How to Improve: The Mountain

[Noahaha]

This video is about an approach that I think can help most people improve. Tell me what you think.

 

[Phillip1847]

I was about to make a point about how practice improves capability somewhat, then I realized that solving to practice familiarity is just practicing the skill of F2L, which is a massive leap up.
In fact, I think it could be reasonable to think that it is a leap up to Feliks or so, but it takes a while to traverse this step. Which could make the mountain really a 3D mountain, with massive boulders to be found around the path up. Some raw boulders(F2L) asre too big to traverse by themselves, and some smaller boulders(IE, lookahead and near the top, multislotting) that can assist in climbing it.
To delve deeper into metaphor, just solving without learning is taking a long and slow, but not very steep way up the mountain, while deliberate practice combined with solving is a path that is moderately steep, but fast enough.

Anyway, cool video, I enjoyed it.

 

[A Leman]

I like this metaphor. To extend on it, mountains are quiet places and the people that can live on them learn how to think for themselves. I have been recently feeling like there is a growing city of people that don't even try to climb the mountain.

 

[jeff081692]

The first part of the mountain is easy. You are full of energy and are just climbing with little effort. Then you get tired and wonder why you are not climbing as fast as when you started. You keep trying to take a step and keep sliding back to your original position, instead of analyzing what is keeping you back and inventing a new technique for your climbing.

Great video, although I do wonder what the time ratio of deliberate practice to solves are for some of these guys that zip to the top. I know more time with deliberate practice lets you improve better in many other fields so I think something like 100 solves and another hour of deliberate practice would be enough but i'm still working out the perfect mix of solves/deliberate practice.

Also maybe some deliberate practice can be done during practice solves if you know what you are focusing on. Like for 3x3, doing normal solves but trying to see cross+1 each time is better deliberate practice than just seeing the cross in inspection and starting your solve.

 

[Noahaha]

Also maybe some deliberate practice can be done during practice solves if you know what you are focusing on. Like for 3x3, doing normal solves but trying to see cross+1 each time is better deliberate practice than just seeing the cross in inspection and starting your solve.

I definitely agree with this. There's no reason why they can't be combined.

 

[Stefan]

Nice concept and video, thanks. I really doubt I could ever be as fast as Feliks, though, now matter what/how I learn and practice. When Nakajima was top, I was able to replay his solves myself at his speed after practicing them a lot. I can't do that with Feliks. Might be an age issue, or maybe that's just an excuse

Btw, I cringed at every "deliberative", I believe it's "deliberate".

 

[Noahaha]

Nice concept and video, thanks. I really doubt I could ever be as fast as Feliks, though, now matter what/how I learn and practice. When Nakajima was top, I was able to replay his solves myself at his speed after practicing them a lot. I can't do that with Feliks. Might be an age issue, or maybe that's just an excuse

Btw, I cringed at every "deliberative", I believe it's "deliberate".

You're right. I've been saying it wrong all these years. Deliberative is a word though, so that has to count for something.

 

[Renslay]

Great video. I always disagreed with people who said "just practice". No, practice won't be enough on its own! For example, if you practice a wrong method or silly fingertricks, that wouldn't help you improve.

 

[antoineccantin]

Great video. I always disagreed with people who said "just practice". No, practice won't be enough on its own! For example, if you practice a wrong method or silly fingertricks, that wouldn't help you improve.

It depends how you define practice.

 

[Faz]

Nice video, some interesting ideas! I'm not sure if I completely agree with everything though

Great video. I always disagreed with people who said "just practice". No, practice won't be enough on its own! For example, if you practice a wrong method or silly fingertricks, that wouldn't help you improve.

My thinking is that when you're doing something over and over, you're naturally becoming more efficient at it. It's what Yu Nakajima did back in the day, just heaps and heaps of solves, and it's basically what I do/did too.

 

[Noahaha]

My thinking is that when you're doing something over and over, you're naturally becoming more efficient at it. It's what Yu Nakajima did back in the day, just heaps and heaps of solves, and it's basically what I do/did too.

If you are naturally becoming more efficient, then you are increasing your capability along the way. If that does happen naturally, then that's good, but I think that lots of people practice solves without really improving anything. I did fail to mention in the video that you can improve while doing lots of solves, as long as you are not just mindlessly solving.

Also hopefully you don't mind that I used your name so much.

 

[jeff081692]

I know for LL I do a lot of U3 and even U4 just for trying to recognize the PLL (and some OLL) and I have been cubing for a long time. One of the things that stops me from fixing this weakness is that if I try to fix it during normal solves I would be solving worse while I try to think about how I would better recognize the case. And so either solves without a timer or just setting up different LL possibilities and coming up with a better recognition system like 2 side for easy cases would be more beneficial than when I tend to ignore the weakness during normal solves. But once small things like that are taken care of I can see the benefit of doing lots of solves since they go by so quickly and it's all about maintaining a certain speed and focusing on turning technique rather than learning new things.

So I guess that is why slow solves are helpful since you don't have to feel pressured by the timer to get the fastest time when you know you have some bad cubing habits.

 

[Escher]

Nice video, some interesting ideas! I'm not sure if I completely agree with everything though

My thinking is that when you're doing something over and over, you're naturally becoming more efficient at it. It's what Yu Nakajima did back in the day, just heaps and heaps of solves, and it's basically what I do/did too.

