CFOP Solve Splits Tool

1 month ago

Speedcubing Advice

One question I’m asked very frequently goes something like: “Hey, I average X seconds, what do you think my time splits for Cross-F2L-OLL-PLL should be at this stage?”.

 

Firstly, it depends a decent amount on your current knowledge base - consider solvers A and B, who both average 20 seconds:

 

Solver A still inserts their cross pieces one at a time into the top layer and has extremely inefficient F2L, but knows full OLL and PLL and can execute these algorithms well. Realistically, the splits for this solver could look like 4-11-2-3.

 

Solver B has a much more move efficient F2L and has quite good lookahead ability, but still uses 4 look last layer. The splits for this solver could look something like 2-9-4-5, which is a vastly different profile to solver A.

 

Without being too pedantic, a better alternative question to “What should my splits be?” is “Based on my current splits of A-B-C-D and the current techniques and algorithms that I have learned, which step(s) of my solve can be improved most easily, or what should I learn next?”

 

This is a much simpler question to answer, and going to the effort to collect the data is the bulk of the work! (CSTimer.net has a nice way to do this - option - timer, multi-phase). Of course, most people probably don’t have a glaringly obvious phase of their solve which needs improvement, but it could be a useful exercise nonetheless.

 

To assist with this, I’ve created a quick tool which allows you to enter your current CFOP solve splits and then compares them to average solve split times which I have put together - partially based on sub 10 solver data gathered in this speedsolving thread

 

This tool won’t provide any magic solutions to improve those stages (that’s what the tutorials are for!), but may help you to identify a part of your solve which can be improved more easily than others. At that point, targeted practice on that stage is something to consider, if you think that your knowledge base is sufficient. For example, if you’re sub 20, know full PLL, but it is taking up 5 seconds in your solve, then deliberately practicing either your recognition or execution of those algorithms will be beneficial.

 

The sheet works for CFOP solvers between 60 and 8 seconds. Beyond sub 8, standard CFOP splits can become a bit difficult to measure, as solvers of that speed tend to frequently use things such as extended crosses, and various OLS techniques, so it’s not particularly relevant.

 

Let me know if you have any questions in the comments section. This short post is thanks to a member of the website who suggested that I cover this topic, thanks Omkar!

 

The sheet is available here :)

 

CS Splits Tool


Thoughts on this blog

Notify me of updates
Matthew  Kleineberg

Matthew Kleineberg Posted 1 month ago

Why do we need permission to get on the sheet


Feliks Zemdegs

Feliks Zemdegs from CubeSkills Posted 1 month ago

Hey Matthew - I've edited the permissions, can you let me know if clicking the link in this blog works now? Cheers :)


nope nope

nope nope Posted 1 month ago

It works now, thank you!


Master Jamin Hendricks

Master Jamin Hendricks Posted 1 month ago

How to edit?


Edward Connell

Edward Connell Posted 1 month ago

Cool - I've always just used the 12:50:16.5:21.5 splits from the article directly, even though I'm well above 10s.  It's interesting to see what you think the splits could be for solvers at different points in their learning.

My cross is at -24% and F2L at +11% - which probably means I should be working on cross+1.


Milan Struyf

Milan Struyf Posted 1 month ago

Cool idea ;)


Matthew  Kleineberg

Matthew Kleineberg Posted 1 month ago

It works now, thanks.


Master Jamin Hendricks

Master Jamin Hendricks Posted 1 month ago

I still can't edit.


Feliks Zemdegs

Feliks Zemdegs from CubeSkills Posted 1 month ago

Hey Jamin -  the link will prompt you to make a copy of the spreadsheet into your personal google drive, at which point you'll be able to edit the sheet.


Ethan Varn

Ethan Varn Posted 1 month ago

This is an amazing tool and also idea, but after an ao50 with splits I’m -2% +3% 0% and -8% respectively so not super sure where to go from here.


Ethan Varn

Ethan Varn Posted 1 month ago

For 11.62


Feliks Zemdegs

Feliks Zemdegs from CubeSkills Posted 1 month ago

That means your splits are pretty well-rounded! :)


Noah Marsh

Noah Marsh Posted 1 month ago

This is super helpful! I wasn't sure if I needed to learn all the OLLs, and now I see that I'm average for my times. Thank you!


Kerry Go

Kerry Go Posted 1 month ago

Thanks! This will be a useful tool. I've wondered about what my splits should be, too, though I never thought about the splits varying based on uneven development. My splits are very inconsistent based on scramble, so I will probably have to get an average of many. For example, a PLL could vary between (aside the obvious case of a PLL skip) getting an case I know, such as a Jb perm, which might take me 4 seconds, or getting a case I don't know, such as a G perm, which might take me 8 seconds, or getting a case that I think I know, but I don't, so I have redo F2L and it takes 30 seconds.


Master Jamin Hendricks

Master Jamin Hendricks Posted 1 month ago

I need to work on cross. lol.


Master Jamin Hendricks

Master Jamin Hendricks Posted 1 month ago

Can you perhaps make one for 4x4?


Sofia Nguyen

Sofia Nguyen Posted 1 month ago

Thank Omkar for the great request and Feliks for such a useful tool. I know little about hardwares so I gotta ask does Gan i have this sorta tool built-in? Coz if it doesn't, well, ;) ;) ;). If it already had the tool, thank you for making it affordable for everyone.


Tan E-Jie

Tan E-Jie Posted 1 month ago

What if I do an XCross?


Tan E-Jie

Tan E-Jie Posted 1 month ago

And does this work for OH?

 

 

 


Noah Marsh

Noah Marsh Posted 1 month ago

If it's an x-cross count it as cross, because your F2L time will be shorter. And yeah, I think it works for Oh


Tan E-Jie

Tan E-Jie Posted 1 month ago

I find that my splits are different for oh though... my f2l is a lot shorter and my ll a lot longer

 

 


Matt Gossett

Matt Gossett Posted 1 month ago

Great idea!


Feliks Zemdegs

Feliks Zemdegs from CubeSkills Posted 1 month ago

Kerry - certainly, this is meant for splits across many solves, best to take an average.

Jamin - Will put that on my list as a future blog idea! :)

Tan - Yeah put it as cross I suppose, but ideally this tool is for those not yet using X-crosses very often in their solves. I'm not sure this is great for OH, typically LL is a higher percentage of OH solves, I'll have to add that to the list for a future blog/tool.


Tan E-Jie

Tan E-Jie Posted 1 month ago

Thanks! By the way, my first name is E-Jie, not Tan

 


Tanner Davey

Tanner Davey Posted 1 month ago

That tool is really cool.

 


Tanner Davey

Tanner Davey Posted 1 month ago

I need to work on cross and oll a lot.


Cal J

Cal J Posted 1 month ago

How do you find your averages???


Tanner Davey

Tanner Davey Posted 1 month ago

It is next to the gold cubes


Tanner Davey

Tanner Davey Posted 1 month ago

On one of them my cross was 90%

 


Sizzles Stein

Sizzles Stein Posted 1 month ago

Hey Feliks can you make a video on how you prefer doing every f2l case.


Matthew Knippen

Matthew Knippen Posted 3 weeks ago

Loved this, thanks for sharing! Putting this together with my GAN 356i was awesome. I averaged my past 10 solves together, and showed that I definitely need to work on my cross.


Kaeden Bak

Kaeden Bak Posted 4 days ago

Meh... I average around 40 on 3x3 PB Is a 19


Kaeden Bak

Kaeden Bak Posted 4 days ago

Also how do learn full oll

 



Become a free member to post a comment about this blog.