Blindsolving Discussion and Help

[Jezuz]

Sorry if this has been answered but this is a HUGE thread.

I think I understand the basic idea of freestyle (using commutators to solve 2 pieces at the time). However, I have a couple of question regarding this:

1. Since you are cycling three pieces at a time you will solve two pieces and put the third in the buffer with every move. This means that every cycle solves two pieces. So how do you deal with an odd number of pieces to solve?

2. Probably related to the previous question: How do you deal with parity?

3. Finally, how does one deal with edges and corners that are already in the right permutation but has the wrong orientation? (In the M2/Old Pochmann method I am currently using I just shoot at one of the stickers and then the second sticker to solve this).

4. As pieces with the correct permutation but wrong orientation are not dealt with in the same manner as I am used to, I guess this changes things for how to memorise them as well. I guess I need to somehow memorise which pieces are oriented wrongly and (for corners) in what way their orientation is.

I understand that many of these things can be dealt with in a variety of ways, but I am just looking for someone fairly experienced in 3BLD to tell me of a good way to deal with these problems?

 

[cmhardw]

1. Since you are cycling three pieces at a time you will solve two pieces and put the third in the buffer with every move. This means that every cycle solves two pieces. So how do you deal with an odd number of pieces to solve?

When you have an odd number of pieces in a cycle, the last location is the buffer piece, which like you said returns to the buffer location.

2. Probably related to the previous question: How do you deal with parity?

If you have parity in the wing edges you can solve down to 2 pieces swapped and then use a parity algorithm to swap those two. If you have parity in the corners or edges of a 3x3x3 then you can solve as much as you can down to 2 corners and 2 edges swapped. Then you use a PLL like T, V, Y, R to swap those pieces.

3. Finally, how does one deal with edges and corners that are already in the right permutation but has the wrong orientation? (In the M2/Old Pochmann method I am currently using I just shoot at one of the stickers and then the second sticker to solve this).

You can twist/flip these pieces (with the buffer piece if required) at the start of the solve, or at the end of the solve.

4. As pieces with the correct permutation but wrong orientation are not dealt with in the same manner as I am used to, I guess this changes things for how to memorise them as well. I guess I need to somehow memorise which pieces are oriented wrongly and (for corners) in what way their orientation is.

I memorize them in a very quick, ephemeral sort of way and solve them first.

I understand that many of these things can be dealt with in a variety of ways, but I am just looking for someone fairly experienced in 3BLD to tell me of a good way to deal with these problems?

As of late I average about 2 minutes on 3x3x3 BLD. At my best I was averaging about 1:15. Take my advice with a grain of salt.

 

[Raviorez]

hello everyone

I really want to learn how to solve a rubik's cube blind.
Which method should I learn?
Classic Pochmann/ OP/M2 / OP/TuRBo??
Which one is the fastest of the 3? Which one do you recommend and why?
Which colourscheme should I use?

 

[ollicubes]

hello everyone

I really want to learn how to solve a rubik's cube blind.
Which method should I learn?
Classic Pochmann/ OP/M2 / OP/TuRBo??
Which one is the fastest of the 3? Which one do you recommend and why?
Which colourscheme should I use?

I recommend you to learn M2 and Old Pochmann. They're not too hard to learn and quite fast when you get used to. OP/TuRBo is fastest of these 3 but diffrence is not huge. Just pick your own color scheme whatever you want.

 

[TheNextFeliks]

hello everyone

I really want to learn how to solve a rubik's cube blind.
Which method should I learn?
Classic Pochmann/ OP/M2 / OP/TuRBo??
Which one is the fastest of the 3? Which one do you recommend and why?
Which colourscheme should I use?

Learn M2/OP. It is relatively easy to learn. It can get really fast (people average under a minute with it). Check out Noah's tutorial on Cubing World

Whatever color scheme you feel most comfortable with. Some choose White/Green since it is WCA scramble orientation. I would either use the one from the tutorial, or do a natural orientation. If you use CFOP, put cross color on D.

 

[Jezuz]

Oh! Did not expect to get an answer from a celebrity! Thanks for taking the time!

Seems like a made a mistake in my first question! What I really meant was what you should do when you have an EVEN number of pieces out of position. In this case, after finishing the last 3-cycle, you will have two more pieces out of position (probably the buffer and one more piece, but could also be two pieces in a separate cycle).

To give an example, let's say that all the edges are out of place and when memorizing I reach the buffer after 11 letters (so that the very last piece is the buffer). In this case I will also have parity so in my usual method I would finish the last piece with an M2-algorithm and then make a parity algorithm (which case rotates the M-layer 180 degrees and switches UL and UB). Then go on to solve the corners using Old Pochmann.

If I would try to solve the same case with commutators I would be left with the buffer and another edge out of place after finishing the edges. Hmm.. I guess in that case I will also end up with the buffer corner and another corner with the wrong permutation! I think I see now. Then I can just make some setup moves and make for example a T-perm (the only one I know at moment).

