Strange Splits (Sub-19 CFOP)

[dodecicosidodecahedron]

Hello,

I have been cubing since late 2017 but I only learnt CFOP in April 2019. I currently average around 18-19 seconds with the method, and I seemed to have plateaued from improving once I hit the 22 second mark. I thought this was strange, and maybe it was something wrong with my practice, so I went ahead and learnt full OLL (already knew full PLL), and can now confidently do 2LLL. This didn't help much, and my times were still not improving.

I decided that I would record my solves to see them from an external perspective, as during solves I can usually justify why I pause as I know what I'm thinking, but through a video that would not be the case. After recording a dozen or so solves and noting the splits for each step, I found that on average, this is what I tend to get (including recognition):

Cross: 2.50
F2L: 10.49
OLL: 2.43
PLL: 3.21

This was weird, as I had the OLL/PLL of someone who averaged sub-15 (I'm sub-19), but I had the F2L of someone who averaged 20+ seconds. Considering that F2L takes up the majority of the solve, it's no surprise as to why I am not improving whatsoever.

I tried doing some F2L only solves, but I don't see much improvement with it. I also tried to solve 2 pairs blindfolded at a time, but I was pretty much unable to do it, as I am incapable of tracking pieces. In the future, I wanted to be sub-10 (pretty much my dream when I got into cubing), but I'm not sure if it will be possible from a realistic standpoint. I don't want this "anything is possible" BS because I know that everyone has their skill ceiling, and I think that with my extremely poor and virtually unimprovable pair tracking, I will never reach this goal.

"Slow down and look ahead" is another generic piece of advice that means nothing without context either, and I hear all of these a lot. Considering all of this, I have 2 not-so-simple questions to ask.

1. Where do I start with pair tracking and how to get good at it?
2. Will I ever be able be sub 10 considering my current splits and improvement rate?

Thank you for reading. Any advice is much appreciated.

 

[xyzzy]

This is potentially very bad advice, so take this with a grain of salt, but what I did to get from 20 seconds (late 2016?) to 14 seconds (now) was to just spam tens of thousands of solves. (I'm also not "sub-10" and worse than plateauing, it seems my average has regressed…)

It seems you already know the "theory" on how to improve, so maybe what follows is completely redundant to you. For tracking F2L pieces, you should generally try to remember which F2L corners/edges are stuck in the wrong slot (try to "get a feel" for this while solving the cross), as those pieces will usually stay put while you solve unrelated pairs/slots. Keep track of the top-facing (or side-facing, choose one) colour of F2L edges you see in the top layer; if you mostly use 2-gen RU/LU F2L algs, edges in the top layer won't flip and you can make use of that fact to predict whether you need to rotate for the next pair. If you already know what the edge for your next F2L pair is, and you know its orientation, all you need to do is look for its matching corner: finding one piece is faster than finding two!

I don't want this "anything is possible" BS because I know that everyone has their skill ceiling

2. Will I ever be able be sub 10 considering my current splits and improvement rate?

I don't believe anything is possible. Growth mindset is a bad meme, educators probably know it, but most of them also won't admit it because it undermines their value proposition. If you just look at how many cubers aren't sub-10, it's clear that not everyone has the potential for it.

But are you among those never-sub-10 cubers? It seems too soon to jump to conclusions, considering that you've been using CFOP proper for only half a year. Keep at it for a few more months, and maybe your improvement (or, touch wood, lack thereof) will surprise you.

 

[PetrusQuber]

I can’t help much, but what I will say is I think you’re being too hard on yourself. I’ve been cubing for nearly the same amount of time as you, and I’m averaging 25. Some people might never manage sub 30, even with more time. Plus, I don’t believe anything is possible, but I’m sure sub 10 is realistic for you. At your stage, you can always improve, and practise will make your F2L better, no matter how long it takes. Feliks is sub 10, and he uses the same method as you, albeit with X Cross and some ZBLL, so you know it is possible. Genetic advantages are something I believe does not exist, or at least to significant extent. It takes time and practise. Be patient . Again, Feliks been cubing for over 10 years, and is only sub 7.

 

[dodecicosidodecahedron]

This is potentially very bad advice, so take this with a grain of salt, but what I did to get from 20 seconds (late 2016?) to 14 seconds (now) was to just spam tens of thousands of solves. (I'm also not "sub-10" and worse than plateauing, it seems my average has regressed…)

It seems you already know the "theory" on how to improve, so maybe what follows is completely redundant to you.

The thing is that although I know the theory on how to improve, I don't know how to fix my F2L efficiently, and I'm scared that spamming thousands of solves will only reinforce my **** habits.

I don’t believe anything is possible, but I’m sure sub 10 is realistic for you. At your stage, you can always improve, and practise will make your F2L better, no matter how long it takes.

Thanks for the uplifting sentence, but I don't think I'll ever make my look ahead sub-10-worthy, unless I completely change the way I view F2L as a whole.

 

[Cubinwitdapizza]

I think you should look at every tutorial possible on f2l and then go from there because every other step is really good. My splits were very similar when I used cfop except for my cross was a bit slower and my f2l was much faster.

 

[Cuberstache]

I also tried to solve 2 pairs blindfolded at a time, but I was pretty much unable to do it, as I am incapable of tracking pieces.

Instead of trying to solve 2 pairs blindfolded, try solving only one. To look ahead well, you should be able to solve one pair blindfolded 100% of the time with complete confidence that it will be solved. That way, you can track your next pair with your eyes while your hands solve the pair without requiring conscious thought. In fact, I tried to solve two pairs blindfolded a few times because you mentioned it. As a sub-10 solver, I can only barely solve two pairs blindfolded. I mess up a lot, I have to look at the cube for a solid 3 seconds before starting, and I have to turn really slowly to track the pieces in my head. Solving two pairs blindfolded is simply not a realistic goal and not how you should be approaching lookahead.

 

[Llewelys]

The thing is that although I know the theory on how to improve, I don't know how to fix my F2L efficiently, and I'm scared that spamming thousands of solves will only reinforce my **** habits.

• Most F2L cases can be solved in 8 moves or less, except for some that take 11 moves. If you're doing more than 11 moves to solve a case you know you're doing something wrong.
• Look for different ways to solve a pair depending on whether you're on last slot or not: most cases are more move efficient when you can take advantage of empty slots.
• If there's a solved corner or edge, use keyhole as soon as possible (D R U R' D' vs U2 L' U L U2 L' U L)

Those are basic advice but can help with F2L efficiency

The most useful advice would be to watch lots of example solves and solve critiques