[Help Thread] Blindsolving

[shadowslice e]

The "best" method for blindsolving is 3-style along with a load of tricks. However, it has an extremely steep learning curve and until you've memoed virtually all the "basic" comms, you wouldn't be able to do a solve and so your memorisation/execution wouldn't be given the chance to improve.

I would therefore recommend you start with OP for corners and M2 for edges and then slowly start to learn the 3-style comms and integrate them into your solve.

I'd recommend checking out zane's and noah's tutorials

Spoiler

edit: also, as stated in the text above this subforum, this is not where you should be asking questions so this thread would be better in either cubing help and questions or blindsolving discussion.

 

[Owen Morrison]

I would therefore recommend you start with OP for corners and M2 for corners and then slowly start to learn the 3-style comms and integrate them into

do you mean edges?

 

[shadowslice e]

do you mean edges?

Oh yeah oops lol. I fixed it now. Thanks for pointing that out

 

[kubesolver]

I don't want to learn a method for it to be redundant further down the line.

I think it's unavoidable, but also it's not a big deal.

Even if method you have learned becomes obsolete, that doesn't mean that the skills you gained by learning it are obsolete nor that the work was wasted.

 

[paul.edmondson.magician]

Ok I'll start with Pochman.

Sent from my CLT-L09 using Tapatalk

 

[paul.edmondson.magician]

do you mean edges?

So it's advisable to learn both methods then? One for edges and the other for corners

Sent from my CLT-L09 using Tapatalk

 

[Owen Morrison]

So it's advisable to learn both methods then? One for edges and the other for corners

Sent from my CLT-L09 using Tapatalk

well you can't solve the corners the same way as you do the edges so you need to know both, although they are pretty similar.

 

[dudefaceguy]

The "best" method for blindsolving is 3-style along with a load of tricks. However, it has an extremely steep learning curve and until you've memoed virtually all the "basic" comms, you wouldn't be able to do a solve and so your memorisation/execution wouldn't be given the chance to improve.

I would therefore recommend you start with OP for corners and M2 for edges and then slowly start to learn the 3-style comms and integrate them into your solve.

I'd recommend checking out zane's and noah's tutorials

Spoiler

edit: also, as stated in the text above this subforum, this is not where you should be asking questions so this thread would be better in either cubing help and questions or blindsolving discussion.

I recently started learning 3style without knowing any other blind methods first. I like using commutators rather than learning algorithms, so it's actually a lot easier for me to learn 3 style than other methods that use CFOP-based algorithms that I don't know.

So far I've just been doing semi-sighted solves, where I memorize one commutator and then execute it without looking. I've also tried writing down the letters and trying to execute the full solve while following the written sequence. Basically I want to get comfortable with executing the commutators before trying memorization.

I'm having a little trouble figuring out how to start new cycles and deal with flipped pieces, since 3 style tutorials already assume you know how this works. But, in general I'm having a lot of fun learning. So, I think it is possible to start with 3 style as long as you are patient.

 

[ProStar]

Learn Old Pochman first. If you like blind solving, you can then learn M2 for edges, but stick with OP for corners. Then, if you want to continue, after practice you could learn orazco(think I spelled that wrong) for corners(so now you'd be using Orozco+M2). Then, if you reeaaallly like blind solving, you could learn the couple hundred algorithms needed for 3-style. This may seem like a lot of steps and uneeded methods, but the stuff you learn won't be in vain. Knowing OP will help you transition onto M2, just like learning beginners method helped before learning CFOP(or another advanced method)

Oh, also, this may be personal preference, but I always thought putting letters on your cube was pointless. The letter scheme is unbelievably easy, just go clockwise around the faces with each letter. Then do the same for corners. That may just be personal opinion though

 

[fun at the joy]

orazco(think I spelled that wrong) for corners(so now you'd be using Orozco+M2).

at least you learned really quickly how to spell orozco

 

[ProStar]

at least you learned really quickly how to spell orozco

Well that was lucky XD

 

[dudefaceguy]

I’ve had some success with my pseudo-3-style blindsolving. I’m wondering if there is anyone else who solves this way, or if it’s just me.

