• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 35,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

The Solve Critique Thread

Aerma

Premium Member
Joined
Apr 1, 2017
Messages
1,195
Location
Galar Region
WCA
2015MANN01
YouTube
Aerma
Random example solve, tips would be appriciated! I average around sub-13, I know full OLL and PLL, COLL minus sunes and antisunes, and some random other cases from various subsets.

R F B2 U R D' R2 F B' R F2 D2 L2 D F2 U2 F2 D R2

y' U F' U L F' U r' U R // Cross (I know it's inneficient but that's what I'd do in a real solve)
L' U2 L U L' U' L // F2L 1
L U' L' R' U' R U' R' U R // F2L 2
R U2 R' L U' L' // F2L 3
R U R' U2 R U2 R' U' R U2 R' // Last slot with winter variation
U (G perm starting with R U R' y') // PLL
 

cubeshepherd

Member
Joined
Feb 18, 2018
Messages
1,972
WCA
2016STEE01
Random example solve, tips would be appriciated! I average around sub-13, I know full OLL and PLL, COLL minus sunes and antisunes, and some random other cases from various subsets.

R F B2 U R D' R2 F B' R F2 D2 L2 D F2 U2 F2 D R2

y' U F' U L F' U r' U R // Cross (I know it's inneficient but that's what I'd do in a real solve)
L' U2 L U L' U' L // F2L 1
L U' L' R' U' R U' R' U R // F2L 2
R U2 R' L U' L' // F2L 3
R U R' U2 R U2 R' U' R U2 R' // Last slot with winter variation
U (G perm starting with R U R' y') // PLL
First off, for the cross the last R should be a Rw so it would be (y' U F' U L F' U r' U r) Cross
And on your 4th F2L pair you needed a U at the beginning before the R.

With that being said, here is your original solve except once I added the needed moves the OLL is a SUNE case and for PLL I believe my version it is a little faster of a perm with no rotation to execute then the one that you were going to have:
https://alg.cubing.net/?setup=R_F_B2_U_R_D-_R2_F_B-_R_F2_D2_L2_D_F2_U2_F2_D_R2&alg=y-_U_F-_U_L_F-_U_Rw-_U_Rw//Cross L-_U2_L_U_L-_U-_L_//_F2L_1 L_U-_L-_R-_U-_R_U-_R-_U_R_//_F2L_2 R_U2_R-_L_U-_L-_//_F2L_3 U_R_U_R-_U2__R_U2_R-_U2R_U-_R-//F2L_4 R_U_R-_U_R_U2_R-//Sune_OLL U_R_U_R-_y-_R2-_Uw-_R_U-_R-_U_R-_Uw_R2// _ &title=Original Solution
y' U F' U L F' U Rw' U Rw//Cross
L' U2 L U L' U' L // F2L 1
L U' L' R' U' R U' R' U R // F2L 2
R U2 R' L U' L' // F2L 3
U R U R' U2 R U2 R' U2R U' R'//F2L 4
R U R' U R U2 R'//Sune OLL
U R U R' y' R2' Uw' R U' R' U R' Uw R2//

And here is a revised solve with your scramble. I believe that the solve that I came up with is a little more efficient, especially for the cross, but please take it with a grain a salt because I only average in the 12 seconds.
y' R' F L D' R//Cross
y' L U' L' U' R U R2 U R U2 R' U2 R U' R' U R//First and second pair
U2 L' U L U2 L' U L//Third pair
U L U' L' U2 L U' L'//Forth pair
R U2 R2 U' R U' R' U2 F R F'//OLL
U' x R' U R' D2 R U' R' D2 R2 x' U'// Aa Perm Pll
https://alg.cubing.net/?setup=R_F_B2_U_R_D-_R2_F_B-_R_F2_D2_L2_D_F2_U2_F2_D_R2&alg=y-_R-_F_L_D-_R//Cross y-_L_U-_L-_U-_R_U_R2_U_R_U2_R-_U2_R_U-_R-_U_R//First_and_second_pair U2_L-_U_L_U2_L-_U_L//Third_pair_ U_L_U-_L-_U2_L_U-_L-//Forth_pair R_U2_R2_U-_R_U-_R-_U2_F_R_F-//OLL U-_x_R-_U_R-_D2_R_U-_R-_D2_R2_x-_U-//_Aa_Perm_Pll &title=Revised Solution

