[Help Thread] Been stuck around the 15 second mark for 4 Years (3x3)

[White KB]

Hoi goys,
I'm Keenan Bruce. I've been cubing since 2017. It took me just under 4 years to get sub-20 (strict parents = no practice time, go figure), then another 8 months to get sub-15. I know full OLL, full PLL, and some advanced stuff like pseudoslotting but not very well. My LSLL takes me about 6 seconds and the rest of the solve takes me about 8-10 seconds. I haven't really cared about 3x3 for the last 4 years because I feel like I hit a wall somewhere around 17k solves. And then college for the last 3ish years has made it so that I don't have that much practice time. I find it hard to plan cross/cross+1 in inspection, but I am dual CN on white/yellow. My lookahead is ok ish. I want to eventually get sub-10. What are some tips based on this info for how I would get sub-10? I feel like I could get there without more algs but if I have to learn new ones I guess I could.

 

[cubemastery321]

Hey Keenan — you're right, you don't need a bunch of new algs to get to sub-10. Your own numbers point to where the time is: about 6 seconds for LSLL, and 8-10 seconds for the rest. So the biggest win is making that last slot + last layer smoother, not learning more cases.

A few things I'd focus on, especially with limited practice time in college:

1. Cross+1, but in stages. Don't try to plan the whole thing at once yet. Use inspection to plan the cross fully, then just spot the first pair and know where it is. Do that for a week or two of untimed solves, then start trying to track that pair while you do the cross. That one habit alone usually cleans up the pause after the cross.

2. Last slot. A lot of that 6 seconds is probably pausing to recognize OLL/PLL and the last pair, not the algs themselves. Try slow solves where you force yourself to look at the last pair while you're doing the second-to-last one, and drill OLL/PLL recognition from different angles — not just the alg, but spotting the case fast.

3. Keep F2L efficient, not fancy. Pseudoslotting is fun, but I'd park it until your normal F2L is pause-free. In slow solves, check if you're doing extra rotations or long insertions — cutting one rotation per pair adds up fast.

A short routine that fits a busy schedule: 5 slow, lookahead-focused solves, then 10 timed solves, and stop there. Consistency beats long sessions you can't keep up.

Sub-10 from 15 is very doable with full OLL/PLL already learned — it's mostly lookahead and last-layer recognition from here.

 

[Puzzle Maniac]

Hey Keenan — you're right, you don't need a bunch of new algs to get to sub-10. Your own numbers point to where the time is: about 6 seconds for LSLL, and 8-10 seconds for the rest. So the biggest win is making that last slot + last layer smoother, not learning more cases.

A few things I'd focus on, especially with limited practice time in college:

1. Cross+1, but in stages. Don't try to plan the whole thing at once yet. Use inspection to plan the cross fully, then just spot the first pair and know where it is. Do that for a week or two of untimed solves, then start trying to track that pair while you do the cross. That one habit alone usually cleans up the pause after the cross.

2. Last slot. A lot of that 6 seconds is probably pausing to recognize OLL/PLL and the last pair, not the algs themselves. Try slow solves where you force yourself to look at the last pair while you're doing the second-to-last one, and drill OLL/PLL recognition from different angles — not just the alg, but spotting the case fast.

3. Keep F2L efficient, not fancy. Pseudoslotting is fun, but I'd park it until your normal F2L is pause-free. In slow solves, check if you're doing extra rotations or long insertions — cutting one rotation per pair adds up fast.

A short routine that fits a busy schedule: 5 slow, lookahead-focused solves, then 10 timed solves, and stop there. Consistency beats long sessions you can't keep up.

Sub-10 from 15 is very doable with full OLL/PLL already learned — it's mostly lookahead and last-layer recognition from here.

Got it—your message perfectly highlights how it is not about learning new algorithms—it's about a commitment to lookahead and fostering robust recognition.

 

[gsingh]

Hoi goys,
I'm Keenan Bruce. I've been cubing since 2017. It took me just under 4 years to get sub-20 (strict parents = no practice time, go figure), then another 8 months to get sub-15. I know full OLL, full PLL, and some advanced stuff like pseudoslotting but not very well. My LSLL takes me about 6 seconds and the rest of the solve takes me about 8-10 seconds. I haven't really cared about 3x3 for the last 4 years because I feel like I hit a wall somewhere around 17k solves. And then college for the last 3ish years has made it so that I don't have that much practice time. I find it hard to plan cross/cross+1 in inspection, but I am dual CN on white/yellow. My lookahead is ok ish. I want to eventually get sub-10. What are some tips based on this info for how I would get sub-10? I feel like I could get there without more algs but if I have to learn new ones I guess I could.

Hey, long time no see

The the most common flaw that I've seen solvers have (even at the ~10s level) is bad F2L solutions. If you ask anyone on reddit, SS, or discord how you should improve your F2L, you will almost certainly be told "just learn intuitive F2L bro", but this is very misleading. While it's true that most F2L solutions for the standard 41 cases seem simple in theory, if you're taught to just use the "intuitive" F2L that's taught on youtube, you will have very poor solutions for a lot of cases.
I would advise that you watch a youtube video that goes over every standard F2L case (this one is my favorite, Lazer0Monkey has a good one as well), and learn the solutions for whichever cases that you're doing inefficiently. You should be able to do all of these from the right and the left side, and you need to be able to do them all from the back as well.
Your LSLL should also not be taking 6s; make sure you can recognize PLL from all angles, and make sure you can predict AUF as well. If you can do that, then it's worth doing a couple hundred reps of each OLL/PLL alg, with proper fingertricks (you can find these on youtube, I think Cubehead has a nice video).