• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 40,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

How to practice

Thecuberlife

Member
Joined
Jun 25, 2015
Messages
9
As of right now I am a 25-60 second solver typically 25-40 50-60 being very rare but occasional. anyway I was wondering from a more experienced solver what should I be practicing and how. Now I’m not asking what to like learn pll I mean actually what do I do so like do blind solves do blah blah blah I don’t know anything else honestly.

Things of help might be whenever I finish cross and f2l it is very often when I look up I’m at 29 seconds so I figure 5-9 secs for cross 20-25 on f2l and my last layer is actually pretty quick (except for e perm which I’m getting faster at but it’s pretty pathetic )I’d say last layer can be from 5-7 or 8. Also something I struggle with that honestly I thing could mess me up by as much a 2-7 seconds is my fumbling. I’ll be doing f2l and looking ahead knowing my algs and then it’s like I lose my grip and then I have to look again cause I forgot where I was and it’s really just annoying. Suggestions here?

So what should I be doing daily I don’t know why but for some reason I can’t word any of this right but like I mean something along these lines

10 normal solves
10 cross solves
F2l slowly for blah

I don’t know if this makes any sense so sorry in advance and also thanks if you read through my horribly explained post.
 
Joined
Aug 16, 2014
Messages
2,987
Location
Webster Groves, MO
WCA
2013BARK01
Honestly my practice routine has for the most part been "do hella solves" and I guess it worked because I'm almost sub 9 now. Really as long as you're practicing and actively trying to eliminate bad habits, you'll get faster. Basically do a bit of trial and error, and when there's error, don't do that again.
 

pjk

Administrator
Staff member
Joined
Mar 13, 2006
Messages
6,686
WCA
2007KELL02
SS Competition Results
Considering this thread was first created 7 years ago, and a lot of practice has happened, and techniques and hardware have been refined, how has practicing changed from when this was first written? How do the fastest solvers practice now vs. what was written here 7 years ago?
 

Escher

Babby
Joined
Jul 23, 2008
Messages
3,374
WCA
2008KINN01
YouTube
Visit Channel
7 years... Wow! It's really heartening for me to see that people still put a 'Like' on this post every now and then, despite so many cubers predominantly using Reddit/YT/Discord nowadays.

Currently I only do a few solves here and there for maintenance. Were I to start practicing properly:

  • Copying the crap out of the top 10. Practically every solve of theirs in comp is recorded and reconstructed, and that's true even for many of the YT at-home vids.
  • Heavy use of Anki for learning alg sets, regularly revisiting deck contents and reviewing stats; and 'lookahead quizzes' for LS -> LL transition.
  • Recording all my solves with a camera, and tagging solves for review (f2l decisions, execution fumbles).
  • Looking more at your hands than at the cube during slow-solves.
  • Finally going ahead and doing in-depth research on some of those old pair-selection ideas with a program or two.
Btw, nice work on the site @pjk (& co.?)
 

pjk

Administrator
Staff member
Joined
Mar 13, 2006
Messages
6,686
WCA
2007KELL02
SS Competition Results
7 years... Wow! It's really heartening for me to see that people still put a 'Like' on this post every now and then, despite so many cubers predominantly using Reddit/YT/Discord nowadays.

Currently I only do a few solves here and there for maintenance. Were I to start practicing properly:

  • Copying the crap out of the top 10. Practically every solve of theirs in comp is recorded and reconstructed, and that's true even for many of the YT at-home vids.
  • Heavy use of Anki for learning alg sets, regularly revisiting deck contents and reviewing stats; and 'lookahead quizzes' for LS -> LL transition.
  • Recording all my solves with a camera, and tagging solves for review (f2l decisions, execution fumbles).
  • Looking more at your hands than at the cube during slow-solves.
  • Finally going ahead and doing in-depth research on some of those old pair-selection ideas with a program or two.
Btw, nice work on the site @pjk (& co.?)
The community has become quite divided across Reddit/YT/FB/Discord, but many, including myself, find it quite hard to have long form constructive discussion there, such as this one. It's a difficult nut to crack.

Reconstructions: Good tip and a great way to find new tricks or techniques to solve different positions.
Anki: I've used Anki extensively for learning languages, and images, but never for learning algs. Can you share some of your decks here? I'd be curious.
Looking for at hands: What are you looking for exactly here?

Great to hear from you again. There isn't a lot of new interesting discussion here these days but would be great to have more of it (but that requires people to participate in interesting discussions ;) ).

Does anyone know how the fastest solvers these days are practicing? Are there any specific techniques or methods of practice? Would be interested to hear if it has changed much over the last few years.
 

Escher

Babby
Joined
Jul 23, 2008
Messages
3,374
WCA
2008KINN01
YouTube
Visit Channel
The community has become quite divided across Reddit/YT/FB/Discord, but many, including myself, find it quite hard to have long form constructive discussion there, such as this one. It's a difficult nut to crack.

Reconstructions: Good tip and a great way to find new tricks or techniques to solve different positions.
Anki: I've used Anki extensively for learning languages, and images, but never for learning algs. Can you share some of your decks here? I'd be curious.
Looking for at hands: What are you looking for exactly here?

