The New Method / Substep / Concept Idea Thread

[WoowyBaby]

TL;DR My last 3 posts on this thread aren't worth much

Woowy's Thistlewaite Reduction FMC Example Solve
Scr: U F' B' U L2 U2 F' D L2 F2 R D2 B2 D2 R F2 L' F2 L B2
(z2 y')
F2 U2 F' D' L2 // EO HTO2x2x2
F2 U' R2 L' U L // ->2x2x3 CP [STEP 1 DONE]
R U2 R U R U'
R2 U' R U2 R' U2
.....
.....
nope

I've concluded that Woowy's Thistlewaite Reduction is not all that great. By doing some testing, I've seen the best way to get yo Half-Turn-Only is most likely Kociemba Phase 1 -> Reduce to HTO. For right now this method should probably be forgotten.

Next method I'm trying to make is a reliable way to do Kociemba Phase 2 efficiently- I want to improve how we do FMC!

 

[ImmolatedMarmoset]

Speaking of Kociemba, how can I learn to do conjugates reliably? You were talking about them as useful for CO in Kociemba but I still don’t really understand how to make a setup move so that it will twist certain corners. Is it more of a trial and error thing, or what? I’m confused.

 

[WoowyBaby]

Speaking of Kociemba, how can I learn to do conjugates reliably? You were talking about them as useful for CO in Kociemba but I still don’t really understand how to make a setup move so that it will twist certain corners. Is it more of a trial and error thing, or what? I’m confused.

I actually don’t use whatever you’re talking about very often because it would add around 7 moves to my Phase 1 solution which is a lot, so really don’t worry about it.
But a similar idea, where you do a setup move to twist 4 corners and move 4 edges is a lot more useful.
- Doing a quarter R move twists 4 corners just like how a quarter F move flips 4 edges.
Example here-
Scr: D2 B2 R2 F' R2 D2 B' D2 F' L' D B2 D F' L2 B R' U'
(z')
U L2 F' U F2 B' // Phase 1 except 4Edges4Corners
L2 F2 D R' // Setup to Single Move, Phase 1 Finish

Sorry if I went off-topic there... so how can you learn to do conjugates reliably? I don't know honestly I'm sorry...


Also I have an idea of how to reliably think about Kociemba Phase 2, but I need to test it before I post, so stay tuned or whatever!
Edit: HTO Reduction might be it...

 

[ImmolatedMarmoset]

I actually don’t use whatever you’re talking about very often because it would add around 7 moves to my Phase 1 solution which is a lot, so really don’t worry about it.
But a similar idea, where you do a setup move to twist 4 corners and move 4 edges is a lot more useful.
- Doing a quarter R move twists 4 corners just like how a quarter F move flips 4 edges.
Example here-
Scr: D2 B2 R2 F' R2 D2 B' D2 F' L' D B2 D F' L2 B R' U'
(z')
U L2 F' U F2 B' // Phase 1 except 4Edges4Corners ( EO Kociemba 2x2x3 and Pairs)
L2 F2 D R' // Setup to Single Move, Phase 1 Finish

Sorry if I went off-topic there... so how can you learn to do conjugates reliably? I don't know honestly I'm sorry...


Also I have an idea of how to reliably think about Kociemba Phase 2, but I need to test it before I post, so stay tuned or whatever!

Plz just tell me now! I won’t have internet for like 4 days and I’d like to know it before then lol so I can do more FMC

 

[WoowyBaby]

Plz just tell me now! I won’t have internet for like 4 days and I’d like to know it before then lol so I can do more FMC

My idea for Phase 2? Well first I need to figure it out if it’s a waste of time or not.
For the meantime, do Half-Turn-Only Reduction, because it’s extremely useful and has even gotten me a 19 and a 20(I got yesterday, it’s also 17stm). HTO Reduction might actually be THE reliable 2-step way of doing Phase 2, only time will tell.
If you want to know more about it, search up Thistlewaite (it’s a computer algorithm that does this) and maybe look in the FMC Thread about it.

 

[ImmolatedMarmoset]

My idea for Phase 2? Well first I need to figure it out if it’s a waste of time or not.
For the meantime, do Half-Turn-Only Reduction, because it’s extremely useful and has even gotten me a 19 and a 20(I got yesterday, it’s also 17stm). HTO Reduction might actually be THE reliable 2-step way of doing Phase 2, only time will tell.
If you want to know more about it, search up Thistlewaite (it’s a computer algorithm that does this) and maybe look in the FMC Thread about it. (if you’re interesting in FMC you should probably know of that thread)

I have multiple posts in the FMC thread.

