Need people to test new bld method

[Kubzy]

Hi. My name's kubzy and I'm a *speed*cuber from London. The biggest struggle so far I've had with cubing is BLD. It's not hard in the sense that it has a lot of algs or different notation or something, it's just I've always had trouble understanding most OP tutorials. I've been working on a method recently which I was able to understand and try learn with relative ease. It's currently a edge only method as I'm trying to figure out a way to go about the corners, but I'll be constantly updating this thread if I do anything.

Method overview:
You start normally with any preferred orientation. The buffer I use is UR and the switch position usually UL or UB if it's a t or y perm respectively. The memorization is the different part though, as you don't insert the edge oriented straight away. Using normal OP, you can just insert the piece into it's place. The "notation" in this method is quite irregular as each edge is a number, not each sticker. So, going clockwise, UB is 1, UR is 2, UF is 3 and UL is 4. FR is 5, DF is 6, FL is 7, DL is 8, BL is 9, DB is 10, BR is 11 and DR is 12. For each of these, you remember the un-oriented pieces as a prime piece. Therefore, when you're done inserting them, you know which pieces are "prime" and you can put them in the slot to flip. The flip spots are 1 and 3 or UB and UF. The flipping algorithms I use is M' U M' U M' U M' U2 M' U M' U M' U M' but I'm aware there are other ones (even the regular OLL one should be fine I believe).

So, in conclusion, I was wondering if someone more advanced/experienced in BLD solving could try out this method, check if it works and give me some constructive feed back on whether you like/dislike it and what I should change.

Thanks for reading,
peace.

 

[h2f]

No feedback so far, so I'll write about it - this method looks very bad. I cant understand why one just cant shoot the piece to its proper place and one has to do this in two steps: shoot in its place; flip when it's needed. It makes 3bld more complicated though one uses the same alg like in OP: Tperm (I guess Yperm is for corners).

 

[xyzzy]

(Disclaimer: I'm terrible at 3bld and I practise very rarely, so I'm very much not an expert here.)

This is pretty similar to 3OP, which used to be the dominant blindfolded solving method many years ago, when people didn't know any better. In 3OP, the pieces are oriented (with stuff like left Sune + right Sune for corners and (M' U)4 for edges) and then permuted with 3-cycles that preserve orientation, which is the opposite order of your proposal. (Orient then permute is better than permute then orient because you can do visual memo of how to orient the pieces and fix it right at the start of execution, rather than thinking of them in terms of like "the yellow-green edge needs to be flipped at the end of the solve" and keeping that in your mind for most of the solve.)

I believe there are two main reasons that 3OP went obsolete: one is that orienting first is super heavy on the move count (somewhat dependent on the algs you use, but most of them are long) and the other is that this requires either multiple memo passes for the edges and corners (each) or memoing two different things (orientation and cycles) at once, which is slower than a "one-pass" memo method. These issues are also applicable to your proposed method.

OP (Old Pochmann / classic Pochmann) and M2 mostly avoid these problems by solving the pieces "completely", rather than having orientation and permutation as separate steps.

tl;dr: Just stick with OP. Also, try switching to a lettering system instead of using numbers.

 

[mark49152]

What the others have said. Plus, I would point out that the fact you struggle with OP tutorials is not a good reason to invent a new method. If you can't understand OP, how do you expect to come up with a better method? Just try harder until you understand OP. It's been popular for many years for good reasons. Good luck!

 

[Kubzy]

The reason? So I can understand it. Someone that already knows and understands everything obviously won't choose this, it's a "method" I proposed for those that are trying to learn BLD, like me. Obviously you won't be getting amazing times with it, but this helped me to start to understand more of how the other methods work. Just because you've probably had knowledge/experience of other methods doesn't mean that it might not help someone else. But of course, easier to just complain than help, isn't it? I'd preferably rather some constructive criticism and someone explaining how I can fix it instead of saying that a different method is better. Just because CFOP and Roux existed, doesn't mean someone couldnt've invented Petrus.. It's clearly not the best method, but it might help someone.

 

[xyzzy]

Wait, what? We're not complaining about anything! I think it's really cool that you're exploring methods on your own, and most of the time when I see someone asking for help, I try to help with the exact method they're exploring, rather than just saying "do X because everyone fast does it".

