The ZZVolution.

[zzcuberman]

I think your wrong. The ergonomics are that much better. RUL as a move set sucks. RUD is better and F turning are fine in many cases. Yes CFOP has rotations. It is hard to say how much of a difference in move sets ZZ has from CFOP but I feel you are making out it is way more than it is.

But here is something what about when pdeuso slotting and F RU F' stuff is included. These techniques give way more options for f2l. What does zz with ZBLL was in this regard?


Also lookahead isn't necessary that much easier. All you need to do is not look at the current pieces your solving but the next ones you are going to solve. Hardness of this is based off how good you are at lookahead.

RUL is great. Your hating in a zz thread were people will defend it

 

[ProStar]

RUL is great. Your hating in a zz thread were people will defend it

EOCross's RUL is good. EOLine's RUL is pretty bad

 

[tx789]

RUL is great. Your hating in a zz thread were people will defend it

R UL sucks cause of all the regrips caused by switching between R and L. RUD is way better.

 

[ProStar]

R UL sucks cause of all the regrips caused by switching between R and L. RUD is way better.

If you use EOCross, then the regrips aren't bad at all. Besides, CFOP is almost never RUD for F2L, only when you're doing a keyhole insert. CFOP F2L is RUFLy, so it could be debated that ZZ ergonomics are better since it's just RUL

 

[Nmile7300]

Let's move this to the method debate thread, please.

 

[Owen Morrison]

RUL is very fast. ZZ with eocross is regripless. Also, VLS + PLL might be similar to ZZ pair + ZBLL, but you have to make the pair first. It saves 3 moves and most people consider ZBLL leaps and bounds better my guy. Saves around 8-9.

Please send me a video of someone or you doing a ZZ solve with 0 regrips

 

[ProStar]

Please send me a video of someone or you doing a ZZ solve with 0 regrips

You couldn't do a CFOP solve with 0 regrips without it being bad

 

[Owen Morrison]

You couldn't do a CFOP solve with 0 regrips without it being bad

I don't see how that is related to what I am talking about.

@zzcuberman is trying to say that ZZ is better than CFOP because it has 0 regrips which is obviously not true.

 

[ProStar]

I don't see how that is related to what I am talking about.

@zzcuberman is trying to say that ZZ is better than CFOP because it has 0 regrips which is obviously not true.

ZZ definitely has regrips, but CFOP also has them. Amount of regrips doesn't make ZZ better than CFOP, but it also doesn't make ZZ worse than CFOP.

 

[Owen Morrison]

ZZ definitely has regrips, but CFOP also has them. Amount of regrips doesn't make ZZ better than CFOP, but it also doesn't make ZZ worse than CFOP.

You are still totally misunderstanding what I am saying, I am not trying to say one method is better or not because of regrips, I am trying to say that zzcuberman's statement about zz having no regrips is wrong.

 

[Nmile7300]

Let's move this to the method debate thread, please.

 

[I'm A Cuber]

do a BLDy-gen solve and get the same times as you normally do and I'll believe you

If there was a BLDy-gen method that had a decent movecount, I would
Edit: wait do you mean a gen similar to moves used in 3bld or one using back, left, and down moves?

 

[TerryD]

If there was a BLDy-gen method that had a decent movecount, I would
Edit: wait do you mean a gen similar to moves used in 3bld or one using back, left, and down moves?

I meant a method using only B, L, and D moves.

 

[OreKehStrah]

The main problem with ZZ lies in the inspection and EO phases. Because you have to plan more during inspection just to do your EOCross, it makes it harder to consistently plan your cross. This snowballs into a bigger problem when you try to plan a first pair alongside the EOCross since you have even less time to plan and have to work around your EO solution.

That being said, F2L and LL are definitely nicer and there are a lot of really cool LS and LL things you can do because of EO, although whether or not some of those things are practical or not can be it's own debate. I personally use CFOP for my main method, and ZZ for fun and to practice/train my TPS.

