ZZ-CT Thread

[4Chan]

I highly disagree, by putting the edge in and doing those algs, you don't get any of the quick, short, easy cases, and then you force the difficult cases.
Not only are they more difficult, but it raises the probability of the "5 twisted corner with edge permuted" cases, which are difficult to recognize and execute. Because of that I don't believe that there is a shorter average move count, and I know for sure that recognition isn't as good.

I intentionally leave the edge in the top layer, so that I can get the nicer easier cases.
By combining this technique with third slot corner control, it forces short, easy to recognize cases more often.

Additionally, neither of them are subsets of eachother. They are both subsets of ZZ-C.
CLS isn't a subset, it's an option.

Also, in regards to CPLS+2GLL, I used to do that for OH, and recognition wasn't so bad:
1. Make pair and hover over DFR slot.
2. Pretend the DFR corner is in UFR.
3. Look at the corner swap.
4. Execute insert that solves CP.
5. 2GLL

 

[PhillipEspinoza]

I guess I meant TSLE is a variation of CLS, not a subset of the CLS algs we know from MGLS. Or maybe it's a variation of ELS where you orient corners as well. That seems to be more fitting.

I have no numbers to back this up but I feel like TSLE skips are more likely if you force edge insertion during block building. I feel like common sense kinda verifies this as the edge is already inserted therefore you can TSLE skip more frequently. Kinda like doing 3rd slot corner control like you were talking about, but a little easier and sometimes the LSE is free and you don't have to work for it. Even if I give you the idea that move count is higher for those edge inserted algs, which I don't really agree with, the overall move count is reduced by the fact that you can start block building from any side and the LS is not fixed. In addition to that, you dont have to worry about learning to recognize the mirrors from all the different LS's.

Those edge inserted (OCLL) cases are still trigger friendly and I actually think the no-corners oriented cases are easier to recognize in this scenario because you only have a handful of cases that it could actually be. It's like recognizing OLL for the cross cases vs recognizing CLS. H and Pi are not hard to recognize for OLL.

Also, wouldn't those rare TSLE cases where simple insertion would be R U' R' be replaced with TSLE skips?

I feel like if you're going to learn to execute and recognize all these different corner orientation cases, you might as well learn CLS as it gives you a higher chance of an LL skip.

EDIT: Also, the CPLS I was referencing was Stachu's circa 2010 where you don't form a pair. It would be like CLS, after ELS, but just permutation of corner while inserting.

 

[genericcuber666]

if i learn tsle and im still using ttll this will still be faster than coll/eplll because the recognition for the first look of ttll is faster than coll recogniton right?

And is anyone else having problems win the tsle anki file?

 

[JTWong71]

I guess I meant TSLE is a variation of CLS, not a subset of the CLS algs we know from MGLS. Or maybe it's a variation of ELS where you orient corners as well. That seems to be more fitting.

I have no numbers to back this up but I feel like TSLE skips are more likely if you force edge insertion during block building. I feel like common sense kinda verifies this as the edge is already inserted therefore you can TSLE skip more frequently. Kinda like doing 3rd slot corner control like you were talking about, but a little easier and sometimes the LSE is free and you don't have to work for it. Even if I give you the idea that move count is higher for those edge inserted algs, which I don't really agree with, the overall move count is reduced by the fact that you can start block building from any side and the LS is not fixed. In addition to that, you dont have to worry about learning to recognize the mirrors from all the different LS's.

Those edge inserted (OCLL) cases are still trigger friendly and I actually think the no-corners oriented cases are easier to recognize in this scenario because you only have a handful of cases that it could actually be. It's like recognizing OLL for the cross cases vs recognizing CLS. H and Pi are not hard to recognize for OLL.

Also, wouldn't those rare TSLE cases where simple insertion would be R U' R' be replaced with TSLE skips?

I feel like if you're going to learn to execute and recognize all these different corner orientation cases, you might as well learn CLS as it gives you a higher chance of an LL skip.

