Blindsolving Discussion and Help

[LordNikolai]

Thank you for your answers Zane,
today I read this tutorial for edge commutators and I try to do some solves first using UF as buffer than UB.
UB it's better I think because I use UBR as buffer for corners and parity is easy to fix leaving last two interchangable edges in UB and UR and using a J/J' or Y perm.
However I have three question to ask you:
1) UB buffer is fingertrick friendly?
2) There are other buffer I can use esier to use?
3) There are other ways to fix parity without leaving last two interchangable edges in UB and UR?
Thank you

 

[Zane_C]

Next question: My second 3bld success was at about 13 minutes, is that very slow for a new blindfold solver, or is it normal?

That's fine, it's perfectly reasonable.

1) UB buffer is fingertrick friendly?

Yes. Any buffer in the M-slice allows for easy and fast commutators using M, M' or M2 as the interchange, without the need of setup moves or rotations.

Sometimes pure BH isn't the fastest way to solve certain cases, so you might find it useful if you can execute U-perms with the bar in front.

2) There are other buffer I can use esier to use?

UF and DF are probably the 2 most popular buffers used by 3-cycle users. The only reason why you could potentially find these buffers easier, is because some people have published a list of algs/comms that they use for each particular case. (ie. here or here.)

3) There are other ways to fix parity without leaving last two interchangable edges in UB and UR?

Yes, you could switch the corners UBL and UBR, then solve the final edge with J, L or T-perm.

It's only after you've memorised the first pieces (corners or edges), that you'll know whether you have parity or not.
So ideally, if you memorise the edges first, you will be leaving 2 corners swapped. If you memorise corners first, you will be leaving 2 edges swapped.

 

[chardison1980]

well ive been trying sited solves with M2/OP trying different methods to track parity, ive been curling my toes during an M2 swap, and relaxing my foot when its with the white center up, (white center up means my foot is in its regular position) (yellow center up i curl my toes which means thats not a normal rest position for my foot) so far its been working. thanks for the ideas

 

[aronpm]

well ive been trying sited solves with M2/OP trying different methods to track parity, ive been curling my toes during an M2 swap, and relaxing my foot when its with the white center up, (white center up means my foot is in its regular position) (yellow center up i curl my toes which means thats not a normal rest position for my foot) so far its been working. thanks for the ideas

:tu Using body cues is a really useful thing to do.

Another thing you can do using your feet is to twist a foot to the left or right, when you have a corner that is twisted. That way you don't have to memorize which direction a corner is twisted, only where it is. You can also make that more complex by twisting your left foot if the corner is on the left side of the cube, and your right foot if the corner is on the right side of the cube, and this lets you do it with two corners.

You could also use your tongue to memorize where a flipped edge is, but that's a bit uncomfortable...

 

[chardison1980]

well for my corners which i know its a no-no because i know i can Ornt and Permute the corners. but it helps me is that I use A=anti-clockwise C=clockwise and my corners are numbered 1-8 started with ulb going clockwise around the top, and then dlf going clockwise so it looks like this
top view (white)
1-2
4-3
bottom (yellow)
5-6
7-8

and its easier for me to remember it that way.
then i do my edges as normal using M2
but im still working on it. thanks

 

[LordNikolai]

Sometimes pure BH isn't the fastest way to solve certain cases, so you might find it useful if you can execute U-perms with the bar in front.

Yes, you could switch the corners UBL and UBR, then solve the final edge with J, L or T-perm.

It's only after you've memorised the first pieces (corners or edges), that you'll know whether you have parity or not.
So ideally, if you memorise the edges first, you will be leaving 2 corners swapped. If you memorise corners first, you will be leaving 2 edges swapped.

Ok I will use UB and I will leave 2 edges swapped when parity happens.
I'm not an expert but I want to become it. In fact I have other question for you :

1) Is faster M2 or BH to solve edge? I know that BH use less moves (avarage 4.5 for one edge and M2 7) but is faster for fingertricks? I ask you this because M2 is very fast to do with finger and in BH not very much (probably because I just started yesterday).

2) Is advisable rotate the cube to do E and S moves in edge commutators?

