# Optimal Pyraminx Method?

#### Carrot

##### Member
Hello =D

I tested the move count of my pyraminx method (WO, the pyraminx method I'm best with)... and I used a little bit over the WCA inspection to be sure I found the (maybe?) best way to do step 1 (I solved it mostly 2-step... so yeah)

I am now thinking, what method is closest to the optimal solution? and if you are good at an existing method, please count your moves and do an average (with qqtimer you can just count the length of the scramble to see what the optimal solution is for the scramble...) so we can compare...

If you have an idea of a method that might be very close to the optimal solution, please tell me the method's steps.. then I will learn it and do an average, unless you do it by yourself (I will learn it anyway lol)

I will start attributing with my average using WO (the move counts in blue are scrambles where I had an optimal solution ):
1. 10[8] B L B' R' U L' B' R'
2. (7[7]) U' L R U' L' R L'
3. 9[7] L' R B' U R' B' L
4. 8[8] R' U' L R' B L' U' L'
5. 13[8] U B' U B' R' U' B' U
6. 9[8] B U' L B' L B R' L'
7. 9[8] R' U L' B R U' L U
8. 10[8] U L' R' U B U R' L
9. 11[8] U L B' R' U B' L U
10. 12[6] L R U' L B U' :fp
11. 10[9] U B R' B' R L' B L' R
12. 10[8] B L' B U' L B U' L

this leaves with an average of 9.8 moves and with an average of 1.8 moves away from optimal solution... (remember possible cancellations... and don't count the tips...)

Last edited:

#### SimonWestlund

##### Member
Amazing Oscar.

My average was 14.2 moves with your scrambles

My best solution was 9 moves on the 5th scramble.

5. U B' U B' R' U' B' U 9[8]

Solution: U B R B' R L' R L R

#### Carrot

##### Member
Simon, You are using double-x-mitsubishi/nuclear/cross/flower/something and then trying to force a LL skip?

because normal LBL would not allow you to do that

I will try see if I can reconstruct my solves later today =)

#### flee135

##### Member
I got 9 on number 10 using simple OKA, but only because I skipped the last layer. It'd be interesting to see your thought process through these, because I cannot see great solutions for the first step every time, much less in the 15-second inspection limit.

#### Carrot

##### Member
I got 9 on number 10 using simple OKA, but only because I skipped the last layer. It'd be interesting to see your thought process through these, because I cannot see great solutions for the first step every time, much less in the 15-second inspection limit.
Well, I'm if I should be within the 15 seconds inspection I won't look any deeper than the first three starts... and I might not even search them deep enough to find the optimal thing for them... so I'm like 12-15 average if you take my speedsolves... but the algs in those are speed-optimized, and not optimal ;D

So by this thread I'm trying to ind a fewest move method, and not a speed-fewestmoves... if you for-example take my petrus style pyraminx method you'll see that I can see the first three moves in inspection, and when those are performed I'm counting the wrong edges... with my fingers -.-' and then after like 10 seconds I can continue in slowmotion

----------------------------------------------------------------------------
For Simon:

I tried to reconstruct them and got this:

1: B' U B' U R' B R' B' R (9/8) well... there was two nearly identical starts on this scramble... so I chose the 1-move shorter top hehe
2: L R' L U R' L' U (7/7)
3: B R B' R U' B' L R L' (9/7)
4: ??
5: U B R B' R L' R L R (9/8)
6: ??
7: L' B L' U' L' B' U' B' U (9/8) I found the same as Simon hehe
8: U R B R B L B' L' B U' (10/8)
9: R U' B' L B U R U' R' U B' (11/8)
10: U L B' L U R' B' R' B R L' (11/6)
11: R' L B' L' R' B R B' U' (9/9)
12: ??

for the ?? I couldn't find any sub 12 solutions ... weird :confused: but whatever... my overall average with this method is about 10-12.5 hehe and I'm color neutral+I know way to many tricks to do the same solve two times in a row ;D so I'm happy that I could get 9/12 of them the same move count or lower =)

Last edited:

#### Brunito

##### Member
17 19 15 15 14 (9) 10 15 11 15 21 (22)

it is a big sh*t. and whit this method my average is under 6 or 5 and official 5.31 hehe

#### Carrot

##### Member
17 19 15 15 14 (9) 10 15 11 15 21 (22)

it is a big sh*t. and whit this method my average is under 6 or 5 and official 5.31 hehe
but your turning speed are good though =) my turning speed is very slow :fp

#### SimonWestlund

##### Member
Simon, You are using double-x-mitsubishi/nuclear/cross/flower/something and then trying to force a LL skip?
That's exactly what I call my method!

