[Help Thread] CFOP Discussion

[Llewelys]

A thing to keep in mind: you can't be sub20 if it takes you 15 seconds to do F2L.
If you reach last layer at 17-18 seconds, knowing WV, ZBLL or COLL won't help at all.

As PetrusQuber said, practicing should do the trick: it'll help you gain back muscle memory and will make recognising patterns easier

 

[WinterIsComing]

Earlier my F2L and Last Layer times were almost same. Now F2L times have just doubled up and Last Layer times have just increased by a few seconds.

 

[Chinmay47]

I need help with the Cross and the F2L!!

Just timed each of my steps by CFOP. Here are the results, Can somebody help?

P1 P2 P3 P4 are Cross F2L OLL and PLL respectively here.

 

[ProStar]

What happened to solve 3? Did you get LL skip and forget to change between cross and f2L?

 

[I'm A Cuber]

Just timed each of my steps by CFOP. Here are the results, Can somebody help?View attachment 12463

P1 P2 P3 P4 are Cross F2L OLL and PLL respectively here.

Plan cross in inspection, watch jperms advanced f2l video

 

[CubingNewfie]

Efficient crosses are important. You should be able to do cross in less than 8 moves. Always try to find a good cross solution. You might already know this, but you can make cross with pieces in the wrong spot, but solved relative to each other.

 

[Chinmay47]

What happened to solve 3? Did you get LL skip and forget to change between cross and f2L?

Solve 3 is the best out of them. Actually I got a nice cross, got 2 F2Ls slotted already. An H- shape OLL and finally a JPerm. It also turns out to be my current PB!

Efficient crosses are important. You should be able to do cross in less than 8 moves. Always try to find a good cross solution. You might already know this, but you can make cross with pieces in the wrong spot, but solved relative to each other.

Fairly a nice trick buddy! But can you please elaborate?

 

[ProStar]

Solve 3 is the best out of them. Actually I got a nice cross, got 2 F2Ls slotted already. An H- shape OLL and finally a JPerm. It also turns out to be my current PB!

You did a Jperm including recognition in .21?

 

[Chinmay47]

You did a Jperm including recognition in .21?

Yes I love J Perms! I can do them even faster than J Perm himself! And surprisingly the Cube I am using currently is the Rubiks brand 3x3 with a bit of cotrolius lube. Its not even adjusted or tensioned in any way.

Sometimes when I watch my own solves, I too feel like, "How did I do this so fast!?" Maybe its just me or its with everyone else! Who knows?

 

[ProStar]

Yes I love J Perms! I can do them even faster than J Perm himself! And surprisingly the Cube I am using currently is the Rubiks brand 3x3 with a bit of cotrolius lube. Its not even adjusted or tensioned in any way.

Sorry dude, but you don't have 61 TPS on a J perm. And that's without pausing for recognition

 

[CubingNewfie]

Fairly a nice trick buddy! But can you please elaborate?

I just mean make sure that you are doing your crosses as efficiently as you can. J Perm has a few videos on good cross tricks.

 

[Sub1Hour]

Sorry dude, but you don't have 61 TPS on a J perm. And that's without pausing for recognition

I did the math and if he is using a 57mm Rubik's brand cube then let's just say that's how far his finger moves during a turn hypothetically. If the J perm is the 13 move one most people use in 0.21 seconds assuming his recog is instant and pauseless. If we take 13 and divide it by .21 we get 61.90476 TPS and I'll round that to 62 for the sake of convenience. If we multiply that by 57 we get 3534mm so thats how far his fingers can go in 1 second. After doing a conversion to meters his fingers can move at around 3.5 m/s on a rubiks brand cube? This math is not 100% accurate in any way shape or form since I dont know what fingertricks he would use and I also did not count any double turns. Just a little thought-provoking "exercise" that's just a bunch of useless math. All I'm saying is @ProStar is correct and its impossible to do this on any modern hardware, let alone a crappy rubik's.

