Fridrich with early edges orientation

[DavidWoner]

6gen F2L >>>>> 3gen f2l.

 

[Kirjava]

On my opinion, this method is not worse than Petrus. Is it? Petrus is a decent method, I guess.

Statements like these will do well to discredit your ideas.

 

[Pyjam]

You're right. I rephrased my statement.

 

[Anonymous]

6gen F2L >>>>> 3gen f2l.

Elaborate?

 

[Kirjava]

EO after 2x2x3 and EO after cross are two extremely different situations.

 

[Pyjam]

EO after 2x2x3 and EO after cross are two extremely different situations.

Could you explain why, please? I'm sincerely interested.

 

[Kirjava]

Recog is easier for EO after a 2x2x3, it produces 2gen and not 3gen, you can quite easily recog/fix two bad edges while connecting a 1x2x2 to a 2x2x2. There are other minor advantages.

The greatest skill of a truly good method developer is the ability to recognise when an idea is poor and abandon it.

 

[Pyjam]

— Recog is easier for EO after a 2x2x3.

I don’t understand why. I tried both situations and, for me, it’s the same thing.

— it produces 2gen and not 3gen

Good point.

— you can quite easily recog/fix two bad edges while connecting a 1x2x2 to a 2x2x2.

I can do the same while finishing the cross.

— There are other minor advantages.

I believe you.

— The greatest skill of a truly good method developer is the ability to recognise when an idea is poor and abandon it.

Certainly, but fortunately I have no such ambition. The cube is also a game and I would like to play with my ideas.

Thank you for taking the time to answer my question.

 

[FatBoyXPC]

The greatest skill of a truly good method developer is the ability to recognise when an idea is poor and abandon it.

You guys should try not to take offence to that statement. He's had a similar conversation (with sincerity) a few times prior to this thread. When it comes to developing methods / learning them, Kirjava is definitely a good resource for help.

Don't forget guys that there are alternate insertion methods for pairs that also allows for partial edge control. Also don't forget that you can't have just one flipped edge, so you can use this to help predict bad edges.

Alternate insertion method: R' F R F' (Sledge Hammer)
Alternate #2: F R' F' R (I believe this is called Hedge Slammer?)
Alternate #3: r U R' U' M

PeterNewton: That Dilbert strip, priceless!

 

[Pyjam]

I feel no offense. I don't consider myself a “method creator”, neither a good speedcuber. I'm playing with my cubes and some ideas. I will not abandon them because someone says they're bad. It doesn't forbid to discuss them, and to learn more about other methods.

Alternate insertion method: R' F R F' (Sledge Hammer)
Alternate #2: F R' F' R (I believe this is called Hedge Slammer?)
Alternate #3: r U R' U' M

I use them (and also (F' r U r') on the left side, and (f R' f') in the rear). I also added (U2) l R U' R' U l' and (U') F R' F' R2 U R' to the Wiki…

…and 10 algo for other cases. Some more useful, I hope.

 

[uberCuber]

cyoubx, if 6 edges are already too difficult for a beginner, what about 12?

The difference here is that the 12 can be looked at during inspection, but the 6 (8?) must be inspected during the solve.

 

[teller]

It's a good type of exercise, regardless. Recognizing edge orientation now, rather than later, is a good thing.

 

[DavidWoner]

Elaborate?

A lot of cases in the <R, U, L> moveset are long and nasty, and you are so restricted. RUL is only advantageous if you have the blockbuilding freedom allowed by ZZ.

 

[Kirjava]

— Recog is easier for EO after a 2x2x3.

I don’t understand why. I tried both situations and, for me, it’s the same thing.

Less pieces, less rules for what makes a 'bad edge'.

— you can quite easily recog/fix two bad edges while connecting a 1x2x2 to a 2x2x2.

I can do the same while finishing the cross.

Not as easily.

— There are other minor advantages.

I believe you.

Think of half the cross edges as a restriction. Wouldn't it be better to do LineEO (not EOLine) instead? This way is just setting yourself up for a higher movecount. There's a reason why "Place DR" isn't a step in Petrus.

This is a bit of a nutty alternative to suggest for EOLL, don't you think?

 

[Anonymous]

A lot of cases in the <R, U, L> moveset are long and nasty, and you are so restricted. RUL is only advantageous if you have the blockbuilding freedom allowed by ZZ.

Oh I see what you're saying. I thought you were including ZZ in your statement.

 

[Pyjam]

Less pieces, less rules for what makes a 'bad edge'.

7 edges to look at for Petrus, 8 for “my method”, no more than 6 has a bad orientation in both case. And I do not have to deal with the DR edge. I'm more trained for my method and actually I find it easier.

I'm convinced that Petrus is a better speedsolving method. I say: if you're trained enough (I think of weeks of training), EO in my method isn't that difficult.

Blockbuilding is only a torture for my brain. You really have no idea of my difficulty to build a 2x2 block. I almost can solve the cube by the time I build one. :fp Planning 12 EO during inspection is too hard for me.

