Commutators vs. M2

[cmhardw]

At the world competition I was racing Stefan and Joey Gouly using our sighted blindfolded systems. Also I know that a lot of people are switching to M2 for edges and R2 for corners. I have been trying to work on my commutator method to get it faster, but I have a feeling that M2 might be faster for edges. I still believe in my corner commutators, though I am keeping an open mind about switching to R2 if my method seems to be slower.

In the interest of friendly competition I did some sighted solves with my blindfolded method, as well as some timed solves while blindfolded. For the sighted solves I inspected the cube for 5 seconds or so before I started. For the blindfolded solves I memorized without the timer, so I probably went a little slower than in a real solve, but I still tried to go relatively quickly to make it fairly realistic.

I timed each phase of the solution as well. For solves where I had parity the time for edges and for corners represents how long it took me down to leaving the last 2 pieces that have to swap. I timed how long it took me to execute the parity fix too.

Sighted solves:
---------------
edges / corners / parity (if applicable)
1) 25.29 / 24.24 / 8.64
2) 29.03 / 17.62 / 7.83
3) 26.30 / 23.19 / 10.70
4) 24.92 / 26.86 / 7.55
5) 25.89 / 25.14 / 8.47
6) 32.27 / 18.71 / ---
7) 29.99 / 14.06 / 4.98
8) 24.89 / 18.42 / ---
9) 31.12 / 16.10 / 12.03
10) 25.32 / 25.63 / 5.83
11) 33.69 / 16.51 / ---
12) 29.16 / 16.36 / ---

Edges average: 27.93
Corners average: 20.19
parity setup and execution (when applicable): 8.25

Blindfolded solves:
-------------------
edges / corners / parity (if applicable)
1) 26.28 / 17.87 / 13.70
2) 22.95 / 17.40 / ---
3) 23.62 / 22.59 / 10.75
4) 24.79 / 30.94 (corners and parity together) / see previous entry

For the 4th solve I forgot to hit the spacebar after solving corners down to the last 2, and I just went straight into the parity execution.

I can do more blindfolded solves, but right now it's late and I am getting tired. I just wanted to have a comparison to sighted and actually blindfolded for at least a few tries.

Please if any of you M2 or R2 users are interested post your times as well. I'm curious which method is faster for each piece type, or if one method is better overall.

---edit---
I don't mean to exclude Turbo, in fact it would be great if anyone using Turbo would also be interested in trying this friendly challenge too. I only personally know of Erik who uses Turbo, but anyone who does please feel free to join this as well.
---edit---

Thanks,
Chris

[masterofthebass]

Chris,
I don't consider myself too fast at BLD in most respects. My PB is 1:48 and I average around the 2:30s. My memo is over 1:15 usually (I think), and I use M2 and the old Pochmann Corners (Y perm w/ set-ups). I will time myself later when it's not 3:45 in the morning and try to contribute a little. I think this can be a really good thread for people considering BLD methods.

[joey]

I just did an average of 5 for sighted solves using old pochmann (T,Y and J). Although sometimes I might throw in an easy 3-cycle.

1)L D2 L' B' L F' R F2 R' B2 R F2 L' R' D2 L' U' F U2 L' D' L D' R U'
        Corners  Parity  Edges    Total
--------------------------------------------
Time:   28.08      4.58    54.84    87.50
Moves:  113        25        218      356
Tps:    4.02       5.46     3.98     4.07

2)U2 D F2 U L2 R' D B2 D' U B2 U' L U2 R2 F2 B U R L D L' F D2 F2
        Corners  Parity  Edges    Total
--------------------------------------------
Time:   23.29      4.27    39.78    67.34 (used two 3-cycles)
Moves:  115        25        147      289
Tps:    5.02       5.85     3.70     4.29

3)L' F' B2 D' R L B2 R' U B2 D' F2 R2 B L' R2 F2 U2 D' R F2 L D2 F' L
        Corners  Parity  Edges    Total
--------------------------------------------
Time:   26.96      0.00    32.02    58.98 (used one 3-cycle)
Moves:  130        00        183      313
Tps:    4.82       0.00     5.72     5.31

4)F L D U' B R2 F R2 B' L' B' L2 R' F2 R' U' L' B' F' U L D2 R2 U' D
        Corners  Parity  Edges    Total
--------------------------------------------
Time:   20.88      0.00    34.58    55.46
Moves:  097        00        156      253
Tps:    4.65       0.00     4.51     4.56

5)U L' B' U' L F2 L R' U2 F2 R' D2 R2 B2 R D2 U' L' U L' F B R B2 L
        Corners  Parity  Edges    Total
--------------------------------------------
Time:   23.12      3.74    36.46    63.32 (used two 3-cycles)
Moves:  097        25        164      286
Tps:    4.20       6.68     4.50     4.52

Average
        Corners  Parity  Edges    Total
--------------------------------------------
Time:   24.47      4.20    39.54    66.52
Moves:  110        25        173      299
Tps:    4.54       6.00     4.48     4.55

This is why I _really_ need to change method. It uses waaay too many moves.
I am actually in the middle of switching methods. Basically I am learning every method, then I can be flexible, and pick the best way to execute. Hopefully this will allow me to execute in around 40s, or below, which is my goal.