3) To do cube faster I have to develop my tps, but If I train me only in blind solves and not so much in speed solves do tsp get slow or I can do only blind solve and tps get fast likewise?
Thank you

 

[AbstractAlg]

tongue? very interesting, I'll give it a try (seriously).

I started using 3-cycles (TuRBo corners) and it's amazing how easy is to determine parity - if your last corner permutation can't be done with a 3-cycle, it's a parity. And later, after M2 edges, do parity fix alg, bring corners to edges do to R-perm and that's it.

@chardison1980
try visual corner memo (tap them, follow route with fingers), it is much faster than converting each permutation in number array, and orientation into letters, it's too much informations. try visually, at first it will be pretty hard to callback memo, but in a week you'll be able to with an ease and you will be much more faster.

 

[aronpm]

1) Is faster M2 or BH to solve edge? I know that BH use less moves (avarage 4.5 for one edge and M2 7) but is faster for fingertricks? I ask you this because M2 is very fast to do with finger and in BH not very much (probably because I just started yesterday).

M2 can be faster because of the algorithms and because of the "mindlessness", because you don't need to think about commutators. It's just recall->solve. However, 3style (it's like BH, but speed-optimal instead of move-optimal) is faster than M2 in my opinion and experience.

2) Is advisable rotate the cube to do E and S moves in edge commutators?

I suggest always rotating to use M slices, or finding a different commutator which uses the M slice. If you do a few more moves to get a much faster commutator, that's still a good thing

3) To do cube faster I have to develop my tps, but If I train me only in blind solves and not so much in speed solves do tsp get slow or I can do only blind solve and tps get fast likewise?
Thank you

If you get fast TPS in blindsolving, you'll have fast TPS in speedsolving too. But you still have to use lookahead in speedsolving, which can slow down your TPS in speedsolving. Actually, I can get 6-7 TPS in speedsolves but in blindsolves my TPS is usually 4-5 because I pause so much.

 

[LordNikolai]

However, 3style (it's like BH, but speed-optimal instead of move-optimal) is faster than M2 in my opinion and experience.

I imagined it. I begin to use 3cycle commutators immediately!

I suggest always rotating to use M slices, or finding a different commutator which uses the M slice. If you do a few more moves to get a much faster commutator, that's still a good thing

Now when I think of a edge commutator I can't think of rotation to optimize it. Probably that will happen when I can use better commutators

 

[chardison1980]

tongue? very interesting, I'll give it a try (seriously).

I started using 3-cycles (TuRBo corners) and it's amazing how easy is to determine parity - if your last corner permutation can't be done with a 3-cycle, it's a parity. And later, after M2 edges, do parity fix alg, bring corners to edges do to R-perm and that's it.

@chardison1980
try visual corner memo (tap them, follow route with fingers), it is much faster than converting each permutation in number array, and orientation into letters, it's too much informations. try visually, at first it will be pretty hard to callback memo, but in a week you'll be able to with an ease and you will be much more faster.

ill give it a try, and get back to you on how it turns out,

 

[mariano.aquino]

Hi everybody! I'm looking for an alg I once saw, but can't find right now...
maybe you guy people can help me. It's the pochmann Y alg, but with flipped edges. Meaning: ULB-RDF,UB-LU.
Anyone...? Thanks!

 

[LordNikolai]

Hi everybody! I'm looking for an alg I once saw, but can't find right now...
maybe you guy people can help me. It's the pochmann Y alg, but with flipped edges. Meaning: ULB-RDF,UB-LU.
Anyone...? Thanks!

I think this algorithm can not exist because in the cube you can flip only 2 edge at the same time and in this case you flip only one edge.

 

[RyanReese09]

Uh...I don't know why you would want this...but

S, Vperm, M'

Depends onhow you execute but after the vperm you should be able to see what you have to do.

I don't know if this is what you meant but...

 

[mariano.aquino]

I currently use freestyle for bld, and when i get parity, I make sure the edge parity ends up in UB-UL, and fix it with old pochmann's Y, targeting the other corner. I thought maybe such algorithm could make setups easier in some cases. Similar to what i just read from aronpm in his 22LL list.

"I don't know if this is what you meant but..."
It was not exactly what i was looking for but thanks for helping anyway!