But, there are two things I can: cross planning during inspection and 8 EO during the solve. Therefore “my method”.

Think of half the cross edges as a restriction. Wouldn't it be better to do LineEO (not EOLine) instead? This way is just setting yourself up for a higher movecount. There's a reason why "Place DR" isn't a step in Petrus.

This is a bit of a nutty alternative to suggest for EOLL, don't you think?

I don't understand everything. I tried 1/2 cross during inspection but it's not enough. I can do more during inspection but not EO. 3/4 cross + partial EO is the next step.

 

[Kirjava]

7 edges to look at for Petrus, 8 for “my method”, no more than 6 has a bad orientation in both case. And I do not have to deal with the DR edge.

You still have to look in more places and detect edge orientation in more than two layers. This alone makes it harder. Not having to deal with DR is offset by the massive burden of having to look for EO on /all/ midges. There are more cases to learn because there are more places the edges can be in. EO will take more moves on average after cross as opposed to after 2x2x3.

I'm more trained for my method and actually I find it easier.

Bias.

I'm convinced that Petrus is a better speedsolving method. I say: if you're trained enough (I think of weeks of training), EO in my method isn't that difficult.

CPLS isn't difficult. It's not good, either.

Blockbuilding is only a torture for my brain. You really have no idea of my difficulty to build a 2x2 block. I almost can solve the cube by the time I build one. :fp Planning 12 EO during inspection is too hard for me.

But, there are two things I can: cross planning during inspection and 8 EO during the solve. Therefore “my method”.

I don't understand everything. I tried 1/2 cross during inspection but it's not enough. I can do more during inspection but not EO. 3/4 cross + partial EO is the next step.

Saying you use something because you can't use something else isn't a good way to advocate a method. But, I know you're not trying to advocate it, so no worries

I'd be interested in seeing some movecount stats from you of using EO vs not using EO. I know that the movecount for CFOP will be lower than this method (even with 2LOLL), but maybe you're gonna produce some funky statistics

This thread kinda reminds me of our foray into RouxZZ. It just doesn't work.

 

[Pyjam]

Movecount isn't my priority.
Depending on the location of the bad edges, EO can be fast or not.
The worse cases aren't those with more bad edges.
4 bad edges on top: F (R U R' U') F' f (R U R' U') f' (12 moves)
8 bad edges: (R2 U R' U' M2' D L' D' r2) y (9 moves)

A couple of examples.

====
Yellow on top, Green on front.
Scramble: B2 R U B' L F2 L F2 B' L2 B' F2 U R' L2 U2 R2 D B' R D2 L D2 R D2
Cross: u L2 u U F R'
I look at the equatorial slice around the cube. Left would be a perfect rear side because FL and BL are oriented: yellow and red beside the orange center. On the right side, both edges are bad (if orange is the back): yellow is on the front, bleu is beside the red center. On the top, 2 good and 2 bad (UR: yellow on the side, UB: green on top). It's an easy case.
EO: R2 U R' U' R2 (y)
It's done. 4 bad edges, 5 moves.
====
Scramble: F2 B D' L U' B2 U L F L2 D2 B F2 L B2 F2 L2 R' B' U2 L' U2 B' U2 L
Cross: L u F R' L' F2
Look around the cube. I see 2 yellow on opposite sides: FL and BR. FR and BL are bad. On top, 2 bad edges. Easy case again.
(F' U' F) u2 (F' U2 F)
It's done. 4 bad edges, 6 moves (u2 completes the cross).
====
Scramble: F2 D U' R' L' F' B U B2 R2 B D2 U2 L2 R' F' B U L' F' U B2 L2 B D2
Cross: U D L D R B'
Look around the cube. Right should be on the rear.
Bad edges: FL + 3 on top. Not a handy case when the bad edge is on the right after rotation.
(y) F' U' (L' U' L U) F u2
It's done. 4 bad edges, 7 moves (u2 completes the cross)
====
Scramble : R2 D' F2 R D' L D R2 D2 B' L' U' F2 U' F' L' R F U' F2
Cross : D' U' R' F B' u2 L2
Look around the cube: 3 bad edges. On top: only 1 if we put Orange/Red on front.
(L U2 L') u2 (y) (R U2 R' F U' F)
4 bad edges, 10 moves.
If we choose Green/Bleu on front :
(u2 F' U' F) (y) (U R2 U R' U' R2) (y)
6 bad edges, 10 moves.

 

[Kirjava]

The worse cases aren't those with more bad edges.
4 bad edges on top: F (R U R' U') F' f (R U R' U') f' (12 moves)
8 bad edges: (R2 U R' U' M2' D L' D' r2) y (9 moves)

4 bad edges on top: L' R' B' R2 U2 B L R (8 moves)

 

[Pyjam]

Thanks.

So, it's probably never more than 9 moves, and maybe average 6 or 7 moves. First of 2-look OLL is 6 moves on average.