but it is an integral part of the most efficient learning process.

If you want me to post my table let me know and I will post it tomorrow then tell you what to do next.
 

mark49152

Premium Member
Joined
Oct 29, 2012
Messages
4,719
Location
UK
WCA
2015RIVE05
YouTube
Visit Channel
The way to learn two sided pll recognition is to make a 4 by 21 table on one piece of paper
Whatever works for you. Most guides start with a list of PLLs and document what they look like from each angle, as you suggest. That didn't work for me, and I preferred to document the colour patterns as seen then deduce from those which PLL it is. Hopefully it helps some people, but it's personal preference, there's no right or wrong way to learn it.
 
Joined
Jun 25, 2017
Messages
2
So, I decided to learn two-sided PLL recognition, just for the fun of it. It looked pretty complicated at first, but it turned out to be much easier than I expected. When trying to recognize from two sides, you see six stickers in a row. There are only 35 patterns for these stickers, and each pattern indicates a unique PLL. Most of the patterns are easy to recognize and remember.
I don't get how some of them that are fully filled out say It is GA/blah blah. If it were fully filled out would'nt the case be definite and not one of the 2?
 

Tom Joad

Member
Joined
Jun 29, 2016
Messages
137
I don't get how some of them that are fully filled out say It is GA/blah blah. If it were fully filled out would'nt the case be definite and not one of the 2?

His guide is telling you that certain features of the pll show you that it is either a Ga or a Gc and you must then use further information to ascertain which one it is. I found the following method the best way to alleviate this problem although there are a handful of subtle cases which are so similar that I still find that I need to cross reference two pieces of information before I know which one it is so I pause for longer as I recognise the pll (this will continue until I start to see the “weights” of colours Mark talks about. Maybe he will elaborate on this).

Anyway:

I average 18 seconds (my f2l and look ahead are weak) so take this with a pinch of salt...

I started focusing on two sided pll recognition five months ago. I've improved vastly but still have a lot to learn.

I made my own sheet of paper with the 21 cases listed down the left hand side and four columns. I named the columns Go!, U, U' and U2.

So I had my 84 plls to recognise, 21 each from 4 angles. "Go!" Was when it was ready for the algorithm I had learnt for the pll.

Now... I sdtarted my slow solving... when I got to the pll, if it was a case from an angle I recognised, then I put a tick in the corresponding box in my table. If I didn't recognise it, I looked around the cube to see which pll it was and how many U turns I needed to do to make it the right angle for my algorithm. I then wrote my own description of the case in the corresponding box in my 84-box table.

I think this is key: Writing your own description. You have to describe the case by its redeeming features which are specific to that case and no other. And it has to be what you recognise. Different people notice different features.

This is when I realised how little I knew and how much I had to learn!

Eventually I had a table with about 34 ticks in it and 50 descriptions. The descriptions were along the lines of:

Three checker in three bar (shifted right), opposite edition.
Three checker in three checker (shifted left), adjacent edition.

These are the harder cases with less succinct descriptions which still remain in my table, which I have redrawn four or five times. Now it has about 74 ticks in it and about 10 descriptions left.

So, once your table is complete...

It is key as you progress to never peak at a third side in your slow solves. Stare at the two sides and try to work it out. If you don't know it, check your table and look back at the cube until you think "I'll remember that next time and I won't have to look at the table"

This method has a nice "compounding reward" associated with it because every time you learn a case, you are also learning all the other cases (because there's one fewer they can be - the one you just learnt!)

By definition, the ten or so descriptions which remain in my table are for the cases which are most similar to each other.

You will change your descriptions as you roll along, making you descriptions of tricky cases more succinct as you describe only the differences which mark them out as unique.

The two example descriptions above are typical in that they pair with one other case (so you should start to notice similar pairs of descriptions across your table)

First you just notice enough information to know "it's either x or y" because you see "three checker in three bar" (for example) Then you notice the extra information (adjacent or opposite) and you know whether it is x or y. Again this is double-rewarding because once you start to learn which one is x, then ipso facto you start to learn which one is y (because it's not x).

Por supuesto, as time passes by, your brain doesn't see "five checker on left" (or whatever), it just recognises the case. This is what your ticks are for in your table. Again this is key. There is no point writing down descriptions of cases you recognise and secondly, each tick is minimising the possibilities for all the cases you don't know!

A couple of final points:

1. You never fully know two sided pll recognition. Like look ahead, it just improves indefinitely i.e. the pause time between oll and pll decreases. But watch the fastest cubers: there is still a pause there. The key is to start off on the road to minimising this pause and the first step is to stop peaking at a third side for further information (so when I say I have 74 ticks in my table, sometimes I still stare at the case for five seconds in a slow solve before my brain recognises it but eventually I think "T perm Go!" and the next time it will be quicker etc)

2. Many cubers, most faster than myself, fundamentally misunderstand this and you will read dreadful advice across this forum like "I find it quicker to just peak at a third side" which is even more flawed than the immortal "you can be sub whatever without full oll so there is no point in learning it"

3. Soon after starting this method, I realised that I was learning more than I set out to. Now I am recognising a case by both perm AND angle and thinking "this is Ja perm from the U2 angle" (for example). To compound this learning during slow solves, close your eyes after you recognise the case and complete both the U turns and perm all at once.

4. Extrapolating this idea, the next step would obviously be to also recognise the auf that will be required at the end so that eventually you can see any pll from any angle and close your eyes and complete the three steps in one go: auf - pll - auf (this is something I look forward to teaching myself once my table has 84 ticks in it )

Happy cubing
 

mark49152

Premium Member
Joined
Oct 29, 2012
Messages
4,719
Location
UK
WCA
2015RIVE05
YouTube
Visit Channel
I don't get how some of them that are fully filled out say It is GA/blah blah. If it were fully filled out would'nt the case be definite and not one of the 2?
In that case, the two options correspond to mirror images.

For example, if you see one side (either left or right) has headlights, and the other has a 2-bar touching those headlights, you know you have either an R perm or a T perm. If the other (single) stickers are the same colour, it's a T perm, which is symmetrical. If they are different colours it's Ra or Rb perm. Which of those it is depends on whether the headlights are to the left or right. They are mirrors of each other.

The guide points out the visual cues you can see during a solve to recognise PLL from only two sides. While learning, it is (IMHO) the simplest way to deduce which case you have, because you base your reasoning purely on the colour pattern in front of you. Over time, it becomes automatic and you simply recognise without thinking or deducing. You don't need to learn all the patterns at once - it's very easy to get started with the obvious patterns, and trickier patterns can be added later.
 
Top