However, in this case, I believe that 3OP and OP are better for learning purposes than the system you're proposing, which is why I mentioned them. You might disagree, and you might be right! Different people can have different approaches to learning new things, and there's nothing wrong about that—if, for you, this works as a stepping stone towards a better BLD method (OP or M2), all the better.

(I realise the tone of my earlier post might have come off as being condescending; that was not my intention and I'm just not very good with words.)

 

[mark49152]

I'd preferably rather some constructive criticism and someone explaining how I can fix it instead of saying that a different method is better.

You asked the community for feedback, and ought to accept that it won't always be exactly what you want to hear. If you just want to have a bit of fun inventing your own method, go ahead. You don't need to take any notice of community feedback.

If you want to learn to do BLD well, the most constructive advice I can offer is to persevere with OP, or even look at M2, TuRBo or commutators. As others have said, BLD methods that permute and orient separately have been pretty much obsoleted and ignored for almost a decade. Of course it's fun to play around with ideas, but ultimately it won't help you to get used to thinking in terms of separate orientation and permutation when it's widely accepted that solving them together is superior.

 

[h2f]

it's a "method" I proposed for those that are trying to learn BLD, like me. (...) it might not help someone else.

As far as I understand your first post you didnt have any success with it. I dont think it's fair to offer this method to someone who wants to learn bld.

 

[Jacck]

I'm not too surprised about the idea - but/because I'm not a good speedcuber.

I started blind with a quite silly own way, then glanced for some hints at OP and especially at M2 for edges.
My problem with M2 was: orientation! Made so many mistakes looking at the incorrect sticker (hadn't understand that at that time) and:
I knew a major-list (words for one- and two-digit-numbers). So it was easier for me to extend the major-list by A and B (12 edges) and I didn't have to learn letterpairs.
Memo of the permutation was easy then, taking two edges as one number (i.e. 35).

The orientation I memoed in a binary way: each layer has 4 edges with value 1,2,4 and 8 if it is flipped. I then memeod the sum as the corresponding letter (i.e. M for 13) which gave me three letters and I formed a word with these.

Execution started with the orientation, normally I took 4 edges together.
Did this system to win the HessenOpen 2014 with 5/7

And another thing, why I did it like that: it was clearly more information to remember, but easy information. I could memo only the orientation and afterwards the permutation which was far easier for me then doing it together (and I couldn't forget a flipped edge in place and I didn't have to learn the algs for the flipped innerslice-edges). It took me quite a while, to get faster in multiblind with the standard M2.

"When it is dark, I prefer easy things, even if they take a bit longer."

That means 1: if you just want a success, a separated orientation can help. If you want to be fast, start learning the more complicated way.
That means 2: an advanced blindcuber won't like this method

 

[G2013]

Hi Kubzy
My name is Guido, I average sub28 using commutators, but I also struggled a lot to understand OP for some reason when I learned it, some 4 or 5 years ago.
What I did, then, is watch a lot of videos about Old Pochmann and try to make up my own method from that what I had seen in the videos. Maybe the main difficulty was that they were all in English and it's not my native language, and I spoke it way worse at that time than I do now

The method I came up with was pretty similar to yours: it first oriented every single piece using a similar algorithm to yours for edges, and using sune/antisune combos for corners (R U2 R' U' R U' R' L' U2 L U L' U L for example)
Then, I'd shoot to each piece using either a T or a J for edges (as in OP but with only 11 possible targets to set up), and with either a Y or a J for corners (same).

I found this easier because I couldn't figure out the setups for each *sticker* while blindfolded, but I could figure them out for each *piece* if it was oriented, for some reason XD
I would memorise the orientation of every piece visually, as well as the permutation of the corners (tracing some imaginary lines, or idk what the hell I did XD)
Then, I'd memo my edges using a 11-letters scheme (every edge but the buffer).
I found it way simpler that way, somehow.

What I suggest that you do is the following:
1) Play a bit with that method, but only until you get quite a good understanding of the concept of solving 1 piece at a time... if you already are quite comfortable with it, DO NOT keep using your method
2) Start learning some setups for pieces that you can't easily figure out by yourself
3) Practice and spam TPS + thinkahead if you want bonus speed That's it

Keep in mind that until this year I used my weird corners visual method and M2 for edges, averaging like 2 minutes or sub2. Now I avg sub28 seconds... why couldn't you?