Update On "CKZOP" System

[dChan]

Just thought I'd update you guys on how 'my' experiment is going. For those of you who dont't know what I'm taling about here is how the method works:

1. Cross
2. Keyhole F2L + Place Last Edge and Orient LL edges with ZBF2L(1 of 24 cases)
3. Orient LL Corners(7 cases)
4. PLL

I found that by using ZBF2L you have a greater chance of experiencing an OLL skip. This is due to using algs from ZBF2L that preserve or flip certain edges. These algs also affect corners- though not directly- so that if you know how the alg works you can presevre LL and achieve an OLL skip. This can sometimes result in a full LL skip because the pieces are either already in place or come together when you preserve the LL orientation with ZBF2L.

Using this method I have gotten my fastest time of 21.45 and an average of 28 sec- though this is not fast for most people it is fast considering the fact that I am a 'beginner' starting only two months ago. So in the hands of an experienced cuber, this system could yeild faster times.

My average Cross requires 2-4 seconds, my F2L needs 13 seconds(Keyhole+ZBF2L), my OLL is 5 seconds, and my PLL is 4 seconds(3 seconds per alg but 1 second for recognition). This is a theoretical average of 26 seconds. If I were to delete OLL I would have a time of 21 seconds which is my fastest time.

That's just my update in case you are interested.

 

[doubleyou]

very interesting strategy! dunno if you watched Erik's F2L (fridrich) vid. but at one point he end up (at the last pair) with no edges oriented. so he insert the pair in an unusual manner to orient some edges. it would require much more insight of the ZBF2L to be able to preserve all edges.

VERY imoressive times after so little time!

 

[Erik]

You can orient all LL edges with only 6 algo's of which at least 3 are easy. They require that the corner and edge are connected already thouhg. I'm not sure how you got to the number 24 (4*6?) but at nevertheless I wouldn't call it ZBF2L because that means you orient the edges from about ANY position...

 

[Dirk BerGuRK]

Thats a really cool method, and it looks like you will get some really fast times with it. Just a thought, it seems kind of inefficient building a cross and then inserting three corners to finish the keyhole F2L (if this is the way you do it, I could be wrong). Whenever I was playing around with keyhole F2L, I always liked to build the first layer (minus the working corner) on the left. By building it on the left, I would build a Roux block and add another corner-edge pair which is surprisingly easy to insert taking advantage of the 'r' turn. Then I could launch right into inserting the rest of the F2L edges with the first layer on the bottom or left (both are very easy to do). Just my two cents.

 

[dChan]

Originally posted by Erik@Apr 23 2007, 11:48 AM
You can orient all LL edges with only 6 algo's of which at least 3 are easy. They require that the corner and edge are connected already thouhg. I'm not sure how you got to the number 24 (4*6?) but at nevertheless I wouldn't call it ZBF2L because that means you orient the edges from about ANY position...

The 24 cases come from the 24 ZBF2L algs need to insert the last corner. Basically you insert the edge in the M-layer and then use the appriate alg for the situation(which edges have not been flipped- use alg x to fix. Is the LL already oriented minus one corner? Use x alg to fix). Of which there are only 24 when you insert the last edge of F2L instead of 96 cases of ZBF2L when you have an 'empty' slate.

Also I've only been using this system for less than two months so that really shows that this system is promising. I used a beginners method for the first weeks then a semi-advanced system, and now I use this.

 

[joey]

It kind of just sounds like the VH method really.

 

[dChan]

Originally posted by joey@Apr 23 2007, 03:57 PM
It kind of just sounds like the VH method really.

Sort of but the VH method requires COLL(40 cases) + Full ZBF2L(96 cases). This system is a 52 algorithm system, muc hsmaller than most advanced methods and still gives results. VH method requires 164 algs, so that's a big difference.

 

[tenderchkn]

Why does it take 5 seconds to do OLL? All of my COLL algs are sub-3, and about half of them are sub-2, so there's no way easy edge-correct cases should take 5 seconds.

 

[joey]

Originally posted by tenderchkn@Apr 24 2007, 03:39 AM
Why does it take 5 seconds to do OLL? All of my COLL algs are sub-3, and about half of them are sub-2, so there's no way easy edge-correct cases should take 5 seconds.

He has probably been doing those 7 OLL algorithms for a shorter time than you, so for him they are slower. For me, those 7 OLL are at 3 or sub-3. Especially the sune!!

 

[AvGalen]

dChan: You are not going to delete OLL-entirely using this method. That would just be a skip/lucky! And tenderchkn and joey are right. 5 seconds just for corner OLL is slow. 13 seconds for keyhole+ZBF2L-edge is fast though. Are you sure it is not 13 seconds keyhole (except last edge), ZBF2l-edge + corner OLL 5 seconds?

A sune takes me about 1 second. Worst case is 3 sunes, so 4 seconds for corner OLL should really be maximum. Even when you are using only 1 alg and get worst case!

At seems we are at the same level right now. You got here much faster than I am, so you will probably be sub-20 before I am.

 

[dChan]

Yeah, I only learned the full 7 cases + some easy ones(T, H, etc. OLLs) about two weeks ago and I am using some B moves right now for it which are slow for me. I'll look up different algs.

The skip needs to be 'worked for.' Sometimes if you do a different alg than what is needed for the case you will get a two'corner only OLL which is easy to do. If you do for example R2 U R2 U R2 U2 R2(I think this is correct) on a solved cube, position the corner over the edge it is supposed to be paired with and do the alg again, there you have an OLL skip. There are many other cases you have to find simply by doing F2L this way, so if you ty it you'll get the idea of what I do. I can't explain it too well because some of the cases are complicated so this is the easiest example.

 

[AvGalen]

What would you do if you got F R U R' U' F' R2 U R2 U R2 U2 R2