Thanks for the reconstructed solves as well.

#### Carrot

##### Member
Simon, You are using double-x-mitsubishi/nuclear/cross/flower/something and then trying to force a LL skip?
That's exactly what I call my method!

Thanks for the reconstructed solves as well.
Simon, that means you are using the polish V hehe =D

Last edited:

#### DavidWoner

##### The Punchmaster
1. B' U B' U R' / B R' B' R (9)
2. L' B' R U L' U / B' R L' R' L' B' L' (13)
3. R' U' B R' B R' / L R' L' R' U' R' U (13)
4. L' B L B L' R' U' L' / B R B' R' (12)
5. U B R' B' / L' R' L B' L B L' (11)
6. L R L U B' / R L' R' L (9)
7. L' B L' U' L' / B' U' B' U (9)
8. L' U R' U L U B / L' R L R' B (11)
9. U' R B U B / U L U' (8)
10. L' U' B' L' / R' U R U' R' (9)
11. U' L' B' R' / U B L' B' L' U L' U (12)
12. R B' R' U' B / R' L' R U' R U R' (12)

average 10/12 = 10.7
mean of 12 = 10.66
Distance from optimal = ~2.7

<3 optimal solution that is not inverse scramble

There are of course some cancellations here that I would not see going full speed, but they would only increase a few solutions by 2 moves.

my turning speed is very slow :fp
Mine is even slower >_> I want to burn this c4y pyraminx, send it straight to hell.

#### fanwuq

##### Member
1.8/8
2.14/7
3.13/7
4.12/8
5.12/8
6.15/8
7.12/8
8.12/8
9.8/8
10.14/6
11.13/9
12.13/8
LBL. 2 optimal forced LL skips.

#### Carrot

##### Member
1. B' U B' U R' / B R' B' R (9)
2. L' B' R U L' U / B' R L' R' L' B' L' (13)
3. R' U' B R' B R' / L R' L' R' U' R' U (13)
4. L' B L B L' R' U' L' / B R B' R' (12)
5. U B R' B' / L' R' L B' L B L' (11)
6. L R L U B' / R L' R' L (9)
7. L' B L' U' L' / B' U' B' U (9)
8. L' U R' U L U B / L' R L R' B (11)
9. U' R B U B / U L U' (8)
10. L' U' B' L' / R' U R U' R' (9)
11. U' L' B' R' / U B L' B' L' U L' U (12)
12. R B' R' U' B / R' L' R U' R U R' (12)

average 10/12 = 10.7
mean of 12 = 10.66
Distance from optimal = ~2.7

<3 optimal solution that is not inverse scramble

There are of course some cancellations here that I would not see going full speed, but they would only increase a few solutions by 2 moves.

my turning speed is very slow :fp
Mine is even slower >_> I want to burn this c4y pyraminx, send it straight to hell.
That is very nice David, are you still using WO? or what method are you using?

ohhh... Fanwuq, you are very consistent =D

I'm gonna try with some other methods later today

EDIT: just tried your solutions... AWESOME!!! =D it's some variant of the polish V? O____o

Last edited:

#### Brunito

##### Member
1. B' U B' U R' / B R' B' R (9)
2. L' B' R U L' U / B' R L' R' L' B' L' (13)
3. R' U' B R' B R' / L R' L' R' U' R' U (13)
4. L' B L B L' R' U' L' / B R B' R' (12)
5. U B R' B' / L' R' L B' L B L' (11)
6. L R L U B' / R L' R' L (9)
7. L' B L' U' L' / B' U' B' U (9)
8. L' U R' U L U B / L' R L R' B (11)
9. U' R B U B / U L U' (8)
10. L' U' B' L' / R' U R U' R' (9)
11. U' L' B' R' / U B L' B' L' U L' U (12)
12. R B' R' U' B / R' L' R U' R U R' (12)

average 10/12 = 10.7
mean of 12 = 10.66
Distance from optimal = ~2.7

<3 optimal solution that is not inverse scramble

There are of course some cancellations here that I would not see going full speed, but they would only increase a few solutions by 2 moves.