All in all you solution was pretty good, especially once you add the right moves:) But I think that if you use the cross solution that I provided you can look ahead at your first F2L pair easier then doing what you did with the lower case r's. Once the cross was done the F2L's were good and your last layer is self explanatory.
 

xyzzy

Member
Joined
Dec 24, 2015
Messages
1,572
Disclaimer: I don't do OH Roux.

1. lol @ hand scrambles

2. It looks like you slam the cube against the table to align the layers very often, and this eats up a lot of time. Practise doing FB+SB without table abuse or hold your cube closer to the table surface. Getting a magnetic cube is also an option, although some people dislike magnets for OH so ymmv.

3. Minimise slice moves during FB/SB. Related to the above point because you need one edge of the cube to be resting against the table for slice moves, whereas you normally have it hovering about a centimetre off.

4. U3' for setting up to CMLL in the first solve is kinda bad. F U R U' R' F' should be about as fast as F R U R' U' F' and these two algs solve the same CMLL case but at different angles.
 
Joined
Sep 25, 2017
Messages
189
Location
Massachusetts
WCA
2016MARS09
<3

@cubingawsumness, I stand by what I said yesterday. Turn faster.

Here are my solves:


Pretty good for me, but definitely room for improvement.

(20.062), 23.462, 22.495, (27.753), 20.873 = 22.277

Lost the scrambles sorry.

I sort of went into Fridrich on the 27. Didn't work out well.

EDIT: the first, second and fifth solves were all clockwise A-perm. weird.
Thee finger tricks are a little bit choppy but nice job with no pauses!
 

CraZZ CFOP

Member
Joined
Jul 30, 2018
Messages
99
WCA
2018HUGE02
Sorry if this critique is incomplete but I am not at the level that you are by the long shot. (Ask somebody else for advice in cross. I use ZZ.)
F2L:
1) I noticed that you used U U instead of U2. This is small, but it can go a long way.
2) I have also seen you do U U', which is two wasted moves.
3) You have small pauses in between each case. At this level, you may want to have no pauses.
 

GWCuber

Member
Joined
Apr 13, 2019
Messages
4
Hope someone can give me tips on how to improve at 3x3

The Solve:

Currently averaging about 1 minute and PB: 36 seconds
 

DanRman

Member
Joined
Jul 23, 2019
Messages
6
I have been practicing quite hard on 7x7x7 during the past month and dropped from around 9 minutes to 6 ;). Now I average 6:30 and my PB is 5:43. But I have plateaued and haven't been able to get any better for a while.

here is an example solve. Any tips will be greatly appreciated.

centers: around 3:00-3:15
edges: ~2:30
3x3:~0:40

 

jakelevine

Member
Joined
May 8, 2011
Messages
25
WCA
2009LEVI01
Watch Kevin Hays walkthroughs and Tuesday tips videos on YouTube. They have helped me a bunch even though I'm not terribly faster than you.

Additionally, I see some unecessary slice back moves on your first center. If no other centers are solved, it isn't necessary to slice back. In general avoid slice moves when at all possible and use wide turns, it's much faster.
 

xyzzy

Member
Joined
Dec 24, 2015
Messages
1,572
I have been practicing quite hard on 7x7x7 during the past month and dropped from around 9 minutes to 6 ;). Now I average 6:30 and my PB is 5:43. But I have plateaued and haven't been able to get any better for a while.

here is an example solve. Any tips will be greatly appreciated.

centers: around 3:00-3:15
edges: ~2:30
3x3:~0:40

Note: t-centres are the "edge" centres, x-centres are the "corner" centres, and the obliques are the rest. I use SiGN notation here (3R = turn only the third layer from the right; 3r = 3Rw = turn three layers from the right), as I do everywhere else.

