ZBF2L

[Jason Baum]

Hi guys,

I just registered for this forum, but I've been cubing for about 2 1/2 years now. I'm sure I've seen some of you at some of the competitions I've been to.

Anyway what I want to talk about is ZBF2L, or ZB in general. Is there anybody else learning this besides me? I know there were a few people learning it a while ago, I just haven't heard anybody talking about ZB for a while. As of now, I know 189 out of the 306 ZBF2L cases. My goal is to know all of ZBF2L by the end of summer. I haven't touched ZBLL yet, and yes it scares the crap out of me haha.

My average for Fridrich F2L is usually around 8-9 seconds (sub 8 on a really good day, hehe). I truly believe that ZBF2L can get down to as low as 9 seconds with practice, and I think this is absolutely essential for the ZB method to have a chance at being faster than Fridrich. Whenever I use ZBF2L in my solving now (with COLL PLL), I average about 15 seconds. I have averaged sub 15 once using this method (discarding solves I get where I do not know the ZBF2L). Keep in mind that my average average for Fridrich solving is low 13, so I still have a ways to go, but I think I am off to a great start with ZBF2L. I do think that eventually, with a lot of practice, my ZBF2L + COLL/PLL average can get pretty close to my Fridrich average. I don't think it will get any faster until I know a good chunk of ZBLL, but that's way down the road.

Am I the only one who has any faith left in the ZB method? It would be nice if there are other people out there who are learning it as well, but as far as I know I am the only one still learning it. ZBF2L isn't that bad at all, it's actually very easily manageable if you put the time into it.

Comments?

 

[Inferno.Fighter.IV]

ZBF2L sounds wicked, I was thinking about it, but I can't find any tutorial pages. I've heard that it's solved in a similar fashion to Fridrich, but for the last pair you not only fill it in but you orient some part of the last layer. I really like the sound of that...

 

[dChan]

I thought ZBF2L only have 96 cases?

I've learned most of the cases that involve an edge already placed and the corner needing to be placed but I've pretty much abandoned it for an easier Petrus style orinetation of edges after a 2x2x3 block has been created.

ZBF2L is more like a Fridrich version of Petrus's orientation step where as instead of by intuition, you orient using an algorithm. ZB method actually sprung out of Petrus' 5+6+7 step LL list by BH.

 

[tenderchkn]

The emphasis these days is on raw speed - something that ZB cannot match. Algorithms are so optimized that there really isn't an advantage to doing ZBF2L followed by ZBLL (especially when those are not optimized and probably never will be). Even if you didn't have to do ZBF2L, do you really think that you can average under 4 seconds doing ZBLL? If it takes you say, 1 second for recognition and 2 seconds for execution, then it'll save you 1 second over what you normally do with OLL/PLL. The problem when the whole ZB method is that it is a lot of work for little or no gain - empirically at least, there has yet to be a single person who is faster at ZB than they are solving with Fridrich.

I do some COLL cases myself (without ZBF2L), but only the ones that aren't overly awkward, because OLL/PLL would be faster. But there are some cases where the gain is so slim (under a second) where I wonder if I should just have the regular OLL.

 

[dChan]

I wouldn't doubt the speed of the ZB method. Look at Chris Hardwick, he has a super fast average with ZB. I think the problem is only getting used to doing ZB. If someone were to doing ZB without delays eg. looking ahead 100%: Cross 2 seconds, 3 c/e pairs + ZBF2L 4 x 2, and ZBLL 3 seconds(recognition + algorithm execution, probably impossibly to look ahead on this step), they could get 13 seconds and maybe lower if they execute their F2L uber faster. Chris Hardwick has demonstrated that you can get really fast times with Zb, you just need to practice.

 

[Erik]

I think a few things:
1. ZB is very fast, it saves you 1.5 to 2 seconds on average when mastered fully I think.
2. ZB is very frightning to people as it has loads of algs.
3. Not a lot people are learning it yet.
4. Err... 1,2 and 3 should already be enough.

 

[pjk]

Welcome to the forum Jason. I have that Dallas '05 DVD of you solving 22+ seconds, and I know from some of your strangepuzzle videos that you have came along way, good work.

I agree with Erik on this, enough said Keep up updated on your progress. Do you do BLD at all?

 

[Harris Chan]

If mastered properly, it is theoretically possible to be faster than Fridrich. For example, you have a scramble, and you create 2 solutions: a Fridrich solution, and a ZB solution. Now execution the solution as fast as you can, with no delays. Supposedly/theoretically, ZB solution would be faster. Currently though, the problem is that the delays are harder to reduce in ZB.