 

[aronpm]

I am far from being 'pro' at 3BLD but I was interested to know how some of the pros memo so quickly, I am looking at the people who can memo sub-1 consistently or even sub 20 seconds, how is that even posssible?

My recommendation is to learn a letter scheme like [wiki]Speffz[/wiki], and use the letters to make sentences for both edges and corners. Memorize edges first and solve them last. Once you are used to your letter scheme, you can switch to an audio-loop system for corners, which is faster, but hard to be fast if you aren't accustomed the the letter scheme. I average about 9 seconds to memorize the cube with sentence/audio edges and audio-loop corners.

Oh, and practise a lot

 

[TheMachanga]

I am far from being 'pro' at 3BLD but I was interested to know how some of the pros memo so quickly, I am looking at the people who can memo sub-1 consistently or even sub 20 seconds, how is that even posssible?

Practice and confidence.

I can memo sub 1 sometimes. Use letters to name your pieces. Don't name the pieces themselves, but the position, or destination. I don't know what method you use, but I use m2, and I name my edges like this (hold it with yellow top, orange front): YG edge: A. GY edge: B. OG edge: C. GO edge: D. And keep going until you get to H. The I switch over to the right, YB edge: I. BY edge: J. OB edge: K. BO edge: L, etc. The RY edge: Q. YR edge: R. YO edge: S, etc, until x. If found this WAY faster. If you use the color method, ex. OG edge: Ogre --> shrek, then you're remembering 2 different letters, and can only memorize 1 pair at a time. Which brings me to my second point.


Memorize in pairs, or if possible, 3 or 4 at a time. Random letters:

QDBKRG. Instead of "Queen Dropped the Book on the Keg while Riding Gary". Do

the QuiDich (or however you spell the harry potter game) BooK is on the RuG.


Confidence is important. I still go over my corner memo (the last thing I memo and the first thing I solve) twice, and spend a lot of time one it, but while solving, it only lasts 10-20 seconds of solving, and I get angry at myself for spending so much time on it, because its already done with. If you memo quicker, it will still be fresh in your memory.

 

[AbstractAlg]

confidence.

to ask 'pros':

1) have you made letter-pair list for each case or just intuitively make up something on place, or you just got used to same naming pairs?
I ask because I started creating list and stopped at ~50th word, but making up on scene takes time...

2) is there some video or written tutorial which explains all types (not just pure) commutators (that is not bldmaster)?

Thanks.

 

[Litz]

^ Audio-loop is basically auditory/echoic memory. It's what makes you remember the sounds you heard in conversations for a few seconds for example. You can use it to loop your letters or whatever you use in your mind without thinking about it and you'll still remember it.

For more information: http://en.wikipedia.org/wiki/Echoic_memory

 

[aronpm]

^Thanks. One other question, I always see people use either M2/classic Pochmann or something else for corners, never seen many people use R2. Is it because it's considered to be inefficient or are there obvious better alternatives? Sorry for the amount of questions, it's just that I've become increasingly interested in blindfold cubing lately.

In the past, I've said that R2 is bad because the algorithms to solve cases on the R layer are too long compared to L layer cases. However, the movecount is still much lower than Old Pochmann, spammin' Y perms all over the joint. There's also the argument that, because R2 uses a DFR buffer, it's harder to switch to a 3-cycle method later on, but I think Marcell Endrey has proved that using a DFR buffer for 3-cycles is still pretty damn fast. I think the only remaining 'disadvantage' is that you have to learn more algorithms for it, but if you're determined to learn R2 it's not a problem.

Of course, R2 is very heavy on <R,U,L> so if you prefer <R,U,D> you might not want to use it. Or you could rotate all algs, so D2 is the interchange and insertions are pure <R,U>, and using DLB as the buffer, but I've never seen this used fast (or at all), and I think using 3-cycles from DLB might be weird.

Questions are fine! It's good to see people becoming interested in BLD.

 

[Divineskulls]

I use Old Pochmann corners and edges at the moment, and I am planning on switching to Turbo edges in the very near future. What should I do about corners? Is there a Turbo-like method that exists for corners, or should I attempt to make the jump straight to simple commutators? OR, should I just stay with Old Pochmann corners? Thanks in advance.