I make up the commutators on the fly during the solve, using this technique. My buffer is U for edges, and V for corners, both on the bottom layer. I always use the bottom layer as the interchange layer. I conjugate one of the other stickers to the bottom layer, and one to the top layer. I always interchange with the buffer. The conjugates are never more than 3 moves, which makes them easy to reverse. I find this easy to visualize because the commutators always have the same orientation.

Does anyone else do this? I imagine not, because it is faster to learn algorithms. I really hate learning algorithms, and I’m not really concerned with speed. I just want a fun and challenging method that I will remember even if I stop practicing it for a year.

 

[Cuberstache]

I’ve had some success with my pseudo-3-style blindsolving. I’m wondering if there is anyone else who solves this way, or if it’s just me.

I make up the commutators on the fly during the solve, using this technique. My buffer is U for edges, and V for corners, both on the bottom layer. I always use the bottom layer as the interchange layer. I conjugate one of the other stickers to the bottom layer, and one to the top layer. I always interchange with the buffer. The conjugates are never more than 3 moves, which makes them easy to reverse. I find this easy to visualize because the commutators always have the same orientation.

Does anyone else do this? I imagine not, because it is faster to learn algorithms. I really hate learning algorithms, and I’m not really concerned with speed. I just want a fun and challenging method that I will remember even if I stop practicing it for a year.

Ah yes, I did something similar for a while, making up algs on the fly based on intuition. Here's what I learned: It sucks. I see from other posts that this is your first blind method and you don't have a success yet. In light of that, I recommend sticking with what you have until you get one success, because that first success feels awesome and your method is good enough to get that. If you want to do 3-style the right way, for speed, here's how I would go about doing that. First off, switch your buffers to UF/UFR (C/C with speffz) as soon as possible. That buffer set has the best comms, and you can continue using your method in a very similar way. When you feel up to it, go to a commutator sheet (I use Daniel Lin's, found here, and you can even customize your orientation and lettering scheme!). Here's the important part: Learning commutators is nothing like learning algorithms. You already understand setups and interchanges and all that, so the sheet just tells you the setups that give you the best commutators. They won't be move-optimal, which can be hard to get over, but they will be speed-optimal. Just follow the setups they have, understand how it moves the pieces around, and treat it just like the ones you came up with, only slightly optimized ones. It will still take a lot of time and effort to remember all of them, but nowhere near what it would take to learn them as algorithms. Best of luck, and don't be afraid to ask further questions!

 

[dudefaceguy]

Ah yes, I did something similar for a while, making up algs on the fly based on intuition. Here's what I learned: It sucks. I see from other posts that this is your first blind method and you don't have a success yet. In light of that, I recommend sticking with what you have until you get one success, because that first success feels awesome and your method is good enough to get that. If you want to do 3-style the right way, for speed, here's how I would go about doing that. First off, switch your buffers to UF/UFR (C/C with speffz) as soon as possible. That buffer set has the best comms, and you can continue using your method in a very similar way. When you feel up to it, go to a commutator sheet (I use Daniel Lin's, found here, and you can even customize your orientation and lettering scheme!). Here's the important part: Learning commutators is nothing like learning algorithms. You already understand setups and interchanges and all that, so the sheet just tells you the setups that give you the best commutators. They won't be move-optimal, which can be hard to get over, but they will be speed-optimal. Just follow the setups they have, understand how it moves the pieces around, and treat it just like the ones you came up with, only slightly optimized ones. It will still take a lot of time and effort to remember all of them, but nowhere near what it would take to learn them as algorithms. Best of luck, and don't be afraid to ask further questions!

That’s for the detailed reply! I will check out those commutator resources. I do enjoy learning more about commutators.

What’s the advantage of using C as the buffer instead of any other sticker? Shouldn’t they all be basically the same?

Also I had my first success today! It did indeed feel amazing.

 

[Cuberstache]

That’s for the detailed reply! I will check out those commutator resources. I do enjoy learning more about commutators.

What’s the advantage of using C as the buffer instead of any other sticker? Shouldn’t they all be basically the same?

Also I had my first success today! It did indeed feel amazing.

