The New Method / Substep / Concept Idea Thread

[supercavitation]

I have a cool idea for an algorithm set. Instead of ZBLL, with 493 algorithms, we can use COALL, with 111 Algs. This can be viewed on the wiki. The difference between the two is efficiency. Many methods like FreeFOP, ZZ, and Petrus setup for ZBLL with no Algs, making ZBLL a good choice (for 1LLL). COALL, however, takes advantage of oriented corners, which then results in less Algs. The thing is, the only way to setup for this is to use CFOP and use WV. I feel like COALL could replace ZBLL altogether. So we need something besides WV.
My idea for the algorithm set I mentioned earlier is to use FreeFOP up to the point where you insert the last edge piece and orient the edges. Instead of orienting edges, we can use an alg set of 28 to orient the corners while inserting the edge.
This isn't the best solution, but it's an idea. There may be a way to set up for this intuitively, like orienting the edges.
I'm not sure how this would work out and even if it is a good idea. But I think it's worth experimenting with.
Let me know what you think.

This has been brought up at least a half dozen times in this thread (including by me a while ago). There are some alg lists, but too many of the cases are really bad.

 

[Shiv3r]

Would a method converting F2B to F2L and then doing LL be any good?
also, ZZ seems like it would be the god of one-handed cubing competitions

 

[Teoidus]

Would a method converting F2B to F2L and then doing LL be any good?
also, ZZ seems like it would be the god of one-handed cubing competitions

Yeah, apparently literally everyone thinks of the Roux-CFOP hybrid at some point. I tried making it work a bit and ended up writing up some stuff here, but after learning more about Roux I think you're better off not mixing the two.

 

[shadowslice e]

Would a method converting F2B to F2L and then doing LL be any good?

Yeah this is always brought up and isn't any better than the vanilla versions of either.

also, ZZ seems like it would be the god of one-handed cubing competitions

ZZ is OP for one handed but I wouldn't discount Roux as the CMLL and SB are great and the LSE is very good once you get used to the LSE. Pertrus is also good after FB as well. I would probably put the methods as

1) ZZ
2) Roux
3) Petrus
4) CFOP (though if you do eo during F2L it can still be pretty fast although I think FreeFOP may still be better).

 

[supercavitation]

ZZ is OP for one handed but I wouldn't discount Roux as the CMLL and SB are great and the LSE is very good once you get used to the LSE. Pertrus is also good after FB as well. I would probably put the methods as

1) ZZ
2) Roux
3) Petrus
4) CFOP (though if you do eo during F2L it can still be pretty fast although I think FreeFOP may still be better).

I would argue that it depends heavily on what you do for 2H. I used to use ZZ for OH, but eventually switched to CFOP after Pavan told me to a few dozen times. I dropped 3.5 seconds in a month and a half. I'm more used to CFOP lookahead than I am to ZZ lookahead, and it was slowing me down.

EDIT: tl;dr CFOP can still be great for OH.

 

[TDM]

Yeah this is always brought up and isn't any better than the vanilla versions of either.



ZZ is OP for one handed but I wouldn't discount Roux as the CMLL and SB are great and the LSE is very good once you get used to the LSE. Pertrus is also good after FB as well. I would probably put the methods as

1) ZZ
2) Roux
3) Petrus
4) CFOP (though if you do eo during F2L it can still be pretty fast although I think FreeFOP may still be better).

I don't know, Petrus' blocks are pretty bad. I'd put them as
1) Roux
2) ZZ
3) CFOP
4) Petrus

 

[shadowslice e]

I don't know, Petrus' blocks are pretty bad. I'd put them as
1) Roux
2) ZZ
3) CFOP
4) Petrus

Yeah I guess though once you get the blocks and Eo good it could be pretty fast for the last bit. Though if you do Eo during F2L I would completely agree that CFOP is better. I just really don't like doing rotations, especially during OH.

Also, you would put roux before ZZ? I've not seen anyone else do that (for OH)

 

[10 Second Cube]

10 Second Cube.

That was what we were working on in 1984 - just when the cube tournaments died off.