1. scramble your cube harder There are like a zillion blocks on the cube that haven't been broken up.
2. Unnecessary slicing-back moves on the first and third centres, as already pointed out.
3. Better solution for the second orange bar. (Probably easiest to see what's going on if you rotate the virtual cube to a back view.)
4. Efficiency is generally on point, except for the unnecessary slices thing. (If I wanted to nitpick I could point out a lot of tiny mistakes, but that would take way too much effort.)
5. At 1:47 you have 4 white bars solved on your left (the right in the video frame), and you have a 4/5 white bar (missing an oblique) on another face. You should prefer to AUF so that the missing oblique is next to the free slice (i.e. so that it's on the left); by doing this, you can use just a normal wide move to realign the 4 white bars you've already solved, rather than using a slice move (which you had to do because you had the missing oblique on the opposite side of the free slice).
6. At 1:53-ish, you can directly form the second blue bar with U' 3l' U2 3l (then insert with U 3r U2 3r'), instead of using M' U2 M to move the inner t-centre to the right spot. (Slice moves take two wrist turns to execute; they're not as slow as doing two separate wide moves, but they're definitely slower than doing only one wide move.)
7. (Last blue bar.) Learn cases 09 and 10 on Meep's L2C list. (Just because the list is for the last two centres doesn't mean that you can't also use those solutions on the other centres.) Also, try to avoid having holes in your bars as you form them; you need to use slice moves to insert pieces into the holes and slice moves are bad.
8. Last two centres. This is the only part I bothered to reconstruct fully, and some of what I say here will also apply to other parts of the centre solving.

Your solution (59 moves not counting a U U' and execution mistakes):
F' 3r U 3r' U' 3R U 3R' 3r U2 3r' // two bars (13 moves)
3-4r' F 3-4r l' U l r U r' // third bar (11 moves)
U' l' U l U l' U' l 3L' U2 3L // fourth bar (13 moves)
U' l' U' l U l' U' l U l' U2 l U' l' U' 3L' U l U' 3L // last bar (22 moves)
alg.cubing.net

Sample solution (roughly how I would've solved it; 37 moves):
F 3l' U 3l U' 3l' U2 3l // two bars (8 moves)
3r U' 3r' U r U l' U2 l r' // third bar (10 moves)
U 3r r' U2 r U2 r' U' 3R' // fourth bar (10 moves)
U2 3L F r' F' 3L' r // last bar (9 moves)
alg.cubing.net

Let's point out all the inefficiencies in your solution. The middle bar was already solved, so there isn't anything to say about that. You have all the pieces needed to form the second bar on the top face, ready to be joined in 3 moves (see the sample solution), but you used a different set of pieces to form a 4/5 bar with a hole (don't make partial bars with holes, as mentioned in point 7). While finishing the second bar, you could also have inserted it at the same time by doing 3R U' 3R'.

The third bar was very weird. I'm not sure if it's because you saw the 1×4 bar on the top face and wanted to preserve that. You generally shouldn't bother preserving blocks for the last two bars, since they show up by chance very frequently anyway. Instead of doing weird slice moves, you could have done U2 r U2 r' U r U2 r'.

Fourth bar was fine, but you could've inserted it with 3l' U2 3l instead of using slice moves 3L' U2 3L. Remember: wide moves are faster than slice moves. For the last bar, you could have used U2 l' U2 l to solve 3/5 of it, then use a single commutator to fix the last two pieces together. What you did was to join the x-centres to 1×2 pair; this can be done in 8 moves, but you used 12 here. (Again, check Meep's L2C site. This is the same case as what you had for the last blue bar.)

(Bonus: there's an even shorter solution that doesn't make use of commutators. This is pretty advanced, though.)

Also, if you're right-handed, you should preferentially keep the solved bars on the left side, so that you'll do most of the moves with your right hand.