EDIT: Also, the CPLS I was referencing was Stachu's circa 2010 where you don't form a pair. It would be like CLS, after ELS, but just permutation of corner while inserting.

When you insert the F2L Edge in, it just becomes a more efficient EJLS (Or EJF2L), where the F2L Corner's Position isn't necessarily at DFR.

TSLE Skips are more likely, of course because the cases are reduced and AUF's don't count due to symmetry reasons. On the Algorithm site there are 3 algorithms requiring 4 Triggers to solve Corner Orientation when the Edge is inserted. Also, from Chris Tran's Seminar, he said that part of the thing with ZBLL Algorithms is that they just break up the F2L, and Restore it. What you're trying to do is Break up the Edge, and Restore it with Solved Corner Orientation.

There isn't really a more likely chance for the simple cases if you don't pay attention to Corner Orientation when placing the Edge. Part of the reason CLS isn't favored as much as TSLE is because the overall Movecount is lower on average, because only 1 piece is being solved, and when a CLS Case comes when there is an Edge and F2L Corner in the correct slot, you are just breaking up the Corner along with the Edge. The normal F2L Cases for those 2 without caring about Corner Orientation are one of the worst ones when EO is solved.

The Movecount from all angles really depends on how you Blockbuild, efficiency-wise, but you still get a ton of choices to choose from when Blockbuilding for ZZF2L. It's very unlikely that after you solve EOLine, Square to build are all bad cases Except only the Square at DFR. Even if that does happen, if you can see it in inspection, just solve EOLine with a (y2) beforehand so the good case is on BLD (Strange name for a Corner Position).

If you look carefully at Corner Orientation, all the cases are just repeated 5 times, with the F2L Edge in different positions in the U-Layer or in the slot.

Say you get the Scramble for TSLE as: (R U R' U2 R U2 R')
Doing it the way you suggested, you would do (R U' R').
The algorithm for that Edge inserted case is (R U' R' U2 R U' R'), making TSLE 10 Moves.
Standard TSLE would just be the inverse, which is only 7 moves compared to 10.

 

[JTWong71]

if i learn tsle and im still using ttll this will still be faster than coll/eplll because the recognition for the first look of ttll is faster than coll recogniton right?

And is anyone else having problems win the tsle anki file?

We believe so partially because Last Slot, COLL, EPLL takes 3 Looks, opposed to TSLE, TTLL, requiring only 2 Looks.
I'm also pretty sure the average movecount for ZZ-CT is lower than COLL/EPLL.

There are times when you get bad cases, and so it is good to know another Subset so that it can back you up so that the movecount doesn't go crazy high.

 

[Hammer]

Could alphasheep please post the tsle spreadsheet again?

 

[GenTheSnail]

Could alphasheep please post the tsle spreadsheet again?

For anyone to lazy to go back and find it...

Have we really gotten so lazy we cant look at the top of this page for it ourselves?
2nd page, 5th or so post from the top was where alphasheep originally posted his spreadsheet.

 

[PhillipEspinoza]

When you insert the F2L Edge in, it just becomes a more efficient EJLS (Or EJF2L), where the F2L Corner's Position isn't necessarily at DFR.

TSLE Skips are more likely, of course because the cases are reduced and AUF's don't count due to symmetry reasons. On the Algorithm site there are 3 algorithms requiring 4 Triggers to solve Corner Orientation when the Edge is inserted. Also, from Chris Tran's Seminar, he said that part of the thing with ZBLL Algorithms is that they just break up the F2L, and Restore it. What you're trying to do is Break up the Edge, and Restore it with Solved Corner Orientation.

There isn't really a more likely chance for the simple cases if you don't pay attention to Corner Orientation when placing the Edge. Part of the reason CLS isn't favored as much as TSLE is because the overall Movecount is lower on average, because only 1 piece is being solved, and when a CLS Case comes when there is an Edge and F2L Corner in the correct slot, you are just breaking up the Corner along with the Edge. The normal F2L Cases for those 2 without caring about Corner Orientation are one of the worst ones when EO is solved.