About myself. I was a cube pioneer. In 1982 I saw the TV show That's Incredible. This was the first cube contest on TV. It was sponsored by Ideal Toy company. They had the rights to the cube (in the USA or worldwide - not sure) at the time. After seeing the show, I went out and bought a cube the next day. I bought a small keychain cube. It took a week to almost solve it and by week two I figured it out. My genius was actually writing things down in a note book. I quickly figured out some basic algorithms.

As time went on I was getting faster. Soon I was under two minutes. Heck, some contestants on the TV didn't finish in one minute. I went out and bought a real Rubik's Cube. Soon I broke one minute. I was excited. I wrote a letter to Ideal Toy company asking about the contest. They sent me a newsletter. In the newsletter was a schedule of tournaments. One was scheduled for Burnsville Mall in Minnesota. I was only 14 so I asked my mom to drive me to it. She hadn't a clue what it was. I competed and hit 30.68 seconds. As it was the last tournament in the region and my time the best, I was instantly named champion and would be going to the finals in Hollywood. My life had just changed dramatically.

Keep in mind in 1982 there was no internet - no collaboration. Until I competed in the contest, I had never met anyone who could solve the cube - much less break one minute. I had my own method - top first, bottom 4 corners, fill in, solve middle and flip last 2 or 4 if necessary. I didn't have all of the algorithms - only the ones I had discovered. I busted my ass trying to get fast for the contest. Just before the contest I had come up with a new method - but it wasn't complete and I wasn't so competent at it (looking at matching and opposing colors).

I went to the contest as the youngest and least experienced (I only had had a cube for 10 months). And this was my first time to collaborate with anyone. All the guys were great. Very intelligent. Most were self taught. David Maze was overall incredibly smart and focused. David Allen had come up with the using his fingers to spin - single and doubles. We were all in awe when he worked the cube. David was obviously the fastest of us. He won the first contest but lost to last year's champion, Min Thai. From Min Thai we learned how to shorted the middle fill in (hard to explain). Min Thai had won the previous year and had a coach. Min was a college student. David was 19 also. The oldest were the best.

This was only the second time I had competed. I was totally unprepared. I had a new method but it took an incredible amount of concentration and I couldn't rise to that level in a contest setting. I couldn't break 30 seconds when competing. Off camera I hit 16 seconds. Oh, we couldn't use our own cubes. Ideal Toy had a bunch of slightly loosened cubes. We got to choose a few from the bunch to use in the Tournament. Of course our times suffered a lot. I was at 16 to 28 seconds with my cube - but above 30 with the stock cubes. They were stiff and we just weren't used to them. David, with his big burly fingers, was able to adapt more quickly.

Anyway, I had a LOT of fun and it was vey inspiring. All of us were invited to attend MIT by one of the professors (Min Thai's coach if I remember correctly). Many of us went to MIT. I just couldn't afford it.

Oh, about half used the CFOP method and half used the corners first method. The top times were all corners first.

Anyway, back to the 10 second cube.

We all collaborated at the contest. We spent a few days together having fun at Disneyland, etc. I spent the next six months in deep research on what would be the fastest method. I broke up each stage and tried to estimate how long it would take for each stage if optimized. No matter how I looked at corners first, I couldn't imagine breaking 15 seconds. I had heard about Jeff Varasano's method from other competitors. It was a revelation as we just hadn't thought of just getting the white and blues in flipped (cubes were white opposite blue then). But Jeff's method was only good for low 20 second times.

Anyhow, the short story is that we came up with the only way to a true 10 second cube (consistently). Sure, I had 8 second times - when the stars, moon and earth happened to align. But consistent 10 second times were just a dream.

What we came up with was:

Flip top and bottom colors (as per the Fridrich method for the bottom). Didn't matter if opposing colors were mixed. We figured why bother with finishing the top as we could do it when doing the bottom.
We assumed we would also figure out a way to flip the center pieces when doing the top and bottom. Once all are aligned it is incredibly fast to finish. Sorry, I am not up on the modern cube lingo to give a good description.

We started developing the algorithms - and our times for the new method were breaking 20 seconds. I thought I would be a professional cubist. Then Ideal Toy company was bought and the show contest cancelled. Then we all went to college and stopped out cubing careers. It was a long while before cube contests started again.

