• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 40,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

ZZLS+1 proof of concept

ottozing

Platinum Member
Joined
Aug 13, 2011
Messages
3,260
Location
Canberra, Australia
WCA
2012MCNE01
YouTube
Visit Channel
Generally speaking, people who use full ZBLL agree that for ZBLL cases, having 2 oriented corners is the best while having 0 oriented corners is the worst (this is accounting for recognition, quality of algs, and how nice OLL>PLL is in comparison)

What I've gone ahead and done is taken every possible LS case you can get with ZZ for the FR slot & found one alternative algorithm that can be used to guarantee that you never get 0 oriented corners for LL. Not all of these are actually worthwhile, but many of them are & the recognition demands really aren't that bad once you get used to them (won't be explaining that part here)

This idea can also apply to ZBLS, and in my opinion will probably become the primary approach to LSLL for advanced CFOP solvers, with the secondary approach being ZBLS, and the tertiary approach being some mix of other stuff like VLS, true OLS, and forcing very nice OLL cases in terms of move count/look ahead into PLL/good alternative OLLCP's

Bolded = worth it
Underlined = almost worth it but needs work/unsure
No bold/no underline = not worth it

F2L 1 - U R U' R' & U2 R U2' R'
F2L 4 - R U R' & R U' R' U' R U' R'
F2L 5 - Reduce to F2L 1
F2L 7 - Reduce to F2L 1
F2L 10 - U2 R U' R' U' R U R' & U' R U R' U R U R'
F2L 12 - R' U2' R2 U R2' U R & R U' R' U R U' R' U2 R U' R' (can also just reduce to F2L 1 with the 2nd or F2L 2 with the 1st)

F2L 14 - U' R U' R' U R U R' & U2' R2 U R' U R U2' R2' (can also reduce to F2L 4 with the 1st)
F2L 15 - Reduce to F2L 1
F2L 17 - Reduce to F2L 4 (could also use R U2' R' U' R U2' R' U R U2' R' as the alt instead of R U2' R' U' R U' R' U' R U' R' which has nicer U layer turns and allows some look ahead into ZBLL since it's got similar properties to R' D' R <U> R' D R)
F2L 19 - Reduce to F2L 1
F2L 21 - Reduce to F2L 1
F2L 23 - Reduce to F2L 1

F2L 25 - Reduce to F2L 1 (U' F' R U R' U' R' F R & S' R U' R' f R' F' R covers all but 1 case)
F2L 27 - Reduce to F2L 1
F2L 30 - Reduce to F2L 4 (R U R' U' R U R' & U' R U2' R' U2 R U R' covers all but 1 case, can also technically do F R' F' R into F2L 2 but it's kinda gross)
F2L 32 - Reduce to F2L 1
F2L 33 - Reduce to F2L 1
F2L 34 - U R' D' R U' R' D R & U R U R' U2' R U R' (can also do D R' U R to reduce to F2L 2)
F2L 38 - R U R' U' R U2 R' U' R U R' or R U' R' U' R U R' U' R U2' R' with the correct pre-AUF (or reduce to F2L 1)
F2L 39 - R U' R' U R U2' R' U R U' R' with the correct pre-AUF (or reduce to F2L 1)


In total, 15/20 are worth it using this system, and of the remaining 5, the only one that definitely isn't worth it in my opinion is F2L 4 aka split free pair. I'd say suck it up for this case and take H/Pi unless you can cancel nicely into the standard OLL (happens for 3 of the 6 cases), but even then it's debatable and I'm very much inclined to think you should always take H/Pi here

Reducing to F2L 4 as such isn't anywhere near as good as reducing to F2L 1, so for that reason I don't like F2L 17 with this system since it only has one way to reduce to a free pair in 3-4 moves (whereas F2L 30 at least has a 2nd way that removes the H/Pi for all but one case). I think that for the 6 cases where the standard insert gives H/Pi, you're better off taking OLL unless there's a good 11 mover (ignoring pre-AUF) for the pure OLS (ones that I found by hand: R U2' R' U' R U R' U2 R U' R' and R U R' U' R U2' R' U2 R U R')

