[WR] Haiyan Zhuang 32.27 BLD single

[joey]

Brain doesn't really mean too much in 3x3 speed.. at least it's not the same sort of thing you need for 3x3 BLD.

No offense, but I find it weird someone who cubes would think that.. Maybe someone who didn't cube though.

 

[Zane_C]

I still can't get over how good this guy is at BLD!!!

 

[4Chan]

It's a similar premise to BH, except with more fingertrick friendly algs.

According to some reputable sources, that IS BH.

Thanks so much for posting, Hai Yan. This makes me feel much better. I was feeling guilty before because I said you were using BH, and then I found out you were apparently using more ordinary freestyle instead. But now I see that you're using what I personally consider to be essentially BH anyway. As Chris says, technically, BH uses exclusively optimal commutators, but you use optimal algorithms for speed instead, so yours is a little different. So it really is freestyle. But it also means that when someone else came on this forum a while back and said you only know some small number of algorithms, they were incorrect - you really do know optimal speed algorithms for every case.

Same reputable source.

I'm not a noob.
I too, browse this forum constantly.

 

[joey]

It's a similar premise to BH, except with more fingertrick friendly algs.

According to some reputable sources, that IS BH.

Thanks so much for posting, Hai Yan. This makes me feel much better. I was feeling guilty before because I said you were using BH, and then I found out you were apparently using more ordinary freestyle instead. But now I see that you're using what I personally consider to be essentially BH anyway. As Chris says, technically, BH uses exclusively optimal commutators, but you use optimal algorithms for speed instead, so yours is a little different. So it really is freestyle. But it also means that when someone else came on this forum a while back and said you only know some small number of algorithms, they were incorrect - you really do know optimal speed algorithms for every case.

Same reputable source.

I'm not a noob.
I too, browse this forum constantly.

anyway, Haiyans method isn't really new or anything. He's just good at it, and actually took it further than others. Like me, who could never be bothered to finish finding algs D:

 

[Stefan]

He should have just memorised the whole cube during the 15 second inspection time and use BLD algorithms to get 15-20 sec speedsolve.

Why? It's not like exec means anything. This way he actually gets to practise the proper BLD procedure.

Well, it might for once get him to the second round so he'd get more official attempts.

 

[r_517]

He should have just memorised the whole cube during the 15 second inspection time and use BLD algorithms to get 15-20 sec speedsolve.

Why? It's not like exec means anything. This way he actually gets to practise the proper BLD procedure.

Well, it might for once get him to the second round so he'd get more official attempts.

BLD is always set between the first and second round of speedsolving in China

 

[mazei]

Just for Haiyan I'm guessing.

 

[zhanglei1943]

He uses Freestyle BLD method.

I heard that haiyan memorizes about 3 thousand of algs of BLD...

 

[KJiptner]

No way. Using a fixed buffer (he does), there are only about 800 cases for corners and edges (3-cycles). Of course he could learn algs for longer cycles but I don't think he does.

 

[irontwig]

No way. Using a fixed buffer (he does), there are only about 800 cases for corners and edges (3-cycles). Of course he could learn algs for longer cycles but I don't think he does.

Here he says 1000+, so I guess 3-cycles+some other stuff:

http://www.speedsolving.com/forum/showthread.php?t=16358

 

[KJiptner]

ah okay. but 3000 is a bit too much

 

[zhanglei1943]

ah okay. but 3000 is a bit too much

sure, you know it is only "heard"...

 

[Konsta]

http://www.popbuying.com/detail.pb/sku.Tape_A_HaiYan_3x3x3_Magic_Cube_DIY_Kit_Black-28303

http://www.popbuying.com/detail.pb/sku.Tape_A_HaiYan_3x3x3_Magic_Cube_DIY_Kit_White-28302

Just bought 2 black ones and 1 white Had to buy a white one too, because Haiyan himself was using it

 

[Mike Hughey]

It's a similar premise to BH, except with more fingertrick friendly algs.

According to some reputable sources, that IS BH.

Thanks so much for posting, Hai Yan. This makes me feel much better. I was feeling guilty before because I said you were using BH, and then I found out you were apparently using more ordinary freestyle instead. But now I see that you're using what I personally consider to be essentially BH anyway. As Chris says, technically, BH uses exclusively optimal commutators, but you use optimal algorithms for speed instead, so yours is a little different. So it really is freestyle. But it also means that when someone else came on this forum a while back and said you only know some small number of algorithms, they were incorrect - you really do know optimal speed algorithms for every case.

Same reputable source.