Great to hear from you again. There isn't a lot of new interesting discussion here these days but would be great to have more of it (but that requires people to participate in interesting discussions ;) ).

Yeah, it's a real shame to me that 'churn' is built so inseparably into Reddit, it's a black hole for content, at least from an exploration pov... People do need to go back to forums (including myself)! I've been reading something called 'Inadequate Equilibria' that's kinda related.

I don't actually have any Anki decks for alg learning: that's what I would use though. I do have one for the Speffz lettering system but as always I'm way too lazy to learn BLD or review those cards. I'm mainly just using it for Python and CS stuff. On that note, it would be really nice to automatically generate anki cards/whole decks and respective images from an alg list.

Looking at hands: I just mean paying attention to the really little things, like where you actually place your gripping fingers on the cube itself. Like I often have my LH-middle on BL and LH-ring on BD, should I get used to the 'feel' of being directly on the stickers and try to make sure I have that stance all the time?

That particular example might be over-engineering a little bit though... Not that I've ever been criminal of that xD
 

pjk

Administrator
Staff member
Joined
Mar 13, 2006
Messages
6,686
WCA
2007KELL02
SS Competition Results
Yeah, it's a real shame to me that 'churn' is built so inseparably into Reddit, it's a black hole for content, at least from an exploration pov... People do need to go back to forums (including myself)! I've been reading something called 'Inadequate Equilibria' that's kinda related.
I've had this discussion with many people, more so about Facebook. Facebook is a great tool, and FB groups can be very useful. But the problem is that content isn't linked (referencing old conversations), and for super active groups, once a thread is made, it disappears within 1-2 days, rarely to be read again. I'm part of a language learning group where people post great topics and spend a lot of time making great lessons. However, after 2 days they are gone. Then a week later someone asks a question which was answered 100 times already, and no one links and references old threads. I built out a tool that enabled Group admins to port their group over to a self hosted forum such that oldest topics could be linked, searched, and discussed further in long form. This helped solve that problem for that language group. The success of FB groups is due to people already being on there and great activity. A community only works when people participate. If people don't participate, nothing interesting is discussed and therefore people leave. The vibe here on the SS forums has changed as people left since not many good conversations were being had. It's a cycle downward, or upward when people participate. It's a tough problem to solve. I do think there are many people like us that much prefer long form discussion on a forum like this over Reddit/FB, but it's likely not the newest generation.

Inadequate Equilibria looks great, how far into are you? Learn anything interesting yet? I'll add it to my to-read list.

I don't actually have any Anki decks for alg learning: that's what I would use though. I do have one for the Speffz lettering system but as always I'm way too lazy to learn BLD or review those cards. I'm mainly just using it for Python and CS stuff. On that note, it would be really nice to automatically generate anki cards/whole decks and respective images from an alg list.
Absolutely, I wonder how many people in the community have used or are using Anki. It's a great tool. As a community maybe we can build out decks for different alg sets to help people drill them. That would be super useful for people wanting to learn new algs with the spaced repetition system (which is proven to be incredible effective for long term memory). I'd use an Anki deck for CMLLs right now if I had one. Great idea.

Looking at hands: I just mean paying attention to the really little things, like where you actually place your gripping fingers on the cube itself. Like I often have my LH-middle on BL and LH-ring on BD, should I get used to the 'feel' of being directly on the stickers and try to make sure I have that stance all the time?

That particular example might be over-engineering a little bit though... Not that I've ever been criminal of that xD
Haha, great idea. I'm guessing that sort of deliberate practice would make a difference. I'd be curious to hear from others on their strategies, if they use any at all. How do you go from 7 sec avg to 6 sec, for example? I suppose to a certain point, just straight practice will improve your times. Then you'd have to deliberately practice the weakness to break off the plateau.
 

Escher

Babby
Joined
Jul 23, 2008
Messages
3,374
WCA
2008KINN01
YouTube
Visit Channel
I've had this discussion with many people, more so about Facebook. Facebook is a great tool, and FB groups can be very useful. But the problem is that content isn't linked (referencing old conversations), and for super active groups, once a thread is made, it disappears within 1-2 days, rarely to be read again. I'm part of a language learning group where people post great topics and spend a lot of time making great lessons. However, after 2 days they are gone. Then a week later someone asks a question which was answered 100 times already, and no one links and references old threads. I built out a tool that enabled Group admins to port their group over to a self hosted forum such that oldest topics could be linked, searched, and discussed further in long form. This helped solve that problem for that language group. The success of FB groups is due to people already being on there and great activity. A community only works when people participate. If people don't participate, nothing interesting is discussed and therefore people leave. The vibe here on the SS forums has changed as people left since not many good conversations were being had. It's a cycle downward, or upward when people participate. It's a tough problem to solve. I do think there are many people like us that much prefer long form discussion on a forum like this over Reddit/FB, but it's likely not the newest generation.

Inadequate Equilibria looks great, how far into are you? Learn anything interesting yet? I'll add it to my to-read list.