F2L 14 is interesting to me. Even though it has two algs, the secondary alg is 3QTM worse and can't be done from home grip. That said, it's basically a sune with an extra R turn, so I'm more inclined to think it's worth it even if it's a bit weird

F2L 25 I'm just kinda sad about lol. I tried a bunch of combos of F' R U R' U' R' F R + inverse & S' R U' R' f R' F' R + inverse but couldn't get 100% with only 2 algs. To get 100% you need to use U' R' F R f' R U R' S and then you get U ZBLL from the U2 angle (or maybe there's a good BLE alg)

Lastly, for the F2L 30 OLS that always gives 4 misoriented corners using the 2 good solutions, you can either do the 7QTM one and get H (which is still nicer than Pi) do F R' F' R d' L' U' L U L' U L (which kinda sucks tbh), or do U' R U2' L U' R' U L' to get T ZBLL from the U2 angle (also kinda sucks)

All in all, using the most conservative version of this system (aka not using gross F2L algs, only 2 per case, and no fancy OLS tricks like WV), you're forced to take H/Pi for 14 out of the 540 possible OLS cases (not sure what the actual stats are on this but I know that just dividing them is incorrect because symmetry or something)

This seems pretty damn solid to me for a start considering that with 0 influence, you'd be getting H/Pi for 120/540 cases

That said, if you were only to use reduction to F2L 1 tricks (excluding F2L 34), the final number of cases where you're stuck getting H or Pi becomes 42/540. This is important because reducing to F2L 1 aka a built pair means you only need to learn one recog system, and you can also do WV on top of that for some cases. This is absolutely the place to start if you're a ZZ solver looking to up your OLS game B)
 

OreKehStrah

Member
Joined
May 24, 2019
Messages
686
Generally speaking, people who use full ZBLL agree that for ZBLL cases, having 2 oriented corners is the best while having 0 oriented corners is the worst (this is accounting for recognition, quality of algs, and how nice OLL>PLL is in comparison)

What I've gone ahead and done is taken every possible LS case you can get with ZZ for the FR slot & found one alternative algorithm that can be used to guarantee that you never get 0 oriented corners for LL. Not all of these are actually worthwhile, but many of them are & the recognition demands really aren't that bad once you get used to them (won't be explaining that part here)

This idea can also apply to ZBLS, and in my opinion will probably become the primary approach to LSLL for advanced CFOP solvers, with the secondary approach being ZBLS, and the tertiary approach being some mix of other stuff like VLS, true OLS, and forcing very nice OLL cases in terms of move count/look ahead into PLL/good alternative OLLCP's

Bolded = worth it
Underlined = almost worth it but needs work/unsure
No bold/no underline = not worth it

F2L 1 - U R U' R' & U2 R U2' R'
F2L 4 - R U R' & R U' R' U' R U' R'
F2L 5 - Reduce to F2L 1
F2L 7 - Reduce to F2L 1
F2L 10 - U2 R U' R' U' R U R' & U' R U R' U R U R'
F2L 12 - R' U2' R2 U R2' U R & R U' R' U R U' R' U2 R U' R' (can also just reduce to F2L 1 with the 2nd or F2L 2 with the 1st)

F2L 14 - U' R U' R' U R U R' & U2' R2 U R' U R U2' R2' (can also reduce to F2L 4 with the 1st)
F2L 15 - Reduce to F2L 1
F2L 17 - Reduce to F2L 4 (could also use R U2' R' U' R U2' R' U R U2' R' as the alt instead of R U2' R' U' R U' R' U' R U' R' which has nicer U layer turns and allows some look ahead into ZBLL since it's got similar properties to R' D' R <U> R' D R)
F2L 19 - Reduce to F2L 1
F2L 21 - Reduce to F2L 1
F2L 23 - Reduce to F2L 1

F2L 25 - Reduce to F2L 1 (U' F' R U R' U' R' F R & S' R U' R' f R' F' R covers all but 1 case)
F2L 27 - Reduce to F2L 1
F2L 30 - Reduce to F2L 4 (R U R' U' R U R' & U' R U2' R' U2 R U R' covers all but 1 case, can also technically do F R' F' R into F2L 2 but it's kinda gross)
F2L 32 - Reduce to F2L 1
F2L 33 - Reduce to F2L 1
F2L 34 - U R' D' R U' R' D R & U R U R' U2' R U R' (can also do D R' U R to reduce to F2L 2)
F2L 38 - R U R' U' R U2 R' U' R U R' or R U' R' U' R U R' U' R U2' R' with the correct pre-AUF (or reduce to F2L 1)
F2L 39 - R U' R' U R U2' R' U R U' R' with the correct pre-AUF (or reduce to F2L 1)