Oh, and also, could you explain your HTO redux method a bit more? I don’t understand what CP means or how to do it, and how do I make and HTO 2x2x3? I’ll search up Thistlethwaite.

 

[WoowyBaby]

I have multiple posts in the FMC thread.

Sorry about that. I guess didn’t recognize you :/ (profile pic change?).

Oh, and also, could you explain your HTO redux method a bit more? I don’t understand what CP means or how to do it, and how do I make and HTO 2x2x3? I’ll search up Thistlethwaite.

I’d think I’ve explained a lot, but here we go-
My HTO Reduction method (I named it Woowy’s Thistlewaite Reduction lol) is not particularly useful right now, so you could honestly forget about it you wanted to.

Anyway, CP stands for Corner Permutation (just like CO or EP or EO), but in this case you’re not actually solving the permutation of the corners (in some other methods you don’t as well), in this you’re placing the corners such that it can be reduced to half-turn-only state using only R and U moves. Yeah. How? One way is to pray for a skip, it’s a 1/6 chance. Of course that’s not the only way

For the HTO 2x2x3 part, it’s like a real 2x2x3 except it can just be opposite colors there. Think of a real 2x2x3 as a 2x2x3 section of a solved cube, then think of an HTO 2x2x3 as a 2x2x3 section of a half-turn-only state. Or it’s just opposite colors in this case, but know that doesn’t quite hold true for the whole. As for how to do it, you just match up blocks, that’s kind of it (not much to explain really).

~New method idea~
"Woowy's R U F Redux"
Steps of this method:
Reduce to [R,U,F] = 2x2x2 Block
Reduce to [R,U,F2] = Edge Orientation
Reduce to [R2,U,F2] = Domino Reduction
Reduce to [R2,U2,F2] = Half-Turn-Only Reduction
Reduce to [] (solved) = HTO Finish

Spoiler: Example Solve + Li'l talk about it

Scr: L2 D2 R2 F2 U' L2 D' L2 U' F2 D2 B R2 B' L U' F' L B' D'
R U L U' B D' //
U F2 R' F //
R2 U' R2 L' U L U' R //
U2 F2 U R2 U2 R2 U' F2 U' F2 U' //
R2 F2 R2 U2 R2 //
34 HTM
Spent like no time looking for this, maybe this is the way to get <40 FMC with minimal effort?
(for FMC): 2x2x2 ~5 , EO ~5 , DR ~8 , HTOR ~10 , Finish ~7 = ~35 moves
Is this good for speedsolving? NO, but I did make a speedsolving method in the past that is EXTREMELY similar to this, called 2OP https://bit.ly/2JFZdDx

 

[presley]

I developed a similar method from the method I posted before.
But I only wrote the Tradition Chinese version article of the new developing method.
Temporarily call it PRUM method and use the idea of mirror reflection.

The step of new developing method are
Step 1, build 2 pairs.
Step 2, retain 2 pairs and into <R, U>.
Step 3, build the 3rd pair.
Step 4, build the 4th pair.
Step 5, extend one pair to 2x2x1 block then end game.

The detail I wrote it on the below website
https://presleygit.github.io/

I also made some video to explain this method roughly
https://www.youtube.com/playlist?list=PLgMFuE1ngI-fP2qOoEUjR_A3rtH0-5Bsg

Hope you like it.

 

[Etotheipi]

Sorry about that. I guess didn’t recognize you :/ (profile pic change?).

I’d think I’ve explained a lot, but here we go-
My HTO Reduction method (I named it Woowy’s Thistlewaite Reduction lol) is not particularly useful right now, so you could honestly forget about it you wanted to.

Anyway, CP stands for Corner Permutation (just like CO or EP or EO), but in this case you’re not actually solving the permutation of the corners (in some other methods you don’t as well), in this you’re placing the corners such that it can be reduced to half-turn-only state using only R and U moves. Yeah. How? One way is to pray for a skip, it’s a 1/6 chance. Of course that’s not the only way

For the HTO 2x2x3 part, it’s like a real 2x2x3 except it can just be opposite colors there. Think of a real 2x2x3 as a 2x2x3 section of a solved cube, then think of an HTO 2x2x3 as a 2x2x3 section of a half-turn-only state. Or it’s just opposite colors in this case, but know that doesn’t quite hold true for the whole. As for how to do it, you just match up blocks, that’s kind of it (not much to explain really).