For us it was different. We had to figure this stuff out on our own - it wasn't about memorizing a bunch of algos - it was about finding them.

I still think this is the fastest consistent method.

Also, I highly doubt the 5 or 6 second times. Sure, they are possible in a fluke - or when a buddy 'mixes' the cube for you. If you cut up each section - it just doesn't add up. There are three stages minimum - The 4 edges (and some corners). The other 4 edges. The bottom. Sure, the bottom can be done in one algo - but it still takes 2 - 3 seconds for that algo leaving 3 or 4 for the rest? Hmmmm....

 

[shadowslice e]

Hello! It's nice to meet a cubing pioneer as always!

Regarding your method, do you mean something like this:
1) orient U and D pieces (only yellow and white facing up and down) and solve the e-slice (the middle of the cube)
2) separate the U and D pieces (yellow facing down, white facing up)
3) solve the U and D face
?

This method sounds like an early version of the Thistlethwaite Algorithm (and was developed into the human thistlethwaite algorithm by Ryan Heise and I would say that one of my methids (SSC) is probably the spiritual succesor/ improvement to those methods).

With regards to your scepticism of the times in the 4-6 second range, they are not the norm no, however, they were set at official competitions where all the scrambles are completely randomised and computer generated. The fastest average anyone has been able to achieve is just below 7 seconds as the average of 100 solves. This is the main reason why the Wca ( the current governing body) registers average of 5 for 3x3 solves to declare winners. The current record for the average of 5 is 6.54 by feliks zemdegs and is a fair bit ahead of any other person ( though less than he used to be- thr time is also the oldest standing record in a tie with MultiBind) and feliks himself is widely regarded as the most successful cuber in history as he also dominated many other events such as all the big cubes and one handed.

Speaking of methods, have you had a look at the modern methods? The most commonly used ones are CFOP, Roux and ZZ in that order.

Sent from my M1005D using Tapatalk

 

[supercavitation]

I just really don't like doing rotations, especially during OH.

As much as I feel like I'm selling what little bit of soul I've managed to retain when I say this, the rotations aren't bad at all when you get used to them.

If you don't use CFOP for 2H, definitely don't use CFOP for OH, though. Even as someone who uses CFOP for OH, I would argue that it's only better for OH when it means your lookahead and tricks from 2H translate.

 

[TDM]

Yeah I guess though once you get the blocks and Eo good it could be pretty fast for the last bit. Though if you do Eo during F2L I would completely agree that CFOP is better. I just really don't like doing rotations, especially during OH.

Yep, the last steps of Petrus are fast; it's similar to ZZ, but with guaranteed 2-gen for the last two blocks, whereas with ZZ you sometimes won't get that (although I think most ZZers do solve one block at a time). I think CFOP has the advantage in the earlier stages of the solve though, which is why it could maybe go either way.

Also, you would put roux before ZZ? I've not seen anyone else do that (for OH)

There are three key things about a good method:
- Efficiency
- Ease of fingertricks (TPS)
- Lookahead

Roux beats ZZ on efficiency easily. I find the fingertricks a lot easier too; not having to do rotations or L moves is really useful. Although FB is horrible OH, EOLine isn't much better. SB>EOF2L for fingertricks, and CMLLs are better than COLLs (on average). LSE and EPLL both have good fingertricks, though I think EPLL is marginally better.

Lookahead is the tricky one to look at. Both methods are rotationless(ish - ZZ can have z rotations), which has its advantages and disadvantages. Although you don't need to keep track of where your next pair/block is while rotating, which can be difficult in CFOP (especially keeping track of orientation of pieces), you do need to know what's in the back. ZZ had BL and BR, and Roux has BL/BR and BD. BD is the hardest case since you can't see either sticker, whereas with the other pieces you at least have some information. So rotationlessness is a bit of a disadvantage to both methods, but more Roux than ZZ.

Then there's the question of where the pieces themselves are moving. With CFOP, this is easy: either they're trapped in slots and completely stationary (thus easy to track), or in the U layer and visible. In ZZ, they can switch between the L and R layers. There also are a greater number of places both edges and corners can go in comparison to Roux, and to CFOP (for edges). Furthermore, there's also the fact that this is OH: with Roux your hand only covers solved pieces. With ZZ it can hide some things you need.

