[Help Thread] Roux Discussion and Help

[Knut]

Thanks a ton for all the help.
I just got a 14.81 and smashed my previous Roux PB and almost beat my CFOP PB. It was a CMLL skip, but I turned pretty slowly and look ahead was nice. Now to go practice more look ahead...

@Dan Cochrane: No, I don't know full CMLL yet, but I know all of the H cases and several other random ones. Still working on that.

 

[Dan Cochrane]

Haha, same. Know all but one h case and that's about it [emoji14]

 

[cashis]

I just got 19 STM for F2B. lost scramble though, but wow.

 

[theROUXbiksCube]

I just got 19 STM for F2B. lost scramble though, but wow.

dang

 

[cashis]

dang

the rest was bad though, so oh well.


On a different note, what are ideal splits for move count? Mine currently are something like 12/20/idk cuz i don't count/28

 

[theROUXbiksCube]

9 for FB, 15-17 for SB, <17 for LSE iirc

 

[cashis]

thinking of going for algorithmic approach to LSE EO. I suck at it, thats where most of the moved come from in my lse. Thanks adam

 

[PenguinsDontFly]

thinking of going for algorithmic approach to LSE EO. I suck at it, thats where most of the moved come from in my lse. Thanks adam

for eo algs, check out alex lau's lse tutorial

 

[GuRoux]

the rest was bad though, so oh well.


On a different note, what are ideal splits for move count? Mine currently are something like 12/20/idk cuz i don't count/28

my movecount is around 8, 16-17, 10, 15-16,
ideally, the FB could be improved to 6-7 and the lse to 13-14 maybe. i think it would be pretty hard to get second block to 12 in a speedsolve, perhaps 14 is more feasable.

 

[NeilH]

Sure. Now, these weren't speedsolves, but solves where I went slow, so I could count the moves, so they are fairly lower than normal, but...
Scramble:U' F2 U' L2 B2 U' F2 D' F2 D U' R' U' B U2 F2 U' B2 F' U F

FB: y' U L' U M' L U2 F' R' U' R2 U B2

SB: r' U' M' R2 U2 r' U' R U M' U'M U2 r U R'

CMLL:U L'U' L U' L' U2 L

EO:U2 M U2 M' U M

LR: M U2 M U2 M2 U'

M-Slice: U' M E2 M' E2

I use CFOP as my main, but wanted to try this scramble with Roux because I might want to switch. Just F2B

y' /inspect

r2 U M2 U2 L U' L' U r U F' U F' /FB

U r' U' r' U' r M2 U2 R U' R' / SB

 

[RomFrta33]

y //inspection
U2 M2 F' B L U x' r2 B' //FB
U r' U M' U2 r' U' r2 U' r2 R U2 r U r' //SB

 

[Ross The Boss]

I did some I' dat dere solve analysising for an average of 12 I filmed. The start and end of a step is the time at which I preform the first and last moves of that step. Here is the average:

[video]https://youtu.be/dYeiQyN5o6U[/video]

I'll come back with a solve by solve breakdown once I regain inthernet access on the computer I did this on, but for now, here is the averages (all of which are approximations save the 'actual').

Actual average: 11.36

FB avg: 2.11
SB square: 1.35
SB pair: 1.09
CMLL: 1.74
LSE: 2.68
avg non-turning/pause time: 2.30

Theoretical average: 11.27


Looks to me like most of the pausing comes with F2B. I need to get them more snappy and learn multi angle CMLL recognition. That, theoretically, will make sub10ish

 

[GuRoux]

I did some I' dat dere solve analysising for an average of 12 I filmed. The start and end of a step is the time at which I preform the first and last moves of that step. Here is the average:

https://youtu.be/dYeiQyN5o6U

I'll come back with a solve by solve breakdown once I regain inthernet access on the computer I did this on, but for now, here is the averages (all of which are approximations save the 'actual').

Actual average: 11.36

FB avg: 2.11
SB square: 1.35
SB pair: 1.09
CMLL: 1.74
LSE: 2.68
avg non-turning/pause time: 2.30

Theoretical average: 11.27


Looks to me like most of the pausing comes with F2B. I need to get them more snappy and learn multi angle CMLL recognition. That, theoretically, will make sub10ish

that's pretty cool, i should do that with my videos, i wonder what percent pauses i have. you turn so smooth but have 20% pauses time? it doesn't really look like it.

for lse, if you paused in lse did you count that toward your pauses or lse? are the pauses just in between steps?

edit: from my 8.24 average of 5, no where near reflective of what i usually solve since it was very lucky, but fun to analyze

fb: 1.7
fb->sb: .24
sb sq: 1.58
sb sp->pr: .05
sb pr: .76
sb->cmll: .64
cmll: .82
cmll->lse: .46
lse: 2.16
total: 8.3
solving: 7.0
transition pauses: 1.3
percent pauses: 16%

so the pauses sb->cmll and cmll->lse amount to 80ish% of all the pauses in my solve, probably in regular solves it's aobut 70%, anyone know how to recognize cmll faster?

 

[Ross The Boss]

Just in between my steps. Most of the time, rather than just down right stopping
I just slow down even more on the final turn like when I'm moving the square out of the way or positioning the first block. I counted those little crawls as a pause. Because really it just gives the illusion of flow once you slow your turning enough to the point that there is really no practical advantage to it

 

[GuRoux]

Just in between my steps. Most of the time, rather than just down right stopping
I just slow down even more on the final turn like when I'm moving the square out of the way or positioning the first block. I counted those little crawls as a pause. Because really it just gives the illusion of flow once you slow your turning enough to the point that there is really no practical advantage to it

yeah, i guess those are pauses, but you seem pretty good at deducing lookahead stuff, you don't really need to turn the cube around or look at the bottom to do lookahead which is really good. prehaps try moving a little faster, you might be surprised that your lookahead can keep up.]

edit: oops, just watched the average of 12 and you do turn the cube around to lookahead a lot, try to minimize that type of stuff.

 

[Ross The Boss]

anyone know how to recognize cmll faster?

I'm not exactly good at it, but I think that might be a time where slowing down at the end of sb pair insertition could actually be helpful

 

[GuRoux]

I'm not exactly good at it, but I think that might be a time where slowing down at the end of sb pair insertition could actually be helpful

yeah, but that's like your illusion of no pauses thing.

 

[Ross The Boss]

yeah, i guess those are pauses, but you seem pretty good at deducing lookahead stuff, you don't really need to turn the cube around or look at the bottom to do lookahead which is really good. prehaps try moving a little faster, you might be surprised that your lookahead can keep up.

I've actually been trying really hard lately to boost my max tps. I think I'm turning well faster in my solves than I was a week or two ago

 

[PenguinsDontFly]

3.75 slow-turning LSE ao 100 with new turning style (thanks guroux)

 

[GuRoux]

3.75 slow-turning LSE ao 100 with new turning style (thanks guroux)

i thought your lse is like 2.5, doesn't look like you should be thanking me.