~New method idea~
"Woowy's R U F Redux"
Steps of this method:
Reduce to [R,U,F] - 2x2x2 Block
Reduce to [R,U,F2] - Edge Orientation
Reduce to [R2,U,F2] - Domino Reduction
Reduce to [R2,U2,F2] - Half-Turn-Only Reduction
Reduce to [] (solved) - Finish

Spoiler: Example Solve + Li'l talk about it

Scr: L2 D2 R2 F2 U' L2 D' L2 U' F2 D2 B R2 B' L U' F' L B' D'
R U L U' B D' //
U F2 R' F //
R2 U' R2 L' U L U' R //
U2 F2 U R2 U2 R2 U' F2 U' F2 U' //
R2 F2 R2 U2 R2 //
34 HTM
Spent like no time looking for this, maybe this is the way to get <40 FMC with minimal effort?
(for FMC): 2x2x2 ~5 , EO ~5 , DR ~8 , HTOR ~10 , Finish ~7 = ~35 moves
Is this good for speedsolving? Not really, but I did make a speedsolving method in the past that is EXTREMELY similar to this, called 2OP https://bit.ly/2JFZdDx

I dont do FMC at all, but i tried this anyway, but i really struggled with the HTOR, could you give me some pointers?

 

[WoowyBaby]

I dont do FMC at all, but i tried this anyway, but i really struggled with the HTOR, could you give me some pointers?

Which method? I'll assume my R U F Redux method for now. Just know that it's not very useful for speedsolving, and I've made a speedsolving method with very similar steps called 2OP.
-Half-Turn-Only Reduction is definitely the hardest step, so don’t worry if you can’t do it. Let’s start with a 2x2 cube. Apply only half turns on it to familiarize yourself with all the possible corner patterns. Notice that its does not simply mean opposite colors on all faces (but that is a requirement to be HTO possible). To be HTO possible, there has to be four identical looking pairs (I sometimes call it quadruplets but I'm not sure if that makes any sense). They all have the same pattern. If there are three "opposite" pairs and one "solved" pair then it's not HTO possible, if there are four "opposite" pairs then it is.
-Now practice on a 2x2. Solve CO, then do Half-Turn-Only Reduction. Then move it over to this 3x3 method step HTOR, and ignore edges for now. Once you get the hang of that, then you can worry about edges.
-Edges are simpler, if it’s opposite colors, it’s reduced to HTO. If they're in their correct slices (and oriented), they're HTO possible. So the WG-edge for example, could be in four spots, solved, UB, DB, and DF.
-You can practice more by doing a R2 U F2 scramble, then just fix the edges. Then scramble it and just fix the corners. Then scramble and try to fix both. It’ll probably be too hard (it’s okay) so to do this in smaller steps, mix it up with R2 U2 F2 moves and then do like 3 moves to mess up HTOR and try to understand how to reverse it, Example: R2 U2 R2 F2 U’ R2 U HTOR: U’ R2 U, and see how it works, and maybe increase the moves to 5 or 6 once you totally understand everything so far.
This idea of starting with just a few moves away and working your way up is actually a practice technique applicable to more than just this.
I have no idea if you need all that, but good luck!

 

[Etotheipi]

Which method? I'll assume my R U F Redux method for now. Just know that it's not very useful for speedsolving, and I've made a speedsolving method with very similar steps called 2OP.
-Half-Turn-Only Reduction is definitely the hardest step, so don’t worry if you can’t do it. Let’s start with a 2x2 cube. Apply only half turns on it to familiarize yourself with all the possible corner patterns. Notice that its does not simply mean opposite colors on all faces (but that is a requirement to be HTO possible). To be HTO possible, there has to be four identical looking pairs (I sometimes call it quadruplets but I'm not sure if that makes any sense). They all have the same pattern. If there are three "opposite" pairs and one "solved" pair then it's not HTO possible, if there are four "opposite" pairs then it is.
-Now practice on a 2x2. Solve CO, then do Half-Turn-Only Reduction. Then move it over to this 3x3 method step HTOR, and ignore edges for now. Once you get the hang of that, then you can worry about edges.
-Edges are simpler, if it’s opposite colors, it’s reduced to HTO. If they're in their correct slices (and oriented), they're HTO possible. So the WG-edge for example, could be in four spots, solved, UB, DB, and DF.
-You can practice more by doing a R2 U F2 scramble, then just fix the edges. Then scramble it and just fix the corners. Then scramble and try to fix both. It’ll probably be too hard (it’s okay) so to do this in smaller steps, mix it up with R2 U2 F2 moves and then do like 3 moves to mess up HTOR and try to understand how to reverse it, Example: R2 U2 R2 F2 U’ R2 U HTOR: U’ R2 U, and see how it works, and maybe increase the moves to 5 or 6 once you totally understand everything so far. This idea of starting with just a few moves away and working your way up is actually be a practice technique applicable to more than just this.
I have no idea if you need all that, but good luck!