I remember back in '09 we were racing to see who could get sub 11 avg 100 first, when we were both around 11.3-5 secs. While I was busy drilling PLLs and breakdowns, you did a huge number of solves and beat me to it, and then kept going There should be some posts deep in the accomplishments thread... As far as I remember the extent of your deliberate practise was just mindfulness while grinding huge number of solves, which is by far the most efficient way to go around doing it, assuming you have the capacity to stay somewhat conscious of your little comments and optimisations. It's what I ended up turning to to get from high 8 averages down to low 8, and although there is a breakpoint for most people whereby slower and more conscious methods (like drills) need to be employed to go further and get back to grinding, I think mindful grinding is the best way to get things going quickly. It's kind of like a cut and bulk phase in weightlifting/body building.

@OP, I thought by mountain you would be talking about the 'asymptote' generated by the graph where x = improvement and y = time invested, but I suppose I like where you've gone with your video.

I quite like the capability/skill distinction, but I think you presented it as little more binary than it really is. Envisage turning your graph by 45 degrees anticlockwise, and consider how whenever you dedicate time specifically towards capability or skill, you're actually pulling yourself laterally away from the goal. It would make sense, then, why those who have improved in a way one can consider as 'optimal' are people like Nakajima, Feliks, and Alex Lau - those who improved by grinding a huge amount while keeping a steady mindset of improvement, rarely pulling themselves in any direction besides the goal. Like I was saying before though, I'm not sure everyone can keep this up.

Personally speaking I know I put FAR too much focus when practising a hobby in one or the other. With guitar, I never put much thought into capability, and alas never got as far as I could have - I simply got good at what I knew, so my mechanical skill was always high. In LoL, I typically spend way, way too much time reading and thinking about strategy, and only recently when I've been grinding have I seen my mechanical ability improve tenfold. I think cubing was the only hobby I've had so far where I found a balance and managed to maintain it - but nowadays I get much more enjoyment out of improving capability than becoming competitive, so rarely do I speedsolve any more. It's such a hard thing to balance, when you're just following your passion!

Good job anyway, it's always great to see stuff like this. I genuinely think the science of improvement is a vast and deep world that can draw on so many different areas for inspiration and development, and along with behavioural economics, is going to become a vitally important field in modern economies in the next 30 years. I love it!

 

[RageCuber]

This is very inspiring in a way. It gave me new insights
into cubing and I feel that this is going to help a lot of
people (Including myself ). Recently I was kind of on
a plateau, but then I noticed my PLL execution was pretty
bad, so I went to badmephisto's site downloaded the PLL
trainer and nailed all my PLLs. Now I am getting times that
I consider to be good much more often. I can't speak for
any one else, but this video was awesome! Keep at it!

 

[mark49152]

Great video. This is how I practice.

One thought on the comments about just grinding out solves: it could be correlation not causation. Perhaps the kind of people with the talent to get that fast are the kind of people who don't need targeted practice to get past their plateaus.

 

[pipkiksass]

Great video. This is how I practice.

One thought on the comments about just grinding out solves: it could be correlation not causation. Perhaps the kind of people with the talent to get that fast are the kind of people who don't need targeted practice to get past their plateaus.

Or perhaps they are the kind of people who can work towards targets while grinding out solves? I think everyone needs targeted practice, but different folks need different levels of conscious effort in order to include targets in their practice. I, for example, need to 100% concentrate on doing something in particular, like planning my cross (which is shocking) in order to work on it. Faz, on the other hand, didn't have to worry about it that much, because he achieved the same thing with less conscious effort.

But yeah, great video Noah - I can see myself smacking my head against the rise of the next step, wondering why I'm not climbing up the mountain... I've known about it a long time, but I suppose it's time to stop blindly solving and start working on some stuff!

 

[mark49152]

@Pip: by "targeted" I mean the kind of practice Noah called "deliberative" (which seems a misnomer). Not just having targets. Sorry for the confusion!

 

[pipkiksass]

@Pip: by "targeted" I mean the kind of practice Noah called "deliberative" (which seems a misnomer). Not just having targets. Sorry for the confusion!

No, I know, as did I. I think there's a varying level of deliberateness required between cubers. So while I have to turn off the timer, sit down and say "now I will practice cross", others might just turn off the timer and do slow solves, and achieve the same ends with less deliberateness/less specific goals/targets.

 

[cmhardw]

I quite like the capability/skill distinction, but I think you presented it as little more binary than it really is. Envisage turning your graph by 45 degrees anticlockwise, and consider how whenever you dedicate time specifically towards capability or skill, you're actually pulling yourself laterally away from the goal. It would make sense, then, why those who have improved in a way one can consider as 'optimal' are people like Nakajima, Feliks, and Alex Lau - those who improved by grinding a huge amount while keeping a steady mindset of improvement, rarely pulling themselves in any direction besides the goal. Like I was saying before though, I'm not sure everyone can keep this up.

Noah's method, as presented, can be a way for us regular folk to emulate the fantastic improvement seen by the greats. I like your way of viewing the graph, Rowan. I will use Noah's approach when practicing, but I will view the graph the way Rowan mentions.

Great video Noah! It gives focus to us who are walking to the right into the same step for so long and aren't sure exactly why we aren't moving up our mountains.

 

[MM99]

The first part of the mountain is easy. You are full of energy and are just climbing with little effort. Then you get tired and wonder why you are not climbing as fast as when you started. You keep trying to take a step and keep sliding back to your original position, instead of analyzing what is keeping you back and inventing a new technique for your climbing.

I really did change my way of climbing by switching to zz but I'm stuck at the 20 second barrier just like I was with cfop... oh well I think the deliberate practice I should be doing is bld eo line that should keep me climbing the mountain