Like this example: , http://uk.youtube.com/watch?v=jcOPR6N2Yl8 if for example I could execute in 45s, the total time would have been ~1:20, instead of 1:48. A definite improvement.

[masterofthebass]

Ok I did my Sighted solves, but they're not too great. Joey kicks my butt with these. I figured in parity into the edges because it's the same algorithm every time, not like 3-cycles where you have to figure out which PLL to use.

Solve 1:
Edges 37.61 Corners 30.68

Solve 2:
Edges 29.88 Corners 35.11

Solve 3:
Edges 36.33 Corners 21.18

Solve 4:
Edges 34.19 Corners 44.90

Solve 5:
Edges 34.56 Corners 37.46

My sighted times may actually be worse than my BLD times, because I have no idea what piece is coming up next. For my Memo'd solves, I'm not going to memorize the entire cube, but rather each step, in order to make sure the executions are executed as best as I could do them.

[Erik]

ok I'll post my times with Pochmann, TuRBo and M2 here. Though for the corners I use the same method for all (I haven't learned all TuRBo algs for those).
so first Pochmann:
sighted solves:
Corners, parity+Edges, Total
1. 23.97 31.19 , 55.15
2. 24.03 50 , 74.03
3. 24.22 37.25 , 61.47
4. 19.83 44.63 , 64.46
5. 23.39 32.19 , 55.58
M2
sighted solves:
Corners, parity+Edges, Total
1. 19.02 20.81 , 39.83
2. 24.03 24.22 , 48.25
3. 23.45 29.75 , 53.20
4. 21.45 24.53 , 45.98
5. 21.56 21.34 , 42.90
TuRBo
sighted solves:
Corners, Edges+parity, Total
1. 22.39 29.94 , 52.33
2. 21.66 24.36 , 46.02
3. 20.91 37.52 , 58.43
4. 23.31 23.61 , 46.92
5. 23.61 29.02 , 52.63

[ExoCorsair]

M2
sighted solves:
Corners, parity+Edges, Total
1. 19.02 20.81 , 39.83

O_O

So would you say that M2 is the fastest?

[Pedro]

ok I'll post my times with Pochmann, TuRBo and M2 here. Though for the corners I use the same method for all (I haven't learned all TuRBo algs for those).
so first Pochmann:
sighted solves:
Corners, parity+Edges, Total
1. 23.97 31.19 , 55.15
2. 24.03 50 , 74.03
3. 24.22 37.25 , 61.47
4. 19.83 44.63 , 64.46
5. 23.39 32.19 , 55.58
M2
sighted solves:
Corners, parity+Edges, Total
1. 19.02 20.81 , 39.83
2. 24.03 24.22 , 48.25
3. 23.45 29.75 , 53.20
4. 21.45 24.53 , 45.98
5. 21.56 21.34 , 42.90
TuRBo
sighted solves:
Corners, Edges+parity, Total
1. 22.39 29.94 , 52.33
2. 21.66 24.36 , 46.02
3. 20.91 37.52 , 58.43
4. 23.31 23.61 , 46.92
5. 23.61 29.02 , 52.63

M2
sighted solves:
Corners, parity+Edges, Total
1. 19.02 20.81 , 39.83

O_O

So would you say that M2 is the fastest?

yeah, that's what I was going to ask

[cmhardw]

Ok I have to admit, M2 is very fast. Also Erik, do you think that Turbo will overtake M2 with more practice? I think probably the reason M2 is faster for you now is because Turbo is a newer method? It seems Turbo is already fairly fast as well though.

As to commutators. I am now going through all my possible cycles out of my buffer and categorizing them into types. There are 440 different possible cycles, but they are of a much smaller number of "types" of cycles. I am also working on triggerizing all of my commutators rather than just executing most of them with wrist turns.

If this does not speed up my edges to be as fast as M2, then I give in. If you can't beat them join them ;-)

But for now I am trying as best as I can to catch up my commutator method to M2 and Turbo and all the other fast methods, for edges at least. I do feel like my corner method is en par with you guys, but I will work on triggerizing all of my cycles as well to get rid of my slower outlier solve times. I have found that for some edge cases I was using long setups turns where no setup turn was even needed. Once I have practiced a bit more I will post my times again in this thread with more sighted solves.

I hope to catch up to M2, and also what it seems that Turbo will become. If I cannot then I will switch to either M2 or Turbo. I guess I'll give myself a deadline of March of next year, which from what I understand is most likely when the Chattahoochee competition will be, or the next tournament I am attending basically.

Happy BLD cubing everyone,
Chris

[Erik]

I have no idea, both method's have their pro's and con's
TuRBo is more effective but requires more thinking, M2 is a bit more moves but doesn't really require thinking inbetween. Maybe a hybrid would work? Solve with M2 but the easier cases solve with a 3 cycle like TuRBo?

[AvGalen]

I am wondering if looking at a cube is a bad thing now. Everyone agrees that looking ahead is really important but Mátyás just solved an entire cube in 38.40 seconds (including memo). Even with only 10 seconds memo that means his blindfolded solves is much faster than everyones sighted solves here. What is his secret?