 

[EngiNerdBrian]

Fairly a nice trick buddy! But can you please elaborate?

From the 11 solve splits you posted it looks to me like your OLL and PLL are the strongest parts of your solve which makes sense because they are algorithm based. There is no trick to the cross other than practice and beginning to recognize relationships between how the pieces move relative to one another during solution...this is a hard skill to learn. The Jperm video referenced talks about using 1 piece to insert another and the "replacement technique." Those are the types of relationships i am talking about you will only be able to see with practice. You had some other threads or questions recently regarding F2L...Again, this is another portion of the solve that is heavily influenced by intuition and recognizing piece relationships. Slow and deliberate solves are your friends here. During OLL and PLL you can let algorithms be your success, i believe cross and F2L require a much more solid understanding of cube mechanics to be able to solve quickly and efficiently.

I did the math and if he is using a 57mm Rubik's brand cube then let's just say that's how far his finger moves during a turn hypothetically. If the J perm is the 13 move one most people use in 0.21 seconds assuming his recog is instant and pauseless. If we take 13 and divide it by .21 we get 61.90476 TPS and I'll round that to 62 for the sake of convenience. If we multiply that by 57 we get 3534mm so thats how far his fingers can go in 1 second. After doing a conversion to meters his fingers can move at around 3.5 m/s on a rubiks brand cube? This math is not 100% accurate in any way shape or form since I dont know what fingertricks he would use and I also did not count any double turns. Just a little thought-provoking "exercise" that's just a bunch of useless math. All I'm saying is @ProStar is correct and its impossible to do this on any modern hardware, let alone a crappy rubik's.

I am going to post a rather useless correction to your calcs. It's all in good fun and i just can't help myself...the lack of relating basic mathematics to real world situations is why i think many people think math is hard, boring, lame, useless, *insert every child's complaint about school here*. I'm happy to see a post like this on the forums.

I would argue that the distance the hand travels during any given turn is equal to 1/4 of the circumference of a circle that can be inscribed within the square represented by the size of the cube (no regrip & no finger tricks). For a 57mm cube the hand actually travels 44.8mm/turn. This gives 2.78m/s per your method which is still quite fast.

As you said, tons of factors we are ignoring & it's just a thought-provoking exercise.

 

[Chinmay47]

OK Ok! I'll reveal what happened actually. Why I call this solve the best solve is because there were some misunderstandings during it. In P1 The cross was only 2 steps far and the first two F2L pairs were already in. And then what happened was that I mistook P1 for Cross+F2L cause it was just so fast that I didn't even realise that I had done both of them so fast. That means that P2 And P3 were not actually F2L or OLL separately! In fact, I use 2 Look OLL currently, and P2 was the first look followed by P3 as the second look. And in that P3 too, that H shape OLL was the one that I recognised as what ends with a J Perm. It was actually my very first prediction of the PLL. Since this rarely showed up during much of the solves so I remembered it fully. And I was very happy with the fact that a J perm was near. So I forgot that I had to lap the timer on finishing the OLL but later got to know that I didn't do it. So I did it super quick and completed the rest of the J Perm I left which was only R2 U' R' U'. This was the actual thing that happened then.

And if you notice, my OLL is pretty strong than any other step, Cause I just learnt to predict the cases of the second look while doing the First look. I mean, I didn't take a tutorial or something. I did some OLL first looks probably 350 times. And In each case I noted the second look that showed up afterwards. With some basic probability and math, I could now integrate my results to an ultimatum. And this is what makes me stronger at OLL. The only thing I lack in here is my speed. My TPS is actually quite low. I calculated it yesterday and it was some 9.38 or something with a J Perm. This is all what I have to say about that funny but actually my best solve.