There really isn't enough data for a truly compelling argument for planning EOCross plus one vs Cross plus two, simply do to the low sample size. Most people never give any thought to ZZ and go straight to CFOP. In an theoretical world where say, all the top 100 CFOP solvers all decided to spend a couple of years practicing EOCross, you could then really start to compare EOCross ZZ vs CFOP.

 

[zzcuberman]

I don't see how that is related to what I am talking about.

@zzcuberman is trying to say that ZZ is better than CFOP because it has 0 regrips which is obviously not true.

i never once said ZZ is better then CFOP. Im trying to say they are equal. Now quit bashing ZZ in a ZZ Thread

 

[tx789]

If you use EOCross, then the regrips aren't bad at all. Besides, CFOP is almost never RUD for F2L, only when you're doing a keyhole insert. CFOP F2L is RUFLy, so it could be debated that ZZ ergonomics are better since it's just RUL

RUD-gen is a better 3 move set of moves. F2L can have also no D moves if you want but adding them is very efficient and pdueso slotting and F RU F' gives F2L way more options while remaining very ergonomic and uses D turns.

For RUFLy vs RUL yes RUL has less moves but F turns can be fine and there are ways to minimise rotations a bit. People here are making out like it is so so much better but the difference is small. This depends on how good your turning is of course. Fancy finger tricks are harder and have more risk but give you more options.

 

[ottozing]

i never once said ZZ is better then CFOP. Im trying to say they are equal. Now quit bashing ZZ in a ZZ Thread

They're not equal, and it's not even close!

Spoiler: long wall of text that's going to give ZZ lovers an existential crisis

ZZ has been around and well known by cubers for more than a decade, and for a long time had similar popularity levels to Roux since at the time, 3x3 hardware made M slices bad and RUL good

If ZZ were equal, the best CFOP solvers in the world wouldn't be more than 2 seconds faster on average than the best ZZ solvers we've seen (ZZ WR official average is 7.6x with EO cross variant, and I don't think we've seen anything faster than a mid 6 avg5 as far as unofficial results are concerned, which is again 2 seconds slower than the best CFOP averages)

I also think that the EOcross variant of ZZ isn't even really ZZ, since every step after EO is exactly the same as the CFOP steps with some advanced LL stuff (which we already see in CFOP) plus an F2L state that can be solved RUL gen (which isn't even optimal because of cases like R D' R' U R D R' and possibly setups to it), but this is a discussion for another time

If you or anyone else wants to prove that ZZ is on the same level as CFOP, then the actual core steps of the method need to fundamentally change, because EO>Line>RUL blockbuilding>ZBLL has obviously failed to compete with CFOP (though I realize that most ZZ users have figured this out and started pursuing EO+Cross)

EO>Cross>F2L>ZBLL also doesn't seem like it would be enough based on a few things

1st, obviously you want to do a mix of EO and Cross at the same time, but on a practical level they're going to be mostly separate steps the majority of the time, so I don't think it's a stretch to say that you're essentially making EO your first step for this method

(If any ZZ solvers want to prove me wrong on this and show me say, an average of 12 where you're consistently doing two or more cross edges before EO is done, then please do because if this assumption is wrong then everything else I'm about to write is at least partially wrong)

With that assumption in mind, you absolutely have to ask the question "is adding this additional step of solving EO worth it?"

The way you work that out is simple (in theory at least). You have to see how long it takes to solve EO time wise on average, and from there, see how much time you save with every other step (those steps being Cross, F2L, and LL)

Cross I would have to believe is worse to be honest, because #1 you're restricted to R,L,U,D,F2,B2 instead of R,L,U,D,F,B and #2 your capacity to plan a good Cross from a fingertrick/efficiency standpoint is never going to be as good as a CFOP solvers Cross

This isn't up for debate based on my original assumption... It's not a case of it being "harder", it's a case of it literally being worse, because you're trying to plan a Cross on a cube that's scrambled while tracking the 4-5 moves of your EO in your head on top of it (all while being restricted in what moves you can and can't do)