The Movecount from all angles really depends on how you Blockbuild, efficiency-wise, but you still get a ton of choices to choose from when Blockbuilding for ZZF2L. It's very unlikely that after you solve EOLine, Square to build are all bad cases Except only the Square at DFR. Even if that does happen, if you can see it in inspection, just solve EOLine with a (y2) beforehand so the good case is on BLD (Strange name for a Corner Position).

If you look carefully at Corner Orientation, all the cases are just repeated 5 times, with the F2L Edge in different positions in the U-Layer or in the slot.

Say you get the Scramble for TSLE as: (R U R' U2 R U2 R')
Doing it the way you suggested, you would do (R U' R').
The algorithm for that Edge inserted case is (R U' R' U2 R U' R'), making TSLE 10 Moves.
Standard TSLE would just be the inverse, which is only 7 moves compared to 10.

Dunno if you know this but EJF2L where the corner isn't in DFR is called CLS. Just an fyi there. So it would be an efficient CLS sub set for TSLE cases where edge is inserted. It's not looking to connect that inserted edge piece with the specific corner piece in the DFR while orienting that corner, which is what EJF2L aims to do, it's looking to pair it up with any corner and insert while orienting all corners. This is why you could use a variety of different CLS cases for any one of the TSLE cases where edge is inserted. EJF2L is a subset of CLS where you have LS edge and corner permuted.

You're failing to understand what I'm suggesting. I'm not suggesting a beginners method insertion of LSE. In your example, my variation would require the 7 move TSLE alg you mentioned where edge is inserted and that's what you would see when you ended up at LS. The goal would be to insert edge while block building, making a 1x1x2 + edge block for either left or right block. Then you get 1/23 cases (remember we have the 7 cross cases which can be triggered pretty easily).

So here are the steps
1) EO Line
2) 1x2x3 block
3) 1x2x2 block + edge
4) CLS/TSLE (1/23 cases)
5) TTLL or PLL

Another way

1) EO Line
2) 1x1x2 block + edge
3) 1x2x3 block
4) CLS/TSLE (1/23 cases)
5) TTLL or PLL

There's no individual step for ELS, it's combined with block building. This will sometimes make the move count during block building lower as you just have to insert an edge as opposed to pairing it up with a corner.

If you combine this with learning to force PLL after TSLE, in other words if you learn CLS, you bring the LL skip probability down to 1/72 from 1/360.

 

[JTWong71]

Dunno if you know this but EJF2L where the corner isn't in DFR is called CLS. Just an fyi there. So it would be an efficient CLS sub set for TSLE cases where edge is inserted. It's not looking to connect that inserted edge piece with the specific corner piece in the DFR while orienting that corner, which is what EJF2L aims to do, it's looking to pair it up with any corner and insert while orienting all corners. This is why you could use a variety of different CLS cases for any one of the TSLE cases where edge is inserted. EJF2L is a subset of CLS where you have LS edge and corner permuted.

You're failing to understand what I'm suggesting. I'm not suggesting a beginners method insertion of LSE. In your example, my variation would require the 7 move TSLE alg you mentioned where edge is inserted and that's what you would see when you ended up at LS. The goal would be to insert edge while block building, making a 1x1x2 + edge block for either left or right block. Then you get 1/23 cases (remember we have the 7 cross cases which can be triggered pretty easily).

So here are the steps
1) EO Line
2) 1x2x3 block
3) 1x2x2 block + edge
4) CLS/TSLE (1/23 cases)
5) TTLL or PLL

Another way

1) EO Line
2) 1x1x2 block + edge
3) 1x2x3 block
4) CLS/TSLE (1/23 cases)
5) TTLL or PLL

There's no individual step for ELS, it's combined with block building. This will sometimes make the move count during block building lower as you just have to insert an edge as opposed to pairing it up with a corner.