With Roux, on the other hand, everything is almost completely limited to being in one of two layers (except for the bit above FB). This does make looking for pieces easier. However the difficulty is greatly increased by edge orientation; with ZZ you don't need to care. You just need to know where the pieces are and you can solve the case. With Roux the edges can be easily flipped with just an Rw move - useful for solving cases, but bad for looking ahead, especially with one edge in the D layer. However with far fewer pieces to be looking for you can concentrate on EO relatively easily, meaning this isn't actually such a bad thing after all. I actually think that only solving five pieces makes lookahead very easy with Roux, especially since it's only really two after you've started solving one block. This is probably the most important point but I can't think of a way of making it more than two sentences. :/

So I think the two methods are about the same with lookahead during F2L/B (Roux maybe having a slight edge), but then that's only half the solve. Looking ahead in ZZ LL is much easier than CFOP LL. However, Roux has CMLL -> LSE, and LSE itself. I'd say this is even easier to look ahead with: I think Kirjava said LSE was "0-look" because so little is being solved at each stage you can see what's coming next easily. One last comment that's almost entirely personal preference: I find it easier learning what effect CxLL has on EO (for CMLL lookahead) than EP (COLL lookahead). You just need to know which edges are effected; with EP, you also need to know how they are effected.

So yeah, you probably didn't want all that, but I didn't want to miss much out.

 

[crafto22]

10 Second Cube.

That was what we were working on in 1984 - just when the cube tournaments died off.

About myself. I was a cube pioneer. In 1982 I saw the TV show That's Incredible. This was the first cube contest on TV. It was sponsored by Ideal Toy company. They had the rights to the cube (in the USA or worldwide - not sure) at the time. After seeing the show, I went out and bought a cube the next day. I bought a small keychain cube. It took a week to almost solve it and by week two I figured it out. My genius was actually writing things down in a note book. I quickly figured out some basic algorithms.

As time went on I was getting faster. Soon I was under two minutes. Heck, some contestants on the TV didn't finish in one minute. I went out and bought a real Rubik's Cube. Soon I broke one minute. I was excited. I wrote a letter to Ideal Toy company asking about the contest. They sent me a newsletter. In the newsletter was a schedule of tournaments. One was scheduled for Burnsville Mall in Minnesota. I was only 14 so I asked my mom to drive me to it. She hadn't a clue what it was. I competed and hit 30.68 seconds. As it was the last tournament in the region and my time the best, I was instantly named champion and would be going to the finals in Hollywood. My life had just changed dramatically.

Keep in mind in 1982 there was no internet - no collaboration. Until I competed in the contest, I had never met anyone who could solve the cube - much less break one minute. I had my own method - top first, bottom 4 corners, fill in, solve middle and flip last 2 or 4 if necessary. I didn't have all of the algorithms - only the ones I had discovered. I busted my ass trying to get fast for the contest. Just before the contest I had come up with a new method - but it wasn't complete and I wasn't so competent at it (looking at matching and opposing colors).

I went to the contest as the youngest and least experienced (I only had had a cube for 10 months). And this was my first time to collaborate with anyone. All the guys were great. Very intelligent. Most were self taught. David Maze was overall incredibly smart and focused. David Allen had come up with the using his fingers to spin - single and doubles. We were all in awe when he worked the cube. David was obviously the fastest of us. He won the first contest but lost to last year's champion, Min Thai. From Min Thai we learned how to shorted the middle fill in (hard to explain). Min Thai had won the previous year and had a coach. Min was a college student. David was 19 also. The oldest were the best.

This was only the second time I had competed. I was totally unprepared. I had a new method but it took an incredible amount of concentration and I couldn't rise to that level in a contest setting. I couldn't break 30 seconds when competing. Off camera I hit 16 seconds. Oh, we couldn't use our own cubes. Ideal Toy had a bunch of slightly loosened cubes. We got to choose a few from the bunch to use in the Tournament. Of course our times suffered a lot. I was at 16 to 28 seconds with my cube - but above 30 with the stock cubes. They were stiff and we just weren't used to them. David, with his big burly fingers, was able to adapt more quickly.