 

[Chinmay47]

From the 11 solve splits you posted it looks to me like your OLL and PLL are the strongest parts of your solve which makes sense because they are algorithm based. There is no trick to the cross other than practice and beginning to recognize relationships between how the pieces move relative to one another during solution...this is a hard skill to learn. The Jperm video referenced talks about using 1 piece to insert another and the "replacement technique." Those are the types of relationships i am talking about you will only be able to see with practice. You had some other threads or questions recently regarding F2L...Again, this is another portion of the solve that is heavily influenced by intuition and recognizing piece relationships. Slow and deliberate solves are your friends here. During OLL and PLL you can let algorithms be your success, i believe cross and F2L require a much more solid understanding of cube mechanics to be able to solve quickly and efficiently.


I am going to post a rather useless correction to your calcs. It's all in good fun and i just can't help myself...the lack of relating basic mathematics to real world situations is why i think many people think math is hard, boring, lame, useless, *insert every child's complaint about school here*. I'm happy to see a post like this on the forums.

I would argue that the distance the hand travels during any given turn is equal to 1/4 of the circumference of a circle that can be inscribed within the square represented by the size of the cube (no regrip & no finger tricks). For a 57mm cube the hand actually travels 44.8mm/turn. This gives 2.78m/s per your method which is still quite fast.

As you said, tons of factors we are ignoring & it's just a thought-provoking exercise.
View attachment 12465

Thank You! even I feel that my OLL and PLL are quite faster though! And this math seems much interesting and lucid for my understanding!

 

[Chinmay47]

I did the math and if he is using a 57mm Rubik's brand cube then let's just say that's how far his finger moves during a turn hypothetically. If the J perm is the 13 move one most people use in 0.21 seconds assuming his recog is instant and pauseless. If we take 13 and divide it by .21 we get 61.90476 TPS and I'll round that to 62 for the sake of convenience. If we multiply that by 57 we get 3534mm so thats how far his fingers can go in 1 second. After doing a conversion to meters his fingers can move at around 3.5 m/s on a rubiks brand cube? This math is not 100% accurate in any way shape or form since I dont know what fingertricks he would use and I also did not count any double turns. Just a little thought-provoking "exercise" that's just a bunch of useless math. All I'm saying is @ProStar is correct and its impossible to do this on any modern hardware, let alone a crappy rubik's.

To be honest, I actually don't know the difference between the J Perms you are talking about here. But I use R U R' F' R U R' U' R' F R2 U' R' U'.

 

[Sub1Hour]

To be honest, I actually don't know the difference between the J Perms you are talking about here. But I use R U R' F' R U R' U' R' F R2 U' R' U'.

That's the standard 13 move JB permutation that I was talking about.

 

[qwr]

Is there a collection of common cross patterns? I know cross is supposed to be intuitive but there must be certain hard cases that have good solutions.
Particularly tricky is if the edge is in the wrong orientation on the cross face because it has to be taken out and put back.
One trick I found just by looking at cross solutions is if there is a flipped edge in cross layer front and another bad edge in top layer right face to do something like F' R' F which turns both edges from bad edges to good edges and swaps their positions

 

[Joe Archibald]

If your looking to get better at cross just watch example solves from fast cubers and look at how they insert different pieces. You should make sure that your cross is under a maximum of 8 moves (normally closer to 6 or less).

 

[Mystheman]

Hello everyone,

When I checked the posts I saw a lot of titles that included questions like this one. Yet, I have some specific questions that I want to ask for becoming sub-12 on 3x3.

First of all, I have been solving cube for more than 5 years. For 2 years of the process, I hadn't solved the cube for the purpose of becoming sub12. Now, I want to enhance my times a little bit less. In order to do this, I searched lots of things but lastly I decided to create a topic that, I thought, could be helpful. So, I haven't completed all the OLL algorithms yet. My PLL is full. For the F2L part, I haven't completed algorithms neither. I just try to solve the F2L parts with the ways that I figured out. What should I do to become sub-12 or faster than that, what way I should be following? Thank you.