In total, 15/20 are worth it using this system, and of the remaining 5, the only one that definitely isn't worth it in my opinion is F2L 4 aka split free pair. I'd say suck it up for this case and take H/Pi unless you can cancel nicely into the standard OLL (happens for 3 of the 6 cases), but even then it's debatable and I'm very much inclined to think you should always take H/Pi here

Reducing to F2L 4 as such isn't anywhere near as good as reducing to F2L 1, so for that reason I don't like F2L 17 with this system since it only has one way to reduce to a free pair in 3-4 moves (whereas F2L 30 at least has a 2nd way that removes the H/Pi for all but one case). I think that for the 6 cases where the standard insert gives H/Pi, you're better off taking OLL unless there's a good 11 mover (ignoring pre-AUF) for the pure OLS (ones that I found by hand: R U2' R' U' R U R' U2 R U' R' and R U R' U' R U2' R' U2 R U R')

F2L 14 is interesting to me. Even though it has two algs, the secondary alg is 3QTM worse and can't be done from home grip. That said, it's basically a sune with an extra R turn, so I'm more inclined to think it's worth it even if it's a bit weird

F2L 25 I'm just kinda sad about lol. I tried a bunch of combos of F' R U R' U' R' F R + inverse & S' R U' R' f R' F' R + inverse but couldn't get 100% with only 2 algs. To get 100% you need to use U' R' F R f' R U R' S and then you get U ZBLL from the U2 angle (or maybe there's a good BLE alg)

Lastly, for the F2L 30 OLS that always gives 4 misoriented corners using the 2 good solutions, you can either do the 7QTM one and get H (which is still nicer than Pi) do F R' F' R d' L' U' L U L' U L (which kinda sucks tbh), or do U' R U2' L U' R' U L' to get T ZBLL from the U2 angle (also kinda sucks)

All in all, using the most conservative version of this system (aka not using gross F2L algs, only 2 per case, and no fancy OLS tricks like WV), you're forced to take H/Pi for 14 out of the 540 possible OLS cases (not sure what the actual stats are on this but I know that just dividing them is incorrect because symmetry or something)

This seems pretty damn solid to me for a start considering that with 0 influence, you'd be getting H/Pi for 120/540 cases

That said, if you were only to use reduction to F2L 1 tricks (excluding F2L 34), the final number of cases where you're stuck getting H or Pi becomes 42/540. This is important because reducing to F2L 1 aka a built pair means you only need to learn one recog system, and you can also do WV on top of that for some cases. This is absolutely the place to start if you're a ZZ solver looking to up your OLS game B)
Conversely, wouldn't this potentially cause problems with imbalanced TUL(SA) ZB and HP skill? If you end up getting HP even less, you're going to end up practicing and using them less. A lot of people still question the value of HP ZBLL already so getting it even less probably isn't going to help. Plus H and Pi are no worse than any other set for recog if you use a system like BH or Twisty PLL recognition.
 

ottozing

Platinum Member
Joined
Aug 13, 2011
Messages
3,260
Location
Canberra, Australia
WCA
2012MCNE01
YouTube
Visit Channel
Conversely, wouldn't this potentially cause problems with imbalanced TUL(SA) ZB and HP skill? If you end up getting HP even less, you're going to end up practicing and using them less. A lot of people still question the value of HP ZBLL already so getting it even less probably isn't going to help. Plus H and Pi are no worse than any other set for recog if you use a system like BH or Twisty PLL recognition.
I don't think this is necessarily a problem if H/Pi are the "problem sets", especially if it's potentially this easy to avoid. People get dot OLL's less often on 3x3 because there are a lot of scenarios where edge control is very easy, but that doesn't mean they suck at dot OLL's when they get them. Obviously I'm comparing an 8 alg set to something a little over 100 algs, but you get the point

