Tracking SpeedCubing Progress

[SpeedCubingLord]

Hey everyone,
I'm looking for inspiration and ideas on the most effective ways to track speedcubing progress. We all want to get faster, but knowing how we are improving (or where we're stuck) is crucial for targeted practice. Simply watching the overall average go down doesn't always tell the whole story.
I'm currently using a pretty basic method (csTimer), but I know many of you have much more sophisticated approaches. I've seen some mentions of detailed spreadsheets and specific software in other threads.
I want to open this up to the community:

• What specific metrics do you track besides your average time (e.g., F2L efficiency, number of pauses, specific algorithm recognition speed, move count)?
• What tools do you use (csTimer features, custom spreadsheets, specific apps, etc.)?
• Do you log every single solve, or just daily/weekly averages?
• Are there any unconventional methods for tracking improvement that you've found surprisingly effective?

My goal is to gather the best practices so we can all improve our training and analysis. Feel free to share screenshots of your tracking graphs or spreadsheet templates if you're comfortable!
Looking forward to seeing your ideas! Happy cubing.

 

[Caleb Valenzuela]

slow solves and recording yourself. Then note down the things that need improvement. Here’s how i do it:
first, do a couple solves to warm up
second, record some of my solves
third, prepare notes and study the solves i recorded in slow motion
fourth, based on the notes, find the most critical weakness and improve it. If i’m trying to improve efficiency, i’ll try to do untimed solves where I:
count moves
experiment with different solutions (like keyhole, pseudoslotting)
on the other hand if it’s recognition and execution that need improvement, i’ll use bestsiteever.ru or other websites that i find on the internet. cstimer also works well if i’m trying to find my splits, where i can also determine which step takes the longest.
One drill that i ALWAYS do is slow solves. Slow solves work well for improving lookahead, turning accuracy, efficiency, etc.

that’s all, but i usually just improved by watching tutorials like “how to be sub-x” on yt, and other times i just do solves and magically improve (i’m not joking it’s weird)

 

[That0newe1rdo]

CubicDB has a thing (cs timer has it to but CubicDB is much easier to do it with) where you can time individual steps.

All you do (on CubicDB) is make a new session and pick the option titled "Multi Step sessions"

I set it to 5 - 2x2x3, EOpair, LXS, OCLL, PLL.

Ofc how many you set it to will influence how many steps you do.

This will tell your average (in the statistics page) of each step, and you can decide on how fast you want each step to be, which will help in seeing what you have to improve on.

 

[qwr]

I know what my main weakness is: cross and f2l. However I have not taken the effort to learn f2l solutions and really practice cross.

For learning algs I tried digital flashcards but the only thing I found I actually use regularly are physical flashcards. For the algs I am pretty sure I have down I can put them aside in a done pile. My memory is average at best and what I found is that it's important to practice all of the new algs for at least a few consecutive days, otherwise they get forgotten easily. For language learning too, it's "use it or lose it".