I'm not a noob.
I too, browse this forum constantly.

anyway, Haiyans method isn't really new or anything. He's just good at it, and actually took it further than others. Like me, who could never be bothered to finish finding algs D:

I think this shows that Chris (who is the "H" of BH and hence qualified to speak about it) probably doesn't consider Haiyan's method to be BH, but I (who am not anything ) do consider Haiyan's method to be BH. It's all a matter of what you call it. But for me, the most significant thing about BH is the idea that you get where all algorithms are instantly recallable and totally braindead and very fast, and Haiyan's method fits that category perfectly. And I think freestyle is not as good a term, because freestyle implies that you decide what you're going to do to solve as you go, but having a fixed choice for every case means there's nothing "freestyle" about it - it's a completely rigid system. But I guess that depends on how you define the term "freestyle". Some definitions imply "ad-libbing", and I guess that's how I think of it, which is why freestyle seems like the wrong term to me for Haiyan's method.

When I finish learning my algorithms for edges (which I'm getting close to), I will consider myself to be totally BH for 3x3x3. And Chris will think I'm not, since half of my algorithms for edges will not be optimal commutators. I'll probably tell everyone I'm doing BH, though. Or at least I'll say I do speed-optimized BH.

If someone asks me what method Haiyan uses, I tell them, "I consider him to essentially use BH." Chris would probably answer the same question by saying that he uses freestyle. Whatever.

 

[cmhardw]

Wow, this is an interesting discussion. If you had asked Daniel and I two years ago what BH was we would have said pretty much what is said on the wiki page for BH:

A blindsolving method based on the use of 3-cycle algorithms. The fundamental idea behind the BH method is to use pre-memorized, move optimal, commutator 3-cycles for all possible 3-cycles starting from a fixed buffer location. For pieces where orientation is defined, the position and orientation of pieces are solved simultaneously.

Perhaps Haiyan's method, speed optimized BH, or move optimal BH are all variations of some larger meta-method? I would include all of Classic Pochmann, M2/R2, TuRBo, and also big cube center U2 methods in this too.

I don't know what the meta method would be called, but the core ideas behind it seem to be:

1) Braindeadedness - no thinking or ad-libbing whatsoever. Every case must have a pre-determined algorithm to solve it. You can use commutator concepts and intuition to come up with this algorithm on the fly so to speak, but you must essentially use the same algorithm(s) for each and every case.

-----side point-----
This point seems hard to phrase, as I would argue that beginner/intermediate level freestyle would not count in this meta-method. However, truly expert level freestyle cubers would eventually develop patterns and always solve certain cycles the same way. This essentially makes their method merge into a variant of this meta method.
-----side point-----

2) Pieces are solved via direct solving, there is no step that only orients pieces. The only exception to this is correctly permuted, but disoriented, pieces. Even Daniel might dispute this, as he sometimes uses commutator 3-cycles instead of commutator mono-flips to solve these cases.

3) Either 1 or 2 pieces are solved with the execution of each step or algorithm.

What does everyone thing about expanding our nomenclature for BLD cubing with some sort of meta method? I think this might be a better solution to the problem? Or at least this post will let us open up the discussion to some new ideas perhaps?

Chris

 

[joey]

Argh I just don't know what to think about it all. (In regards to Chris' post)

ps MHOBH, Mike Hughey Optimised BH... no, let's not go that far!

 

[Micael]

No way. Using a fixed buffer (he does), there are only about 800 cases for corners and edges (3-cycles). Of course he could learn algs for longer cycles but I don't think he does.

Here he says 1000+, so I guess 3-cycles+some other stuff:

http://www.speedsolving.com/forum/showthread.php?t=16358

I wonder how many parity algs he has. Someone know about it?

 

[Mike Hughey]

Argh I just don't know what to think about it all. (In regards to Chris' post)

ps MHOBH, Mike Hughey Optimised BH... no, let's not go that far!

Hopefully it won't catch on, or we'd have MHOBH Rule!

I bet yoruichi thinks this is the silliest sub-thread he's ever read. And perhaps he's right - we'd all get better at BLD if we quit talking about it and just did more BLD practice.

But Chris, you're probably just about the only one highly-respected enough that you could come up with a BLD method and metamethod classification system that people might actually use. Feel free to do so if you want - I'd probably use it.

 

[brunson]

12 and a half seconds (28%) faster than the #2 record. What I find awesome is that he set a WR in the first round, then broke it in the second. If all his 3x3 results are BLD solves, as reported, he has set the top 8 BLD times.

Truly awe inspiring!

 

[CitricAcid]

Why doesn't he just use his BLD method for regular 333 speedsolving?

He would be getting OK times.