If you combine this with learning to force PLL after TSLE, in other words if you learn CLS, you bring the LL skip probability down to 1/72 from 1/300+.

I do know what the subsets are, I'm saying that a way to think about what you are doing is a more efficient EJF2L, by not caring about the DFR Corner's Permutation.

And about the skips you were talking about if you insert the Edge in the middle of the solve. If the Adjacent Corner from the Inserted Edge isn't Oriented, it isn't possible to skip TSLE unless you mess with the slot during some other part of F2L because the Corner will stay misoriented, until your Edge inserted TSLE. You really seem to want to keep the algorithm count very low, and really if you are going to insert the Edge in the middle of the solve, you might as well just use CLS so that you always get a PLL.

 

[gyroninja]

Could alphasheep please post the tsle spreadsheet again?

All the algs are on my site.

http://gyroninja.net/zzct/zzct-tsle.html

 

[PhillipEspinoza]

I do know what the subsets are, I'm saying that a way to think about what you are doing is a more efficient EJF2L, by not caring about the DFR Corner's Permutation.

And about the skips you were talking about if you insert the Edge in the middle of the solve. If the Adjacent Corner from the Inserted Edge isn't Oriented, it isn't possible to skip TSLE unless you mess with the slot during some other part of F2L because the Corner will stay misoriented, until your Edge inserted TSLE. You really seem to want to keep the algorithm count very low, and really if you are going to insert the Edge in the middle of the solve, you might as well just use CLS so that you always get a PLL.

Wow that's actually a brilliant point. If you could do corner control while looking for that free edge, you'll not only increase your chances of a TSLE skip but you'll minimize the chances of the nasty cases and just have cross cases for TSLE right? That would bring your TSLE skip down to 1/27. Wow.

This would create a forced LL skip to happen 1/27 times?!

 

[JTWong71]

Wow that's actually a brilliant point. If you could do corner control while looking for that free edge, you'll not only increase your chances of a TSLE skip but you'll minimize the chances of the nasty cases and just have cross cases for TSLE right? That would bring your TSLE skip down to 1/27. Wow.

If you put it that way, yes, but that just turns TSLE into OCLL, and then TTLL.

 

[PhillipEspinoza]

OK as crazy as it sounds I have just discovered a full LL skip method that only requires 72 TTLL algs and 28 CLS algs. Anyone interested in hearing it? It might be ridiculous but it's valid.

 

[wir3sandfir3s]

OK as crazy as it sounds I have just discovered a full LL skip method that only requires 72 TTLL algs and 28 CLS algs. Anyone interested in hearing it? It might be ridiculous but it's valid.

Yes. Of course. Go.

 

[PhillipEspinoza]

Brief description and remember the disclaimer, might be ridiculous but it's legit.

1) EO Line
2) 1x2x3 block (either left or right)
3) L3E - Last 3 Edges (build block of edges paired with oriented bottom layer corners *4 times as many corners to choose from should make pairing easier*)
4) CLS - for back most slot, should be 1/28 cases to identify via position of white/top layer sticker combined with easy OCLL cases
5) TTLL

Steps 2 and 3 are chronologically interchangeable. You can build the L3E block first if it's easier.

The hardest and most controversial thing about this idea is the block building with bottom layer corners. I suck at block building but something tells me though hard to retrain your eye at first, there should be no reason you couldn't master this new way of block building. Thoughts or questions?

EDIT: Sometimes during building of L3E it will be most convenient to form a "true pair" in which case you will still do CLS in step 4 but step 5 will just be PLL.

 

[Hammer]

All the algs are on my site.

http://gyroninja.net/zzct/zzct-tsle.html

Oh thanks

 

[Tao Yu]

Tip for people learning TTLL: Try reversing algs that you like and seeing what cases the reversed algs solve. You'll get some nice algs sometimes.