Anyway, I had a LOT of fun and it was vey inspiring. All of us were invited to attend MIT by one of the professors (Min Thai's coach if I remember correctly). Many of us went to MIT. I just couldn't afford it.

Oh, about half used the CFOP method and half used the corners first method. The top times were all corners first.

Anyway, back to the 10 second cube.

We all collaborated at the contest. We spent a few days together having fun at Disneyland, etc. I spent the next six months in deep research on what would be the fastest method. I broke up each stage and tried to estimate how long it would take for each stage if optimized. No matter how I looked at corners first, I couldn't imagine breaking 15 seconds. I had heard about Jeff Varasano's method from other competitors. It was a revelation as we just hadn't thought of just getting the white and blues in flipped (cubes were white opposite blue then). But Jeff's method was only good for low 20 second times.

Anyhow, the short story is that we came up with the only way to a true 10 second cube (consistently). Sure, I had 8 second times - when the stars, moon and earth happened to align. But consistent 10 second times were just a dream.

What we came up with was:

Flip top and bottom colors (as per the Fridrich method for the bottom). Didn't matter if opposing colors were mixed. We figured why bother with finishing the top as we could do it when doing the bottom.
We assumed we would also figure out a way to flip the center pieces when doing the top and bottom. Once all are aligned it is incredibly fast to finish. Sorry, I am not up on the modern cube lingo to give a good description.

We started developing the algorithms - and our times for the new method were breaking 20 seconds. I thought I would be a professional cubist. Then Ideal Toy company was bought and the show contest cancelled. Then we all went to college and stopped out cubing careers. It was a long while before cube contests started again.

For us it was different. We had to figure this stuff out on our own - it wasn't about memorizing a bunch of algos - it was about finding them.

I still think this is the fastest consistent method.

Also, I highly doubt the 5 or 6 second times. Sure, they are possible in a fluke - or when a buddy 'mixes' the cube for you. If you cut up each section - it just doesn't add up. There are three stages minimum - The 4 edges (and some corners). The other 4 edges. The bottom. Sure, the bottom can be done in one algo - but it still takes 2 - 3 seconds for that algo leaving 3 or 4 for the rest? Hmmmm....

Regarding your doubt in the 5 and 6 second times, they are certainly possible, I know from experience. You can't possibly think that the computer generated scrambles are fake. 5, 6 or even 4 second times are definetely possible, if you looked at the time it takes for each phase of the solution, it adds up. I myself have gotten a couple of 6 second times, a result of lucky scrambles of course, but I know consistent 6 second times are possible, and Feliks Zemdegs is already almost there.

 

[shadowslice e]

As much as I feel like I'm selling what little bit of soul I've managed to retain when I say this, the rotations aren't bad at all when you get used to them.

If you don't use CFOP for 2H, definitely don't use CFOP for OH, though. Even as someone who uses CFOP for OH, I would argue that it's only better for OH when it means your lookahead and tricks from 2H translate.

Yeah but that doesn't mean the method is actually good. It may be better for a specific person but that doesn't mean the method us objectively better. For example, just because rotations "aren't that bad" doesn't mean it wouldn't be better without them.

So yeah, you probably didn't want all that, but I didn't want to miss much out.

No actually, it was quite nice to read a post where someone actually bothered to completely justify their thought processes.

A little bit shorter would've been nice though (not that I can really talk )

 

[TDM]

No actually, it was quite nice to read a post where someone actually bothered to completely justify their thought processes.

A little bit shorter would've been nice though (not that I can really talk )

Haha, I'll try for next time!

though i've done worse before...
E: some of my opinions on that link have probably changed since then, so I wouldn't bother reading it. I'll edit it if I ever want to use it again at some point.

 

[supercavitation]

Yeah but that doesn't mean the method is actually good. It may be better for a specific person but that doesn't mean the method us objectively better. For example, just because rotations "aren't that bad" doesn't mean it wouldn't be better without them.