If people decide that H in full is worth it but only 2/3rd of Pi is, for example, then we can simplify the OCLL reduction further by only focusing on the 4/27 cases that lead to Pi instead of the 6/27 that lead to 4 corners flipped, all while taking advantage of H being good (if it is)

As for BH/twisty PLL recognition, I want someone to actually make a video explaining why it's at least potentially better than standard "look for patterns" style recognition (assuming one doesn't already exist), because to me it just seems straight up worse (even if it makes learning ZB easier)
 

ottozing

Platinum Member
Joined
Aug 13, 2011
Messages
3,260
Location
Canberra, Australia
WCA
2012MCNE01
YouTube
Visit Channel
+10 effort. How did you gen the algs?
Just used general F2L knowledge and brute force tried them out on 6 3x3's with the 6 unique 4 corner twist angles to see if I could make two mice algs work. This is why I want my OLS tool developed lmao doing this by hand took like an hour when it should have taken like 5-15 mins
 

OreKehStrah

Member
Joined
May 24, 2019
Messages
686
I don't think this is necessarily a problem if H/Pi are the "problem sets", especially if it's potentially this easy to avoid. People get dot OLL's less often on 3x3 because there are a lot of scenarios where edge control is very easy, but that doesn't mean they suck at dot OLL's when they get them. Obviously I'm comparing an 8 alg set to something a little over 100 algs, but you get the point

If people decide that H in full is worth it but only 2/3rd of Pi is, for example, then we can simplify the OCLL reduction further by only focusing on the 4/27 cases that lead to Pi instead of the 6/27 that lead to 4 corners flipped, all while taking advantage of H being good (if it is)

As for BH/twisty PLL recognition, I want someone to actually make a video explaining why it's at least potentially better than standard "look for patterns" style recognition (assuming one doesn't already exist), because to me it just seems straight up worse (even if it makes learning ZB easier)
I can make a video on BH/Twisty if you want, but TLDR: Block based recog is really good when you get the AUF naturally for the blocks you recog from but otherwise you're going to need to AUF to the recog angle, then exec angle. As such, block based recog is very difficult to learn for all 4 angles. BH is a very consistent system because you always look at the same pieces every time vs recognizing blocks and can be learned for all 4 angles, although it does take a fair amount of work to learn for all 4 angles, but it is at least consistent and systematic at what is checked and learned. Twisty is similar to BH. You recog CP and take note of where headlights are, if any, and then mentally twist the UFR corner to orient it, and recognize what PLL has that 2x2 block of UF, UFR, UR pattern relative to the headlights. The main benefit of this is that it should be easier to learn all 4 angles for.
 

abunickabhi

Member
Joined
Jan 9, 2014
Messages
2,955
Location
Yo
WCA
2013GHOD01
YouTube
Visit Channel
Generally speaking, people who use full ZBLL agree that for ZBLL cases, having 2 oriented corners is the best while having 0 oriented corners is the worst (this is accounting for recognition, quality of algs, and how nice OLL>PLL is in comparison)

What I've gone ahead and done is taken every possible LS case you can get with ZZ for the FR slot & found one alternative algorithm that can be used to guarantee that you never get 0 oriented corners for LL. Not all of these are actually worthwhile, but many of them are & the recognition demands really aren't that bad once you get used to them (won't be explaining that part here)

This idea can also apply to ZBLS, and in my opinion will probably become the primary approach to LSLL for advanced CFOP solvers, with the secondary approach being ZBLS, and the tertiary approach being some mix of other stuff like VLS, true OLS, and forcing very nice OLL cases in terms of move count/look ahead into PLL/good alternative OLLCP's

Bolded = worth it
Underlined = almost worth it but needs work/unsure
No bold/no underline = not worth it

F2L 1 - U R U' R' & U2 R U2' R'
F2L 4 - R U R' & R U' R' U' R U' R'
F2L 5 - Reduce to F2L 1
F2L 7 - Reduce to F2L 1
F2L 10 - U2 R U' R' U' R U R' & U' R U R' U R U R'
F2L 12 - R' U2' R2 U R2' U R & R U' R' U R U' R' U2 R U' R' (can also just reduce to F2L 1 with the 2nd or F2L 2 with the 1st)