Trying to do CN ZZ only makes this problem worse... Technically on one EO axis you have 4 cross options if you do z/z'/z2 or no rotation after EO, but then you have to check all 4 after committing to an EO axis, and if you find a nice cross that requires a z2, you also have to ideally rotate and change your EO solution so that your solve start isn't "solve EO>z rotation>solve Cross"

If you want to say that hypothetically it's possible to plan EOCross+1 every solve, then I'm going to say that it's hypothetically possible to plan Cross+2 with CFOP every solve. Inspection wise, you're always going to be behind CFOP by at least one "step" (counting each F2L pair as a step) if it's truly a level playing field (I would argue more but I want to give y'all some breathing room)

Based on all of the reasonable theory-crafting I can do, I'm going to go ahead and say that EO first makes Cross a lot harder since once you commit to an EO axis, you're down to 4 possible Crosses instead of 6, you can't use F or B quarter turns, and if you're picking the nicest cross out of the 4, on average, you're going to have to adjust your EO solution 3/4 times if you want to be optimal and avoid doing a z/z'/z2 between steps

Good luck planning a first pair on top of that in 15 seconds (really more like 12 because everyone panic starts on the 8 or 12 warning)

With that now in mind, I think the next question to ask is "which of the two post-inspection scenarios is better?"

a) A Cross solved with all of your edges oriented for F2L/LL
b) A Cross solved with your first pair also solved

This wall of text is already getting very long, but I'm going to go ahead and assume that anyone who's gotten decently good with both CFOP and ZZ would agree that B is a lot nicer

People have already said a bunch about move-group stuff in this thread, but I'll go ahead and say that there's a reason most people don't use RULgen algs for PLL, even though virtually every adj/diag swap PLL has an RULgen alg that's the same movecount (or better) than the RUF/RUD equivilents (a good example is the R U2 R' U' R U2 L' U R' U' L Jperm)

It's because RUL is worse than RUF, RUD, and a variety of other move-groups. Again, RULgen PLL algs exist, and I'm fairly certain all of them are very close to the optimal FMC movecount, so if RUL was a good move-group, all of the fastest cubers would be using these algs since low movecount+good move-group is a complete no brainer

Finally, is ZBLL better than OLL>PLL? Probably, but not by more than half a second in my honest opinion

The only full ZBLL sets that I can safely say have been proven worthwhile are U, T, and L

Sune and Antisune have the worst CP recognition of the 7 sets, along with the fastest normal OLL's of the 7 sets. This makes ZBLL for these OLL's very hard to justify, even though on paper the algs are actually quite good (especially for Sune)

H and Pi have a unique issue where they have the easiest CP recognition (though only slightly better than U/T/L IMO), but also the worst ZBLL recognition due to the lack of any oriented corner

At least with Sune/Antisune, you can use a 1x2x2 oriented square to recognize things somewhat nicely, but with H/Pi you have to be a lot more creative with how you recognize the cases

I've talked with the best ZB users in the world, and even they agree that they only think parts of H/Pi are worth it for ZB, while some COLL sets aren't worth it

On the other hand, even though I don't have the best TPS or even turning, I've been able to get most PLL's sub 1 framecounted (can probably do the last two which are Fperm and one of the Nperms but I'm lazy), and almost every OLL sub .8 framecounted (minus an OLL where I do FRURUF>T ZBLL in actual solves), and with ROLL/JOLL recog tricks, I know that without learning anything new I can probably make my LL 2.5ish on average even if I stopped using ZBLL stuff/1LLL stuff/VLS stuff entirely

Still, I think all edges oriented last layer (aka a mix of ZB and OLL>PLL) when mastered would probably be low or sub 2, so this is the one area where I think EO as a first step could make a half second difference in theory

Bottom line, EO makes Cross worse in the sense that you're one step behind a CFOP solver inspection wise (if not more, I personally think it's closer to 2 than 1), but in theory makes LL half a second faster

Is RULgen 4 pairs is better than random-gen 3 pairs? Considering fast cubers can do a y2 in sub .4 framecounted (and y/y' a lot faster), along with the fact that good CFOP solvers won't rotate more than twice in a solve via rotationless F2L tricks like f R f' stuff, r U' R' U R U r', and a billion other neat tricks I could list, I highly highly doubt it

ZZ is a lot worse than Roux & CFOP, and that's OK

 

[Username: Username:]

They're not equal, and it's not even close!