But who knows, that's what people used to think of Fridrich initially too, no? But being able to maintain the recognition and execution of ALL the algorithms would be harder, because the frequency of the algorthms is less, much less than Fridrich.

Perhaps it's a dead end...may be people will invent another method, that's faster and more efficient. For now, work on the F2L, and find better LL algos

-Harris

 

[Johannes91]

I know almost 45% of ZBLL, but I only use 35% or so because recognition and recalling the alg takes too long on some cases. I don't believe in ZB (or Fridrich) though, starting with cross just wastes a bunch of moves. I build F2L in blocks, Petrus or Heise style. There's one guy who memorized whole ZBF2L and ZBLL pretty quickly using some memory system, but he isn't a speedcuber, just interested in solving for few moves in a reasonable amount of time.

 

[Jason Baum]

dChan- ZBF2L has 306 cases, or 158 not counting reflections.

tenderchkn- You bring up some good points, but I have to disagree. I think that ZB can match Fridrich in terms of raw speed. Like I was saying in my first post, I really feel that ZBF2L can come very, very close to matching Fridrich F2L. I know almost 200 cases, and I can only think of 2 or 3 cases that I've learned so far that I really don't like, meaning that there is NO WAY those cases would ever be faster than just doing the normal Fridrich F2L for that pair. For the rest of the cases, I believe that they can all come very close to being as fast as Fridrich with enough pracice. I haven't looked at ZBLL yet, but I'm sure the reason that the algs aren't optimized is that people haven't worked on it as much as OLL/PLL. As long as your ZBF2L can come close to matching your Fridrich F2L, then ZB stands an extremely good chance at being the faster method once mastered. I also agree that the difference might turn out to be very small, maybe only a second or so. However, that one second might turn out to be the difference between averaging sub 11 and averaging sub 10, which is a huge difference to me.

PJK- Yes I do BLD, but I hardly ever practice it. I ususally get anywhere in between 3 and 5 minutes, but my completion rate is just awful (I would estimate 33%). I have only gotten two sub 3 BLD solves. I would like to get faster at it, but IMO it is very boring to practice.

 

[El Veintitres]

So I've done a little research on ZB on my own as well and there seem to be some conflicting facts. In some places (like this thread) I hear people say there are 306 ZBF2L algorithms, and yet in some other places I hear people saying that there are over 1700. Most of the stuff I've read on reputable sources lead me to belive that there are 306, but I just wonder where this misconception about there being 1700 cases comes from.

 

[mcciff2112]

I think people tend to exaggerate when they talk about ZB. I trust Jason more than anyone, since he was one of the few who attempted (and succeeded) to learn all of it. But most hear that there are a lot of algorithms, and the amount just grows from there. According to Jason's page on his site, there are 800 cases in ZB (306 ZBF2L + 494 ZBLL), but I've heard people say that there are close to 2000 algorithms (But these people are just saying what they have heard, not what they learned).

 

[Johannes91]

I think people tend to exaggerate when they talk about ZB. I trust Jason more than anyone,

Fortunately there's no need to trust anyone, since verifying the numbers isn't hard to do.

since he was one of the few who attempted (and succeeded) to learn all of it.

Did he finish ZBLL? Not long ago he wasn't even close. Who are the other few?

 

[Jason Baum]

Nope, the most ZB I knew at one point was 100% ZBF2L and a little over 30% ZBLL. I did however spend a lot of time with ZB and at one point could consistently average sub 12 with what I knew.

As for the number of cases, there are 305 ZBF2L cases, 306 including the solved case. There are 493 ZBLL algorithms, 494 including the solved case. So that makes a total of either 798 or 800 cases altogether, depending on how you look at it. 800 is more aesthetically pleasing than 798, so I usually consider ZB to have 800 cases. Not sure where this talk about close to 2000 cases came from, but it definitely isn't true.

 

[royzabeast]

I was on somebody's website, I forgot whose, but I do recall that they had started learning the algorithm's a year and a half ago and have since been halfway done.

More or less .

 

[Cride5]

There are 493 ZBLL algorithms, 494 including the solved case.

Did you find that mirroring and/or inverting the ZBLL algs wasn't really practical for speedcubing?

 

[Jason Baum]

Did you find that mirroring and/or inverting the ZBLL algs wasn't really practical for speedcubing?

For me personally, no not really. My left hand is horrible compared to my right hand, so I never really mirrored much of anything. I did use inversions, but to me an inversion feels like a completely different alg.

 

[blah]

How about mirroring across the S slice then? That's what I do because my left hand sucks too, definitely worse than yours