Thanks! I'll practice this a bit.

 

[WoowyBaby]

New Skewb Method - Sky Method

1. Build a TriBlock at UB (avg ~5? moves)
2. Centers + Corner Permutation (5 algs, intuitive, avg. 6 moves)
3. Orient Last 4 Corners (5 algorithms)

The goal of this method is to try to utilize the R' L R L' trigger, similar to how Sarah's method utilizes the Sledgehammer. I will name R' L R L' Sky move and L R' L' R A-Sky, unless there is already a name for this, then please tell me. I personally think that Sky move is exactly as fast as sledgehammer, and sometimes faster, because you have to do a little regrip before every sledge, but that is not the case with Sky. To prove my point with real life testing (more ppl should do this!), I can do triple Sky in 1.10 seconds but triple Sledge is 1.80 at best. I'm not sure why there hasn't been a method made around this move, probably because it doesn't preserve a layer/face like Sledgehammer and everyone's biased towards faces (not a joke).

Another goal of this method is to have a low alg count, which it does at only 10, but also can also become faster by combing steps/adding algsets (see advanced).

For this post, scrambles are in WCA notation while everything else, solutions and algorithms, are in rubikskewb notation (unless otherwise stated).
*However, z and z' rotations will be around DF/UB corners instead.
Sky = r' l r l' A-Sky = l r' l' r

Algorithms and Example Solves
~Algorithms~
If you ever don’t know how to hold the cube before you start it, apply inverse of algorithm to see.
Centers + CP Cases:
CP Only= Sky Sky Sky
C Only, CW= A-Sky A-Sky
C Only, CCW= Sky Sky
CP+CW C= Sky
CP+CCW C= A-Sky
1.5 Sky’s on avg for CP+Centers (incl solved+prob)
CO Cases:
Peanut= r' B' r b' r' B r B' b B -or- (z*) Sky Sky z A-Sky A-Sky
Sune= (z') A-Sky z A-Sky z A-Sky
A-Sune= (z) Sky z' Sky z' Sky
4-Twist, middle CW= F l' B' F' l' F r F' B l' r' -or- Sky alg?
4-Twist, middle CCW= B' l' r F' l' B r' F r l' r' -or- Sky alg?
Algorithms on a sheet w/pictures(yay!) can be found here.
I will try to improve the algorithms over time.
- It is possible to have every algorithm (even all of L7P) to be Sky/A-Sky combos, I just haven't figured them out yet.

So the 4-twist algs are currently the equivalent of bad NS algs from Sarah's....

~Example Solves~
Scramble: L' R B' R' U L B' R'
(y)
r' F L' f' L' (y') r' // TriBlock
Sky Sky Sky // C+CP
(z) r' B' r b' r' B r B' b B // Peanut

Scramble: R' U R U R' U B' L' U
(y x y')
f' r F' // TriBlock
A-sky z2^ Sky Sky Sky // L7P
3 normal moves + 16 moves in sky triggers = 19
(^rubikskewb z2)
This is kind of?(not rly) an example of why Sky is as good as Sarah's

More example solves coming soon

A random tip for building your TriBlock is to think about it as 3 center/corner pairs that you join together, OR just focus on 2/3 of it then think of the last pair. Thinking about it as 3 centers around 1 corner -> Expand won't work, because the other corners that part of the full TriBlock will be always be twisted, due to the nature of Skewb corner tetrads.

Variations
If you think building a full TriBlock is too difficult, then you might like Twisty Sky:
1. TriBlock minus 1 corner
2. Fix corner + CP (7 algs)
3. Centers + Orient L4C (~30 algs)
Was thinking on 2) Fix corner + centers 3) Last 4 Corners, but I think it's worse.
Combining the 2nd and 3rd steps of Sky to make a 2-step method is Advanced Sky Method:
1. TriBlock, same as normal
2. L7P (~90 cases)
Still easier to learn than Sarah’s advanced, because they are both a both series of Sledge/Sky triggers, but this is 90 vs. 130 cases.
I will have an algsheet here for all L7P cases.
- Just realize I purposefully split L7P into 2 parts for the Normal Sky Method in the most even (+best?) way possible, it’s only 10 algs. Any other 2-step way either has 30+ algs or weird recog.

I should add a explanation/tutorial video soon, but can't promise anything right now- (video link once I do it)

Also I should actually TEST IF THIS IDEA IS GOOD before throwing it out there? I know, I know, it can be annoying to just have someone post a stupid 3x3 method idea they came up with in 1 minute that has zero potential or usefulness, but I think this is different.

I believe this will be quite revolutionary to Skewb solving, what are your thoughts?