Hmm... I guess it is still not completely clear how to make sure that the 4 last pieces in this case end up in the correct orientation... I guess that if you are unlucky the last PLL makes the last pieces end up with the wrong orientation.

Another troublesome case would be if you have a pair of interchanged pieces or any cycle with an even number of pieces.

Is there any good tutorial on the full method somewhere? Preferably with a few example solves...

 

[Raviorez]

I recommend you to learn M2 and Old Pochmann. They're not too hard to learn and quite fast when you get used to. OP/TuRBo is fastest of these 3 but diffrence is not huge. Just pick your own color scheme whatever you want.

Thanks, I'll learn M2/OP

Learn M2/OP. It is relatively easy to learn. It can get really fast (people average under a minute with it). Check out Noah's tutorial on Cubing World

Whatever color scheme you feel most comfortable with. Some choose White/Green since it is WCA scramble orientation. I would either use the one from the tutorial, or do a natural orientation. If you use CFOP, put cross color on D.

Thanks, I'll learn M2/OP. I use CFOP so I'll use yellow on top, Blue on front. Is that a good orientation?

 

[cmhardw]

Oh! Did not expect to get an answer from a celebrity! Thanks for taking the time!

Glad to help, but I consider myself more of a cubing old-timer

Seems like a made a mistake in my first question! What I really meant was what you should do when you have an EVEN number of pieces out of position. In this case, after finishing the last 3-cycle, you will have two more pieces out of position (probably the buffer and one more piece, but could also be two pieces in a separate cycle).

To give an example, let's say that all the edges are out of place and when memorizing I reach the buffer after 11 letters (so that the very last piece is the buffer). In this case I will also have parity so in my usual method I would finish the last piece with an M2-algorithm and then make a parity algorithm (which case rotates the M-layer 180 degrees and switches UL and UB). Then go on to solve the corners using Old Pochmann.

If I would try to solve the same case with commutators I would be left with the buffer and another edge out of place after finishing the edges. Hmm.. I guess in that case I will also end up with the buffer corner and another corner with the wrong permutation! I think I see now. Then I can just make some setup moves and make for example a T-perm (the only one I know at moment).

Those are all good methods for what to do in this case. The root of this issue is that you cannot solve such a case using commutators. Commutators will always contain an even number of quarter turns, and are thus considered even permutations on the pieces. When you solve everything in a piece orbit, but still have two pieces swapped that is odd permutation parity, which must be solved using an odd permutation on the pieces, and so commutators are no longer an option.

Hmm... I guess it is still not completely clear how to make sure that the 4 last pieces in this case end up in the correct orientation... I guess that if you are unlucky the last PLL makes the last pieces end up with the wrong orientation.

Try to always setup to the same layer, say the U layer. Then think of it as sticker switching. Which sticker on the first corner switches with which sticker of the second corner - remember that when you switch a sticker, the whole corner piece comes with.

For example, the standard U perm swaps UBR with UFR and UL with UR. However it also swaps RBU with FRU and LU with RU. There are other ways to think of the swaps that also are true.

Another troublesome case would be if you have a pair of interchanged pieces or any cycle with an even number of pieces.

Is there any good tutorial on the full method somewhere? Preferably with a few example solves...

A pair of interchanged pieces can be solved using commutators.

Two methods to solve the cycle case (AB)(CD) which is A and B swap with each other and C and D swap with each other are below:

1) Do the cycle (ABC) followed by (ADC).

If you start with (AB)(CD) and do the three cycle (ABC) you will solve the B location, and you will put the piece that belongs in the A location into the C location. The piece that started in the C location needs to go to spot D, and you moved it to spot A. That leaves the cycle (ADC) to solve all pieces.

2) You can do the cycle (ACD) followed by (ACB).

If you start with (AB)(CD) and do (ACD) you are "breaking into a new cycle", meaning you are cycling a piece into the (CD) cycle. When you break a new cycle and solve it through to the end, you have to "close" the cycle by shooting to the same spot where you broke into the cycle before moving on (in this case moving on means solving the rest of the (AB) cycle).

So finish the solution with (ACB) which will close the (CD) cycle that you broke into, then finish off solving the (AB) cycle.

P.S. Notice that method 1 is also using the technique of "breaking into a new cycle"

 

[Jezuz]

A pair of interchanged pieces can be solved using commutators.

Two methods to solve the cycle case (AB)(CD) which is A and B swap with each other and C and D swap with each other are below:

1) Do the cycle (ABC) followed by (ADC).

If you start with (AB)(CD) and do the three cycle (ABC) you will solve the B location, and you will put the piece that belongs in the A location into the C location. The piece that started in the C location needs to go to spot D, and you moved it to spot A. That leaves the cycle (ADC) to solve all pieces.