9. You spent four seconds thinking about what to do between finishing centres and edge pairing. This isn't a bad thing per se, and as you do more solves you can try to make the transition more fluid.
10. Lots and lots and lots of cube rotations during edge pairing. If you want to do E-slice edge pairing, you must use flicks to do Uw/3Uw moves instead of wrist turns; if you want to do M-slice edge pairing, wrist turns are fine (not like you have a choice), but you shouldn't ever be doing any rotations other than x and x'; if you want to do S-slice edge pairing, you shouldn't.
11. Assuming you want to do E-slice edge pairing, because that's what most people do these days, if you can't find the edge pieces you need on the top layer, just do a z2 rotation to switch the top and bottom layers. You can use R D R' and similar inserts if you happen to notice the pieces on the bottom, but you mostly shouldn't bother peeking at the bottom layer.
12. stop rotating so much during L4E aaaaaaaaaaaa

Learn your M-slice flipping algs if you want to do M-slice L4E. U' R' U R' F R F' is the popular one, I believe.
13. Your OLL parity alg execution has like eight regrips; standard algs require only three or four. Look up fingertrick videos for OLL parity algs.
14. For two-look PLL, use T perm (or J perm) and Y perm to permute the corners. A perms are really bad on big cubes.
 

DanRman

Member
Joined
Jul 23, 2019
Messages
6
Note: t-centres are the "edge" centres, x-centres are the "corner" centres, and the obliques are the rest. I use SiGN notation here (3R = turn only the third layer from the right; 3r = 3Rw = turn three layers from the right), as I do everywhere else.

1. scramble your cube harder There are like a zillion blocks on the cube that haven't been broken up.
2. Unnecessary slicing-back moves on the first and third centres, as already pointed out.
3. Better solution for the second orange bar. (Probably easiest to see what's going on if you rotate the virtual cube to a back view.)
4. Efficiency is generally on point, except for the unnecessary slices thing. (If I wanted to nitpick I could point out a lot of tiny mistakes, but that would take way too much effort.)
5. At 1:47 you have 4 white bars solved on your left (the right in the video frame), and you have a 4/5 white bar (missing an oblique) on another face. You should prefer to AUF so that the missing oblique is next to the free slice (i.e. so that it's on the left); by doing this, you can use just a normal wide move to realign the 4 white bars you've already solved, rather than using a slice move (which you had to do because you had the missing oblique on the opposite side of the free slice).
6. At 1:53-ish, you can directly form the second blue bar with U' 3l' U2 3l (then insert with U 3r U2 3r'), instead of using M' U2 M to move the inner t-centre to the right spot. (Slice moves take two wrist turns to execute; they're not as slow as doing two separate wide moves, but they're definitely slower than doing only one wide move.)
7. (Last blue bar.) Learn cases 09 and 10 on Meep's L2C list. (Just because the list is for the last two centres doesn't mean that you can't also use those solutions on the other centres.) Also, try to avoid having holes in your bars as you form them; you need to use slice moves to insert pieces into the holes and slice moves are bad.
8. Last two centres. This is the only part I bothered to reconstruct fully, and some of what I say here will also apply to other parts of the centre solving.

Your solution (59 moves not counting a U U' and execution mistakes):
F' 3r U 3r' U' 3R U 3R' 3r U2 3r' // two bars (13 moves)
3-4r' F 3-4r l' U l r U r' // third bar (11 moves)
U' l' U l U l' U' l 3L' U2 3L // fourth bar (13 moves)
U' l' U' l U l' U' l U l' U2 l U' l' U' 3L' U l U' 3L // last bar (22 moves)
alg.cubing.net

Sample solution (roughly how I would've solved it; 37 moves):
F 3l' U 3l U' 3l' U2 3l // two bars (8 moves)
3r U' 3r' U r U l' U2 l r' // third bar (10 moves)
U 3r r' U2 r U2 r' U' 3R' // fourth bar (10 moves)
U2 3L F r' F' 3L' r // last bar (9 moves)
alg.cubing.net

Let's point out all the inefficiencies in your solution. The middle bar was already solved, so there isn't anything to say about that. You have all the pieces needed to form the second bar on the top face, ready to be joined in 3 moves (see the sample solution), but you used a different set of pieces to form a 4/5 bar with a hole (don't make partial bars with holes, as mentioned in point 7). While finishing the second bar, you could also have inserted it at the same time by doing 3R U' 3R'.