Spoiler: long wall of text that's going to give ZZ lovers an existential crisis

ZZ has been around and well known by cubers for more than a decade, and for a long time had similar popularity levels to Roux since at the time, 3x3 hardware made M slices bad and RUL good

If ZZ were equal, the best CFOP solvers in the world wouldn't be more than 2 seconds faster on average than the best ZZ solvers we've seen (ZZ WR official average is 7.6x with EO cross variant, and I don't think we've seen anything faster than a mid 6 avg5 as far as unofficial results are concerned, which is again 2 seconds slower than the best CFOP averages)

I also think that the EOcross variant of ZZ isn't even really ZZ, since every step after EO is exactly the same as the CFOP steps with some advanced LL stuff (which we already see in CFOP) plus an F2L state that can be solved RUL gen (which isn't even optimal because of cases like R D' R' U R D R' and possibly setups to it), but this is a discussion for another time

If you or anyone else wants to prove that ZZ is on the same level as CFOP, then the actual core steps of the method need to fundamentally change, because EO>Line>RUL blockbuilding>ZBLL has obviously failed to compete with CFOP (though I realize that most ZZ users have figured this out and started pursuing EO+Cross)

EO>Cross>F2L>ZBLL also doesn't seem like it would be enough based on a few things

1st, obviously you want to do a mix of EO and Cross at the same time, but on a practical level they're going to be mostly separate steps the majority of the time, so I don't think it's a stretch to say that you're essentially making EO your first step for this method

(If any ZZ solvers want to prove me wrong on this and show me say, an average of 12 where you're consistently doing two or more cross edges before EO is done, then please do because if this assumption is wrong then everything else I'm about to write is at least partially wrong)

With that assumption in mind, you absolutely have to ask the question "is adding this additional step of solving EO worth it?"

The way you work that out is simple (in theory at least). You have to see how long it takes to solve EO time wise on average, and from there, see how much time you save with every other step (those steps being Cross, F2L, and LL)

Cross I would have to believe is worse to be honest, because #1 you're restricted to R,L,U,D,F2,B2 instead of R,L,U,D,F,B and #2 your capacity to plan a good Cross from a fingertrick/efficiency standpoint is never going to be as good as a CFOP solvers Cross

This isn't up for debate based on my original assumption... It's not a case of it being "harder", it's a case of it literally being worse, because you're trying to plan a Cross on a cube that's scrambled while tracking the 4-5 moves of your EO in your head on top of it (all while being restricted in what moves you can and can't do)

Trying to do CN ZZ only makes this problem worse... Technically on one EO axis you have 4 cross options if you do z/z'/z2 or no rotation after EO, but then you have to check all 4 after committing to an EO axis, and if you find a nice cross that requires a z2, you also have to ideally rotate and change your EO solution so that your solve start isn't "solve EO>z rotation>solve Cross"

If you want to say that hypothetically it's possible to plan EOCross+1 every solve, then I'm going to say that it's hypothetically possible to plan Cross+2 with CFOP every solve. Inspection wise, you're always going to be behind CFOP by at least one "step" (counting each F2L pair as a step) if it's truly a level playing field (I would argue more but I want to give y'all some breathing room)

Based on all of the reasonable theory-crafting I can do, I'm going to go ahead and say that EO first makes Cross a lot harder since once you commit to an EO axis, you're down to 4 possible Crosses instead of 6, you can't use F or B quarter turns, and if you're picking the nicest cross out of the 4, on average, you're going to have to adjust your EO solution 3/4 times if you want to be optimal and avoid doing a z/z'/z2 between steps

Good luck planning a first pair on top of that in 15 seconds (really more like 12 because everyone panic starts on the 8 or 12 warning)

With that now in mind, I think the next question to ask is "which of the two post-inspection scenarios is better?"