Some algs I found by reversing:

U' R U2' R' U R U2' R D R' U2 R D' R' U R' (mirror is good too)
R' D' R U' R' D R U R U' R' U R U R' U R U' R'
U' R' U' R' U R2 U2' R' U2' R' U' R U R

Some other algs I found by combining WV algs with TSLE algs

R U2' R' U2 R U2' R2' F R U R U' R' F'
F R U R' U' R' F' R2 U2' R' U2 R U2' R' (inverse of above)
U' R U' R' U R U' R2' F2 R2 U' R' U' R U R' F2

 

[JTWong71]

Brief description and remember the disclaimer, might be ridiculous but it's legit.

1) EO Line
2) 1x2x3 block (either left or right)
3) L3E - Last 3 Edges (build block of edges paired with oriented bottom layer corners *4 times as many corners to choose from should make pairing easier*)
4) CLS - for back most slot, should be 1/28 cases to identify via position of white/top layer sticker combined with easy OCLL cases
5) TTLL

Steps 2 and 3 are chronologically interchangeable. You can build the L3E block first if it's easier.

The hardest and most controversial thing about this idea is the block building with bottom layer corners. I suck at block building but something tells me though hard to retrain your eye at first, there should be no reason you couldn't master this new way of block building. Thoughts or questions?

EDIT: Sometimes during building of L3E it will be most convenient to form a "true pair" in which case you will still do CLS in step 4 but step 5 will just be PLL.

So is this what you mean?

Scramble: D2 R D2 R2 D2 R' D2 R' U2 L R2 U B' L D2 R D' B' F' U2

x2 //Inspection
R F' U R B' D' //EO + BL Square
L D' L' //XEOLine
F2 R' (U' D) R2 D' R2 //Oriented F2L + FR Square
U2 L' U L U2 L' U2 L U2 L' U L //Orient Corners + Permute 1 Corner
z R D2 L U2 L' D2 L U2 (L' R') //TTLL

36 STM

 

[wir3sandfir3s]

Brief description and remember the disclaimer, might be ridiculous but it's legit.

1) EO Line
2) 1x2x3 block (either left or right)
3) L3E - Last 3 Edges (build block of edges paired with oriented bottom layer corners *4 times as many corners to choose from should make pairing easier*)
4) CLS - for back most slot, should be 1/28 cases to identify via position of white/top layer sticker combined with easy OCLL cases
5) TTLL

Steps 2 and 3 are chronologically interchangeable. You can build the L3E block first if it's easier.

The hardest and most controversial thing about this idea is the block building with bottom layer corners. I suck at block building but something tells me though hard to retrain your eye at first, there should be no reason you couldn't master this new way of block building. Thoughts or questions?

EDIT: Sometimes during building of L3E it will be most convenient to form a "true pair" in which case you will still do CLS in step 4 but step 5 will just be PLL.

That's actually not too bad. I like it. Maybe do a D move so the back corner is in front? Or we can get Algs when the back corner we are inserting is in the front.
However, one might argue:
"If you are creating two pseudo pairs, why not just solve F2L-1 slot, create the pseudo pair and insert with wv?"
Only one less alg, but I think your idea is nicer. Plus, you have freedom to pick the slot you can do CLS on, and you can possibly avoid a bad case.

 

[JTWong71]

That's actually not too bad. I like it. Maybe do a D move so the back corner is in front? Or we can get Algs when the back corner we are inserting is in the front.
However, one might argue:
"If you are creating two pseudo pairs, why not just solve F2L-1 slot, create the pseudo pair and insert with wv?"
Only one less alg, but I think your idea is nicer.

So basically you don't have to build a 1x2x3 Block first, as it is likely that a random Corner is paired up nicely with an F2L Edge earlier in the solve, and it can be placed anywhere on the cube, and so yeah, you just offset the D-Layer to solve the Corner and Orientation. For me I think the only problem would be Look ahead because if you try to look for the most efficient solution between all the Corners, it'll take some time to check.