2) You can do the cycle (ACD) followed by (ACB).

If you start with (AB)(CD) and do (ACD) you are "breaking into a new cycle", meaning you are cycling a piece into the (CD) cycle. When you break a new cycle and solve it through to the end, you have to "close" the cycle by shooting to the same spot where you broke into the cycle before moving on (in this case moving on means solving the rest of the (AB) cycle).

So finish the solution with (ACB) which will close the (CD) cycle that you broke into, then finish off solving the (AB) cycle.

P.S. Notice that method 1 is also using the technique of "breaking into a new cycle"

Ah! Now I see. I guess the "worst" case is that you end up with two corners and two edges swapped, which is then solved by a PLL.

Thanks again for the help!

 

[Tao Yu]

Ah! Now I see. I guess the "worst" case is that you end up with two corners and two edges swapped, which is then solved by a PLL.

Thanks again for the help!

What a lot of BLDers do is influence which two corners or two edges are swapped, so it's easier to figure out the PLL to use.

What buffers do you use? And do you execute edges or corners first? I can suggest a way of doing it.

 

[giorgi]

How fast can you get in 3x3 BLD using classic pochmann? My PB is 6:45.00 without using M2 edges and I am just interested how fast can I get with the method I use Thanks

 

[Goosly]

Worldclass BLD'ers can memo in ~10 seconds. Let's assume, if you execute fast with OP, you need 3 seconds for every target (including setupmoves) and for parity. That would make 10+(11+1+7)*3 = 67 seconds for a standard 11/7 scramble.

 

[Jezuz]

What a lot of BLDers do is influence which two corners or two edges are swapped, so it's easier to figure out the PLL to use.

What buffers do you use? And do you execute edges or corners first? I can suggest a way of doing it.

I use DF and UBL as buffers. I think I will stick to M2 for edges and first learn freestyle on the corners.
Currently I do edges, (parity if needed) and then corners. This is just because I found the parity algorithm in M2/Old Pochmann easier this way.

I would love to hear your suggestion though!

 

[Tao Yu]

I use DF and UBL as buffers. I think I will stick to M2 for edges and first learn freestyle on the corners.
Currently I do edges, (parity if needed) and then corners. This is just because I found the parity algorithm in M2/Old Pochmann easier this way.

I would love to hear your suggestion though!

This parity fix method would be perfect for you (Noah also uses UBL/DF with edges first. It's also what I do).



Edit: Your current parity alg will actually still work, believe it or not, as long as you use UBL/DF with edges first. Just solve the last edge target using M2, do your parity alg, solve the corners and then do the last target with OP. Slower than Noah's method though.

 

[Cubenovice]

How fast can you get in 3x3 BLD using classic pochmann? My PB is 6:45.00 without using M2 edges and I am just interested how fast can I get with the method I use Thanks

Sub 60 is possible, as demonstrated by Cornelius.

 

[sk8erman41]

I understand cycles and identifying them and everything but my question is what is the best way to find an empty spot to shoot the buffer to when you get a new cycle? When this happens to me I have to retrace all the memo I've already done and it takes long/screws me up. I have heard of putting a finger on each cubie that is used, but that seems awkward to me. Thoughts?

 

[A Leman]

I understand cycles and identifying them and everything but my question is what is the best way to find an empty spot to shoot the buffer to when you get a new cycle? When this happens to me I have to retrace all the memo I've already done and it takes long/screws me up. I have heard of putting a finger on each cubie that is used, but that seems awkward to me. Thoughts?

I think many people have a few targets that they give preference to. My buffer is UF and I like to break into a new cycle at UB,UR,DF,UL the most and I tend to passively take note of already shooting to the places that I typically break into. For really fast people, it becomes natural and they just do it without much thinking.

 

[JasonDL13]

Here's my question. I'm currently trying to get my first success with bld solves. I'm just screwing up on the edges a bit. Currently I use Old Pochmann. When I get my first bld solve I want to get a better method (for edges) since Old Pochmann is very slow. What method do you recommend? I have seen M2 and I know the basics but it looks hard to me. Probably just because I haven't learned all of it.

EDIT: I'm not going to make another post but thanks to the people below that helped me. I do know how to do 2x2 bld using Old Pochmann though.

 

[Goosly]

Switching from OP to M2 right after your first success won't help a lot. But if you really want to switch, I recommend M2.

 

[TheGrayCuber]

Practicing for a while with OP until you can consistently get successes will help you as a BLD solver. I do suggest switching to M2 at some point, but not until you have more successes under your belt and are more familiar with BLD. OP is a great way to familiarize yourself with BLD solving before confusing yourself with the harder methods and it will give you time to really get to know your memo, so by the time you switch to more advanced methods, you won't have to worry about the simple, background things. I would advise you to use OP until you have more practice and are more consistent.