The third bar was very weird. I'm not sure if it's because you saw the 1×4 bar on the top face and wanted to preserve that. You generally shouldn't bother preserving blocks for the last two bars, since they show up by chance very frequently anyway. Instead of doing weird slice moves, you could have done U2 r U2 r' U r U2 r'.

Fourth bar was fine, but you could've inserted it with 3l' U2 3l instead of using slice moves 3L' U2 3L. Remember: wide moves are faster than slice moves. For the last bar, you could have used U2 l' U2 l to solve 3/5 of it, then use a single commutator to fix the last two pieces together. What you did was to join the x-centres to 1×2 pair; this can be done in 8 moves, but you used 12 here. (Again, check Meep's L2C site. This is the same case as what you had for the last blue bar.)

(Bonus: there's an even shorter solution that doesn't make use of commutators. This is pretty advanced, though.)

Also, if you're right-handed, you should preferentially keep the solved bars on the left side, so that you'll do most of the moves with your right hand.

9. You spent four seconds thinking about what to do between finishing centres and edge pairing. This isn't a bad thing per se, and as you do more solves you can try to make the transition more fluid.
10. Lots and lots and lots of cube rotations during edge pairing. If you want to do E-slice edge pairing, you must use flicks to do Uw/3Uw moves instead of wrist turns; if you want to do M-slice edge pairing, wrist turns are fine (not like you have a choice), but you shouldn't ever be doing any rotations other than x and x'; if you want to do S-slice edge pairing, you shouldn't.
11. Assuming you want to do E-slice edge pairing, because that's what most people do these days, if you can't find the edge pieces you need on the top layer, just do a z2 rotation to switch the top and bottom layers. You can use R D R' and similar inserts if you happen to notice the pieces on the bottom, but you mostly shouldn't bother peeking at the bottom layer.
12. stop rotating so much during L4E aaaaaaaaaaaa

Learn your M-slice flipping algs if you want to do M-slice L4E. U' R' U R' F R F' is the popular one, I believe.
13. Your OLL parity alg execution has like eight regrips; standard algs require only three or four. Look up fingertrick videos for OLL parity algs.
14. For two-look PLL, use T perm (or J perm) and Y perm to permute the corners. A perms are really bad on big cubes.

Better answer than i ever could have hoped for, ill be back in a month with results
 

DanRman

Member
Joined
Jul 23, 2019
Messages
6
Note: t-centres are the "edge" centres, x-centres are the "corner" centres, and the obliques are the rest. I use SiGN notation here (3R = turn only the third layer from the right; 3r = 3Rw = turn three layers from the right), as I do everywhere else.

1. scramble your cube harder There are like a zillion blocks on the cube that haven't been broken up.
2. Unnecessary slicing-back moves on the first and third centres, as already pointed out.
3. Better solution for the second orange bar. (Probably easiest to see what's going on if you rotate the virtual cube to a back view.)
4. Efficiency is generally on point, except for the unnecessary slices thing. (If I wanted to nitpick I could point out a lot of tiny mistakes, but that would take way too much effort.)
5. At 1:47 you have 4 white bars solved on your left (the right in the video frame), and you have a 4/5 white bar (missing an oblique) on another face. You should prefer to AUF so that the missing oblique is next to the free slice (i.e. so that it's on the left); by doing this, you can use just a normal wide move to realign the 4 white bars you've already solved, rather than using a slice move (which you had to do because you had the missing oblique on the opposite side of the free slice).
6. At 1:53-ish, you can directly form the second blue bar with U' 3l' U2 3l (then insert with U 3r U2 3r'), instead of using M' U2 M to move the inner t-centre to the right spot. (Slice moves take two wrist turns to execute; they're not as slow as doing two separate wide moves, but they're definitely slower than doing only one wide move.)
7. (Last blue bar.) Learn cases 09 and 10 on Meep's L2C list. (Just because the list is for the last two centres doesn't mean that you can't also use those solutions on the other centres.) Also, try to avoid having holes in your bars as you form them; you need to use slice moves to insert pieces into the holes and slice moves are bad.
8. Last two centres. This is the only part I bothered to reconstruct fully, and some of what I say here will also apply to other parts of the centre solving.