a) A Cross solved with all of your edges oriented for F2L/LL
b) A Cross solved with your first pair also solved

This wall of text is already getting very long, but I'm going to go ahead and assume that anyone who's gotten decently good with both CFOP and ZZ would agree that B is a lot nicer

People have already said a bunch about move-group stuff in this thread, but I'll go ahead and say that there's a reason most people don't use RULgen algs for PLL, even though virtually every adj/diag swap PLL has an RULgen alg that's the same movecount (or better) than the RUF/RUD equivilents (a good example is the R U2 R' U' R U2 L' U R' U' L Jperm)

It's because RUL is worse than RUF, RUD, and a variety of other move-groups. Again, RULgen PLL algs exist, and I'm fairly certain all of them are very close to the optimal FMC movecount, so if RUL was a good move-group, all of the fastest cubers would be using these algs since low movecount+good move-group is a complete no brainer

Finally, is ZBLL better than OLL>PLL? Probably, but not by more than half a second in my honest opinion

The only full ZBLL sets that I can safely say have been proven worthwhile are U, T, and L

Sune and Antisune have the worst CP recognition of the 7 sets, along with the fastest normal OLL's of the 7 sets. This makes ZBLL for these OLL's very hard to justify, even though on paper the algs are actually quite good (especially for Sune)

H and Pi have a unique issue where they have the easiest CP recognition (though only slightly better than U/T/L IMO), but also the worst ZBLL recognition due to the lack of any oriented corner

At least with Sune/Antisune, you can use a 1x2x2 oriented square to recognize things somewhat nicely, but with H/Pi you have to be a lot more creative with how you recognize the cases

I've talked with the best ZB users in the world, and even they agree that they only think parts of H/Pi are worth it for ZB, while some COLL sets aren't worth it

On the other hand, even though I don't have the best TPS or even turning, I've been able to get most PLL's sub 1 framecounted (can probably do the last two which are Fperm and one of the Nperms but I'm lazy), and almost every OLL sub .8 framecounted (minus an OLL where I do FRURUF>T ZBLL in actual solves), and with ROLL/JOLL recog tricks, I know that without learning anything new I can probably make my LL 2.5ish on average even if I stopped using ZBLL stuff/1LLL stuff/VLS stuff entirely

Still, I think all edges oriented last layer (aka a mix of ZB and OLL>PLL) when mastered would probably be low or sub 2, so this is the one area where I think EO as a first step could make a half second difference in theory

Bottom line, EO makes Cross worse in the sense that you're one step behind a CFOP solver inspection wise (if not more, I personally think it's closer to 2 than 1), but in theory makes LL half a second faster

Is RULgen 4 pairs is better than random-gen 3 pairs? Considering fast cubers can do a y2 in sub .4 framecounted (and y/y' a lot faster), along with the fact that good CFOP solvers won't rotate more than twice in a solve via rotationless F2L tricks like f R f' stuff, r U' R' U R U r', and a billion other neat tricks I could list, I highly highly doubt it

ZZ is a lot worse than Roux & CFOP, and that's OK

Jay, you did a good job destroying ZZ lol and that's also ok.

 

[mukerflap]

Please send me a video of someone or you doing a ZZ solve with 0 regrips

obvious eo cross has regrips but in his f2l he stays in home grip, i think thats what he meant to say

RUD-gen is a better 3 move set of moves. F2L can have also no D moves if you want but adding them is very efficient and pdueso slotting and F RU F' gives F2L way more options while remaining very ergonomic and uses D turns.

For RUFLy vs RUL yes RUL has less moves but F turns can be fine and there are ways to minimise rotations a bit. People here are making out like it is so so much better but the difference is small. This depends on how good your turning is of course. Fancy finger tricks are harder and have more risk but give you more options.

pseudo slotting is actually much easier in ZZ beacuse you dont have to worry about the EO

 

[zzcuberman]

obvious eo cross has regrips but in his f2l he stays in home grip, i think thats what he meant to say

pseudo slotting is actually much easier in ZZ beacuse you dont have to worry about the EO

yes thats what i meant, obviously lol these guys are dense.