It's certainly good, though I won't argue that it's better than ZZ by any means (I would agree that it's worse in many cases). That second sentence was basically my point. CFOP is certainly better for many people, for the same reasons I got faster when I switched (lookahead and tricks being similar to 2H). I agree that rotationless is better for OH, though I will note that I said they "aren't bad," not that that "'aren't that bad'." I would argue that they don't necessarily hurt the method significantly, as, with practice, they both get significantly faster, and tend to decrease.

 

[wir3sandfir3s]

I have an idea that definitely needs work but has potential.
Everybody focuses on making LL variants for ZZ, but what about F2L variants?
So I hate LL because I suck at it, pretty much. Solving it during f2l would be awesome. If you do OLL and PLL for ZZ, there are a total of 29 LL cases because edges are oriented. 7 are OLL. 21 are PLL. Pretty light on the algorithms.
I am not really used to ZZF2L either. I can do one block pretty decently, but I struggle with the other for some reason.
With all this in mind, I came up with the idea of the DLS technique (Double Last Slot), or L2S (last 2 slots).
This probably has been thought of before, but I thought I would post it anyway because I have several more ideas involving it.
So to set up a DLS, solve the cube with ZZ up to the point where you have I 1x2x3 block and LL left to solve. Solve the last D edge, and pair the last 2 corners and edges. Place the last 2 pairs in the top layer, and they should be adjacent to each other because edges are already oriented. AUF until they are both on the left. Insert the first pair while orienting corners using 1 of 7 algorithms. Insert the second using 1 of 21 algorithms. Cube solved.
Sounds great, right? LL while inserting the last F2L pairs? Probably not because you have probably figured out by now that the recognition sucks.
I'll work on recognition and post solutions here, otherwise this seems like an okay idea.

 

[xyzzy]

Solve the last D edge, and pair the last 2 corners and edges. Place the last 2 pairs in the top layer, and they should be adjacent to each other because edges are already oriented.

This sounds like the hard part. The first pair is easy enough to make, but making the second pair while keeping the first pair in the U layer is nontrivial, especially if the edge is stuck in an inconvenient location. There're also four possible positions (pairs are parallel, pairs are perpendicular, one of the pairs already solved, or F2L skip), not just one.

Insert the first pair while orienting corners using 1 of 7 algorithms. Insert the second using 1 of 21 algorithms.

The first part sounds like WV, except much more constrained and thus less efficient. You actually have more than just eight cases here because the pair forces a particular AUF, and you also have to deal with the orientation of the corner in the second slot.

 

[wir3sandfir3s]

This sounds like the hard part. The first pair is easy enough to make, but making the second pair while keeping the first pair in the U layer is nontrivial, especially if the edge is stuck in an inconvenient location. There're also four possible positions (pairs are parallel, pairs are perpendicular, one of the pairs already solved, or F2L skip), not just one.

It's actually pretty easy, I'll do a video after enough improvement (if any). And there is only one position after both pairs are in the U layer. Bad wording on my part. Sorry.

The first part sounds like WV, except much more constrained and thus less efficient. You actually have more than just eight cases here because the pair forces a particular AUF, and you also have to deal with the orientation of the corner in the second slot.

It's actually the same constrainment, just seems like more because of recog. And yes, if my math doesn't fail again, there are actually 28 cases, which still isn't absolutely awful. Bad math on my part. And you don't actually have to keep track of the corner in the second slot (if I am correct, I seem to be very wrong lately) because of corner orientation patterns, like you don't have to keep track of the corner in the slot in WV.
Hope I've justified this all.

 

[xyzzy]

And there is only one position after both pairs are in the U layer. Bad wording on my part. Sorry.

R U2 R2 U' R and R U R' U' R' U' R have the pairs in different positions. (Or am I misunderstanding your posts?)

like you don't have to keep track of the corner in the slot in WV.

Right; I forgot about that. My bad.

 

[wir3sandfir3s]

R U2 R2 U' R and R U R' U' R' U' R have the pairs in different positions. (Or am I misunderstanding your posts?)

Solve with either Petrus or ZZ up to this point, then try making the pairs with the last D edge in place. In the U layer, they should be adjacent to each other and there should be a "T" of unsolved pieces (if that makes any sense). I can upload a picture if you are still confused.