Absolutely, I wonder how many people in the community have used or are using Anki. It's a great tool. As a community maybe we can build out decks for different alg sets to help people drill them. That would be super useful for people wanting to learn new algs with the spaced repetition system (which is proven to be incredible effective for long term memory). I'd use an Anki deck for CMLLs right now if I had one. Great idea.

Haha, great idea. I'm guessing that sort of deliberate practice would make a difference. I'd be curious to hear from others on their strategies, if they use any at all. How do you go from 7 sec avg to 6 sec, for example? I suppose to a certain point, just straight practice will improve your times. Then you'd have to deliberately practice the weakness to break off the plateau.

Yeah, I think it'll take a sea change in the younger generation to revitalise forums for all of us.

I'm about halfway through Inadequate Equilibria. I'm not an economist or an AI researcher but I think the tl;dr is: markets in which every actor has collectively perfect information, but individually imperfect ability to act, will eventually reach a nash equilibrium or 'local maximum', and not necessarily the best one. And it takes complete cooperation to escape it, not necessarily always in the best direction for some actors (or even every actor). Sometimes we can't fix things without completely breaking things first. There's a roughly similar thing in AI where you do what seems in principle to be correct but your model gets stuck and can't 'see' the global maximum. Eliezer is really good at beginning with very precisely and technically stated topics or principles and ending with the big, persistent problems of the human condition!

I'll think more on that Anki topic! Especially making things practically easier. From personal experience it's actually pretty good for quick recall.

I wouldn't know, but I reckon the absolute fastest people have discovered something which works well for them, and sheer repetition alongside the occasional conscious tweaks is enough. I suspect hardware has always been a limiting factor and is more responsible for the biggest changes in times compared to any revelations about the cognitive process. Not to say that good old deliberate practice won't be the most effective way of improving for the rest of us ;)
 

pjk

Administrator
Staff member
Joined
Mar 13, 2006
Messages
6,686
WCA
2007KELL02
SS Competition Results
Yeah, I think it'll take a sea change in the younger generation to revitalise forums for all of us.

I'm about halfway through Inadequate Equilibria. I'm not an economist or an AI researcher but I think the tl;dr is: markets in which every actor has collectively perfect information, but individually imperfect ability to act, will eventually reach a nash equilibrium or 'local maximum', and not necessarily the best one. And it takes complete cooperation to escape it, not necessarily always in the best direction for some actors (or even every actor). Sometimes we can't fix things without completely breaking things first. There's a roughly similar thing in AI where you do what seems in principle to be correct but your model gets stuck and can't 'see' the global maximum. Eliezer is really good at beginning with very precisely and technically stated topics or principles and ending with the big, persistent problems of the human condition!
I was at the book store a couple weeks back and came across "Oxford Shorts", which are shorter books on various topics which introduce you to a wide variety of topics (there one ones on wars, depressions, psychology, religions, etc.). I figured if I can read one a week or so, I'd get a decent overview of various topics. The first one I started was "Intro to Game Theory". It's funny you mention the above, as they intro Nash Equilibrium and various other intro game theory scenarios, similar to what you described above. As a result, I downloaded the movie "A Beautiful Mind" and have started watching the movie on Nash. Interesting guy. I will have to read the Inadequate Equilibria soon.

I'll think more on that Anki topic! Especially making things practically easier. From personal experience it's actually pretty good for quick recall.
Indeed. We should start a thread to see if others in the community would like to help build Anki decks. They'd be super useful.

I wouldn't know, but I reckon the absolute fastest people have discovered something which works well for them, and sheer repetition alongside the occasional conscious tweaks is enough. I suspect hardware has always been a limiting factor and is more responsible for the biggest changes in times compared to any revelations about the cognitive process. Not to say that good old deliberate practice won't be the most effective way of improving for the rest of us ;)
Agreed. I also suspect part of it is a) the more time passing which means more practice, and b) more money/competition these days which means more motivation to improve. Nonetheless, better hardware, better algs/techniques, more time for repetition, etc. has all contributed. I met a younger solver at a competition a couple months back who'd been solving only for 6 months, yet averaged around 11 seconds. 10 years ago that rate of improvement would have been unheard of. I think Macky got sub-20 in less than 1 year and that was remarkable at the time.
 

Davsalooki

Member
Joined
Jan 3, 2019
Messages
10
Get a good cube, slowly work your way up. Do blindfolded cross (you can do 1-4 edges, depending on what you CAN do), and f2l pairs. You should start learning full PLL and use 2-look OLL.
 

Hazel

Premium Member
Joined
Apr 1, 2017
Messages
1,681
Location
in your walls :3
Get a good cube, slowly work your way up. Do blindfolded cross (you can do 1-4 edges, depending on what you CAN do), and f2l pairs. You should start learning full PLL and use 2-look OLL.
note: good cube does not mean an expert-level one, at your speed a budget cube will be perfectly fine. However, if you have the money to spend and you want the best on the market, feel free to buy a GTS3 M or a GAN356 X, but don't expect the expert cubes to drastically decrease your times from a budget cube like a Yuxin Little Magic or an MF3RS2.
 
Top