my turning speed is very slow :fp
Mine is even slower >_> I want to burn this c4y pyraminx, send it straight to hell.
i use the same method as david .odder

#### Carrot

##### Member
1. B' U B' U R' / B R' B' R (9)
2. L' B' R U L' U / B' R L' R' L' B' L' (13)
3. R' U' B R' B R' / L R' L' R' U' R' U (13)
4. L' B L B L' R' U' L' / B R B' R' (12)
5. U B R' B' / L' R' L B' L B L' (11)
6. L R L U B' / R L' R' L (9)
7. L' B L' U' L' / B' U' B' U (9)
8. L' U R' U L U B / L' R L R' B (11)
9. U' R B U B / U L U' (8)
10. L' U' B' L' / R' U R U' R' (9)
11. U' L' B' R' / U B L' B' L' U L' U (12)
12. R B' R' U' B / R' L' R U' R U R' (12)

average 10/12 = 10.7
mean of 12 = 10.66
Distance from optimal = ~2.7

<3 optimal solution that is not inverse scramble

There are of course some cancellations here that I would not see going full speed, but they would only increase a few solutions by 2 moves.

my turning speed is very slow :fp
Mine is even slower >_> I want to burn this c4y pyraminx, send it straight to hell.
i use the same method as david .odder
ohh... it looks like OKA (youtube style) now =D

#### DavidWoner

##### The Punchmaster
EDIT: just tried your solutions... AWESOME!!! =D it's some variant of the polish V? O____o
I just do w/e looks good to get down to last 3 edges. These just looked a bit easier with Oka-style start, although I don't fingertrick it Oka-style and I do tips at the end still. I don't know, these scrambles felt easy, I think my normal movecount it probably 13 or so.

#### Carrot

##### Member
EDIT: just tried your solutions... AWESOME!!! =D it's some variant of the polish V? O____o
I just do w/e looks good to get down to last 3 edges. These just looked a bit easier with Oka-style start, although I don't fingertrick it Oka-style and I do tips at the end still. I don't know, these scrambles felt easy, I think my normal movecount it probably 13 or so.
hehe.. ofc they felt easy xD I took my best average of 12 in a session of about 50

I don't know why people chose my scrambles xD (because you guys are too lazy? hehe )

okay to be serious... I actually also made up a method for pyraminx (old one) that has a maximum move count of: 3 + 6 + 7 = 16... and I guess average is 2.5 + 3 + 5 = 10.5... which is pretty awesome... the problem is that I never seem to find the optimal solution from the last case... but the move count for the last step is only a guess... so I might be wrong...

but for people who wanna know it:
solve a 1x2x1 block
solve the rest 2-gen..

The story behind this method:
I wanted a method that ALWAYS gave you an extended-flower/mistubishi/something, that why I build the block... and with a maximum of 3 moves (I think so... but if you can find a situation where you can't do it in less than 4 moves I would be very happy =) )...
Then I thought... gosh... I hate to do LL -.-' So I thought, why not just solve bad edges... then you also restrict it to 2-gen? huh? xD
And then I got like... NO!!! you can still get a sune... Then I realized that you by solving one of the top edges and then solve the bottom you will always get LL skip... well... not "optimal" but it's indeed not a bad method for fewest moves...

I made this method up in like 10 minutes =) (I call it FMC or Petrus Style )

Could anyone else please try this method and/or tell me if my move-counts is correct for this method?

#### miniGOINGS

##### Member
http://www.speedsolving.com/wiki/index.php/Pyraminx_Methods said:
Petrus
Step 1: Create a 'block' by matching one edge to the two adjacent corners, and also fixing those two tips.
Step 2: Orient the remaining edges, Petrus-style, so that they can all be solved using only U and R turns.
Step 3: Finish all of the edges using only U and R turns, then fix the last two tips on the U and R corners.

#### fanwuq

##### Member
Odder, you never heard of this method before? Tim Sun was using this over 2 years ago. The only problem I see with it is that the bad edges are a bit hard to recognize.

#### miniGOINGS

##### Member
Odder, you never heard of this method before? Tim Sun was using this over 2 years ago. The only problem I see with it is that the bad edges are a bit hard to recognize.
Yea, the only thing I could think of would be to have a fixed colour scheme. That might help with EO.