Awesome! Congrats on your success! The advantage is that overall, the commutators for C are faster than other ones. Having the buffers in the front of the cube means more commutators are RUD or RUM or other fast movegroups. I personally don't know a whole lot about how fast different buffers are, but I'm pretty sure that other people have tested out different buffers, and UF/UFR have been determined to be fastest. If you're going to learn commutators, there's really no reason not to learn the best ones.

 

[dudefaceguy]

After many more failures and a couple of successes, I have a question about memorization accuracy.

I see that WCA regulations take the best of 3 solves for 3bld, which means that you can fail twice and it doesn’t matter as long as you get a good time on the third solve. It seems like high-level blind techniques are geared towards a high-risk, high-speed strategy for this reason. I've heard that the fastest blind solvers only have about a 60% success rate.

As for me, I do not hope to be fast - I am an old man and I am doing this to relax and essentially meditate. I’m more interested in accuracy than speed. Do you have any suggestions on different memorization techniques for increased accuracy?

So far I am memorizing images for edges rather than just sounds. It takes me a long time to memorize and execute, so audio memorization doesn’t really work for me. Is it normal for a beginner to memorize using more robust techniques, and then transition to audio/image once they get faster?

 

[Cuberstache]

After many more failures and a couple of successes, I have a question about memorization accuracy.

I see that WCA regulations take the best of 3 solves for 3bld, which means that you can fail twice and it doesn’t matter as long as you get a good time on the third solve. It seems like high-level blind techniques are geared towards a high-risk, high-speed strategy for this reason. I've heard that the fastest blind solvers only have about a 60% success rate.

As for me, I do not hope to be fast - I am an old man and I am doing this to relax and essentially meditate. I’m more interested in accuracy than speed. Do you have any suggestions on different memorization techniques for increased accuracy?

So far I am memorizing images for edges rather than just sounds. It takes me a long time to memorize and execute, so audio memorization doesn’t really work for me. Is it normal for a beginner to memorize using more robust techniques, and then transition to audio/image once they get faster?

Yeah, using images and slowly transitioning towards audio is normal. Early on, it's good to spend a lot of time thinking of good images and how to connect them together. As you improve, you'll rely on the words themselves more than the images they create. Getting a vague feel for the direction tracing goes can be a bigger deal when you're fast. Also, people don't generally switch to full audio for edges until sub-1:30 at the earliest so don't worry about it.

 

[Habsen]

Is it normal for a beginner to memorize using more robust techniques, and then transition to audio/image once they get faster?

As a total beginner, you have to recall the information a couple of minutes after memorizing it. More robust techniques or multiple memo passes are the way to go here.
Top cubers memorize in less than ten seconds and solve it within the next 15 seconds. They need to remember the information for a very short time only which makes techniques such as audio more feasible.
Using images/words for everything should be fine right now. You can try to use audio for more pieces once you get faster at tracing and execution.

 

[dudefaceguy]

Yeah, using images and slowly transitioning towards audio is normal. Early on, it's good to spend a lot of time thinking of good images and how to connect them together. As you improve, you'll rely on the words themselves more than the images they create. Getting a vague feel for the direction tracing goes can be a bigger deal when you're fast. Also, people don't generally switch to full audio for edges until sub-1:30 at the earliest so don't worry about it.

As a total beginner, you have to recall the information a couple of minutes after memorizing it. More robust techniques or multiple memo passes are the way to go here.
Top cubers memorize in less than ten seconds and solve it within the next 15 seconds. They need to remember the information for a very short time only which makes techniques such as audio more feasible.
Using images/words for everything should be fine right now. You can try to use audio for more pieces once you get faster at tracing and execution.

Great, thank you guys!

I see your point about getting a vague feeling for the direction of the cycles. I had a successful solve tonight in which I forgot my corner memo completely for about a minute. I eventually remembered it by remembering the first corner swap visually, and recalling the first letter from that.

Thanks again - blindsolving is incredibly fun. As an intuitive solver, I feel like it's designed for me.

 

[dudefaceguy]

I've recently started executig all edge and corner flips first, which seems to improve my accuracy. When do you guys usually execute your flips? Thanks.