 

[DesertWolf]

New Skewb Method - Sky Method

1. Build a TriBlock at UB (avg ~5? moves)

I believe this will be quite revolutionary to Skewb solving, what are your thoughts?

I can't fully understand the TriBlock. What exactly are we creating?

 

[WoowyBaby]

I can't fully understand the TriBlock. What exactly are we creating?

Read more than the first two lines and you’ll easily figure it out.
Just to tell you, its the block that is preserved when you apply the Sky move, r’ l r l’.

 

[Deleted member 36445]

i'm impressed at your ability to find new ways to solve cubes

 

[DesertWolf]

Read more than the first two lines and you’ll easily figure it out.

It’s the block that is preserved when you apply the Sky move, r’ l r l’.

I see it now. Like on the second example solve i must've done a mistake cause i couldn't seem to get it. Also yes i went through more then the first two lines. anyway now that i have a better grasp of the method here's my opinion (doesn't have to be correct, it's an opinion): The biggest issue i see with it is recognition. Not to say it's bad but it's not as good as Sarah's advanced IMO. The way the TriBlock is built would be the reason. Obviously with the advanced version that is effectively fixed as one looking then becomes an option. Before i talk though about the advanced version i'll just say that i definitely see potential for the method and i think it can bring someone to very good times with low alg counts which is always good hehe. Now obviously if someone would want to be world class then advanced would be required. The question is how does it compare to Sarah's advanced? What worries me mainly are some of the algs. Yes doing a Sky alg is fast but the way some of them combine can be quite weird (taking those with more than 1/2 (if you include initial rotation) rotation as an example) . Also those that arn't solvable with a combination of sky's we'd have to wait and see how good that alg would compare to the current as some i feel arn't as good ergonomically as some of the non sledge and hedge sarah's advanced cases. Finally, it's a method definitely worth exploring but no real judgement which isn't only based on opinions can be made till someone actually goes through and starts actively practicing the method.

 

[WoowyBaby]

I see it now. Like on the second example solve i must've done a mistake cause i couldn't seem to get it. Also yes i went through more then the first two lines. anyway now that i have a better grasp of the method here's my opinion (doesn't have to be correct, it's an opinion): The biggest issue i see with it is recognition. Not to say it's bad but it's not as good as Sarah's advanced IMO. The way the TriBlock is built would be the reason. Obviously with the advanced version that is effectively fixed as one looking then becomes an option. Before i talk though about the advanced version i'll just say that i definitely see potential for the method and i think it can bring someone to very good times with low alg counts which is always good hehe. Now obviously if someone would want to be world class then advanced would be required. The question is how does it compare to Sarah's advanced? What worries me mainly are some of the algs. Yes doing a Sky alg is fast but the way some of them combine can be quite weird (taking those with more than 1/2 (if you include initial rotation) rotation as an example) . Also those that arn't solvable with a combination of sky's we'd have to wait and see how good that alg would compare to the current as some i feel arn't as good ergonomically as some of the non sledge and hedge sarah's advanced cases. Finally, it's a method definitely worth exploring but no real judgement which isn't only based on opinions can be made till someone actually goes through and starts actively practicing the method.

Thanks for your feedback! I appreciate that you actually look into it consider it and give me constructive criticism, unlike some people.
I do agree with you that Sarah’s still has an edge over Sky, so I don’t think this method will become very popular.

At this current moment it’s hard to tell how good it is, and yeah, I probably should’ve made more L7P algorithms and more example solves before posting that, once I work on this method more and do a bunch of test solves I’ll give you an update.

 

[Cogman09]

I have an idea for a new method and need feedback

Steps:
Cross
F2l-1
Orient edges
insert last f2l pair and solve Oll with one algorithm (the corner and edge do not need to be paired up)
Pll

would this be somewhat viable?

 

[WoowyBaby]

1) This belongs in New Method / Idea Thread
2) This is literally Winter Variation (Edit: ZZ-C OLS)
Winter variation is a useful set, don’t get me wrong, it’s definitely viable, it’s just not a new idea :/
Edit: Wait I’m sorry! You said it doesn’t have to be paired up, which would be ZZ-C OLS which has around 500 algorithms.
Just by the way, last slot methods have been explored and are now basically wrung dry, but know that there are plenty of other kinds of ideas that the “rag is still full”!

 

[Hazel]

At this point, virtually every last slot+last layer method has been proposed and has either been turned into an alg set or deemed inviable. What you proposed is just using after the first 3 slots to force a Winter Variation case, which is slower than just inserting the slot and doing OLL regularly. Don't let this stop you from striving to look for new, good methods though