Your solution (59 moves not counting a U U' and execution mistakes):
F' 3r U 3r' U' 3R U 3R' 3r U2 3r' // two bars (13 moves)
3-4r' F 3-4r l' U l r U r' // third bar (11 moves)
U' l' U l U l' U' l 3L' U2 3L // fourth bar (13 moves)
U' l' U' l U l' U' l U l' U2 l U' l' U' 3L' U l U' 3L // last bar (22 moves)
alg.cubing.net

Sample solution (roughly how I would've solved it; 37 moves):
F 3l' U 3l U' 3l' U2 3l // two bars (8 moves)
3r U' 3r' U r U l' U2 l r' // third bar (10 moves)
U 3r r' U2 r U2 r' U' 3R' // fourth bar (10 moves)
U2 3L F r' F' 3L' r // last bar (9 moves)
alg.cubing.net

Let's point out all the inefficiencies in your solution. The middle bar was already solved, so there isn't anything to say about that. You have all the pieces needed to form the second bar on the top face, ready to be joined in 3 moves (see the sample solution), but you used a different set of pieces to form a 4/5 bar with a hole (don't make partial bars with holes, as mentioned in point 7). While finishing the second bar, you could also have inserted it at the same time by doing 3R U' 3R'.

The third bar was very weird. I'm not sure if it's because you saw the 1×4 bar on the top face and wanted to preserve that. You generally shouldn't bother preserving blocks for the last two bars, since they show up by chance very frequently anyway. Instead of doing weird slice moves, you could have done U2 r U2 r' U r U2 r'.

Fourth bar was fine, but you could've inserted it with 3l' U2 3l instead of using slice moves 3L' U2 3L. Remember: wide moves are faster than slice moves. For the last bar, you could have used U2 l' U2 l to solve 3/5 of it, then use a single commutator to fix the last two pieces together. What you did was to join the x-centres to 1×2 pair; this can be done in 8 moves, but you used 12 here. (Again, check Meep's L2C site. This is the same case as what you had for the last blue bar.)

(Bonus: there's an even shorter solution that doesn't make use of commutators. This is pretty advanced, though.)

Also, if you're right-handed, you should preferentially keep the solved bars on the left side, so that you'll do most of the moves with your right hand.

9. You spent four seconds thinking about what to do between finishing centres and edge pairing. This isn't a bad thing per se, and as you do more solves you can try to make the transition more fluid.
10. Lots and lots and lots of cube rotations during edge pairing. If you want to do E-slice edge pairing, you must use flicks to do Uw/3Uw moves instead of wrist turns; if you want to do M-slice edge pairing, wrist turns are fine (not like you have a choice), but you shouldn't ever be doing any rotations other than x and x'; if you want to do S-slice edge pairing, you shouldn't.
11. Assuming you want to do E-slice edge pairing, because that's what most people do these days, if you can't find the edge pieces you need on the top layer, just do a z2 rotation to switch the top and bottom layers. You can use R D R' and similar inserts if you happen to notice the pieces on the bottom, but you mostly shouldn't bother peeking at the bottom layer.
12. stop rotating so much during L4E aaaaaaaaaaaa

Learn your M-slice flipping algs if you want to do M-slice L4E. U' R' U R' F R F' is the popular one, I believe.
13. Your OLL parity alg execution has like eight regrips; standard algs require only three or four. Look up fingertrick videos for OLL parity algs.
14. For two-look PLL, use T perm (or J perm) and Y perm to permute the corners. A perms are really bad on big cubes.
Also how do I do less rotations on last 4 edges?
 
Top