F2L 14 - U' R U' R' U R U R' & U2' R2 U R' U R U2' R2' (can also reduce to F2L 4 with the 1st)
F2L 15 - Reduce to F2L 1
F2L 17 - Reduce to F2L 4 (could also use R U2' R' U' R U2' R' U R U2' R' as the alt instead of R U2' R' U' R U' R' U' R U' R' which has nicer U layer turns and allows some look ahead into ZBLL since it's got similar properties to R' D' R <U> R' D R)
F2L 19 - Reduce to F2L 1
F2L 21 - Reduce to F2L 1
F2L 23 - Reduce to F2L 1

F2L 25 - Reduce to F2L 1 (U' F' R U R' U' R' F R & S' R U' R' f R' F' R covers all but 1 case)
F2L 27 - Reduce to F2L 1
F2L 30 - Reduce to F2L 4 (R U R' U' R U R' & U' R U2' R' U2 R U R' covers all but 1 case, can also technically do F R' F' R into F2L 2 but it's kinda gross)
F2L 32 - Reduce to F2L 1
F2L 33 - Reduce to F2L 1
F2L 34 - U R' D' R U' R' D R & U R U R' U2' R U R' (can also do D R' U R to reduce to F2L 2)
F2L 38 - R U R' U' R U2 R' U' R U R' or R U' R' U' R U R' U' R U2' R' with the correct pre-AUF (or reduce to F2L 1)
F2L 39 - R U' R' U R U2' R' U R U' R' with the correct pre-AUF (or reduce to F2L 1)


In total, 15/20 are worth it using this system, and of the remaining 5, the only one that definitely isn't worth it in my opinion is F2L 4 aka split free pair. I'd say suck it up for this case and take H/Pi unless you can cancel nicely into the standard OLL (happens for 3 of the 6 cases), but even then it's debatable and I'm very much inclined to think you should always take H/Pi here

Reducing to F2L 4 as such isn't anywhere near as good as reducing to F2L 1, so for that reason I don't like F2L 17 with this system since it only has one way to reduce to a free pair in 3-4 moves (whereas F2L 30 at least has a 2nd way that removes the H/Pi for all but one case). I think that for the 6 cases where the standard insert gives H/Pi, you're better off taking OLL unless there's a good 11 mover (ignoring pre-AUF) for the pure OLS (ones that I found by hand: R U2' R' U' R U R' U2 R U' R' and R U R' U' R U2' R' U2 R U R')

F2L 14 is interesting to me. Even though it has two algs, the secondary alg is 3QTM worse and can't be done from home grip. That said, it's basically a sune with an extra R turn, so I'm more inclined to think it's worth it even if it's a bit weird

F2L 25 I'm just kinda sad about lol. I tried a bunch of combos of F' R U R' U' R' F R + inverse & S' R U' R' f R' F' R + inverse but couldn't get 100% with only 2 algs. To get 100% you need to use U' R' F R f' R U R' S and then you get U ZBLL from the U2 angle (or maybe there's a good BLE alg)

Lastly, for the F2L 30 OLS that always gives 4 misoriented corners using the 2 good solutions, you can either do the 7QTM one and get H (which is still nicer than Pi) do F R' F' R d' L' U' L U L' U L (which kinda sucks tbh), or do U' R U2' L U' R' U L' to get T ZBLL from the U2 angle (also kinda sucks)

All in all, using the most conservative version of this system (aka not using gross F2L algs, only 2 per case, and no fancy OLS tricks like WV), you're forced to take H/Pi for 14 out of the 540 possible OLS cases (not sure what the actual stats are on this but I know that just dividing them is incorrect because symmetry or something)

This seems pretty damn solid to me for a start considering that with 0 influence, you'd be getting H/Pi for 120/540 cases

That said, if you were only to use reduction to F2L 1 tricks (excluding F2L 34), the final number of cases where you're stuck getting H or Pi becomes 42/540. This is important because reducing to F2L 1 aka a built pair means you only need to learn one recog system, and you can also do WV on top of that for some cases. This is absolutely the place to start if you're a ZZ solver looking to up your OLS game B)
+10 for the effort. Had to read the post in order to understand all the points.
 
Top