The Solve Critique Thread

[JianhanC]

You really must turn faster. But if your cube is not allowing you to you should take advantage of that to lookahead, so that you can optimise your moves. I see a lot of unnecessary moves. Also you need to control your rotations, so that you don't lose track of what you're doing. Nothing much to add about edges, lookahead faster than you turn. 3x3 will improve by practising 3x3, and when you get comfy with turning on bigcubes. And lastly, you could have used R' U R U' to setup the diagonal edge to the opposite, then doing the PP alg. Much less painful.

 

[windhero]

This has been a bad day at cubing for me. I've had an Ao100 of 20.50 seconds and an Ao12 of 18.54 seconds. I'm trying to get constant sub 20.
EDIT: I just tried look ahead with a metronome. My limit for lookahead seems to be around 160bpm. Any faster than that and I start to fail if there are hard cases where no corners are on the top layer.

TIMES:
24.84, 20.77, 21.56, 19.42, 21.20
number of times: 5/5
best time: 19.42
worst time: 24.84

current avg5: 21.18 (σ = 0.40)
best avg5: 21.18 (σ = 0.40)

session avg: 21.18 (σ = 0.40)
session mean: 21.56

 

[sneaklyfox]

Spoiler

This has been a bad day at cubing for me. I've had an Ao100 of 20.50 seconds and an Ao12 of 18.54 seconds. I'm trying to get constant sub 20.
EDIT: I just tried look ahead with a metronome. My limit for lookahead seems to be around 160bpm. Any faster than that and I start to fail if there are hard cases where no corners are on the top layer.

TIMES:
24.84, 20.77, 21.56, 19.42, 21.20
number of times: 5/5
best time: 19.42
worst time: 24.84

current avg5: 21.18 (σ = 0.40)
best avg5: 21.18 (σ = 0.40)

session avg: 21.18 (σ = 0.40)
session mean: 21.56

Nice cube. I like your turn style. A little hard to tell from this angle but maybe you need better cross and better transition to the first F2L case. Or maybe sometimes better transition to the next F2L... there are a couple of huge pauses. Not bad though.

 

[MarcelP]

This is a very good average for me. What are the quickwins for me to get better
Average of 5: (23.31) 23.62 24.60 23.77 (29.79) = 24.00
[video=youtube_share;KM2yFL3A8lw]http://youtu.be/KM2yFL3A8lw[/video]

 

[uniacto]

This is a very good average for me. What are the quickwins for me to get better
Average of 5: (23.31) 23.62 24.60 23.77 (29.79) = 24.00
~video~

Right now it's not bad, but it seems that you've still got a bit of pauses between steps and F2L pairs.

for the first solve, I'd recommend you work on that Z perm. You could have gotten a 20-21 second solve if it was a bit faster. Here's a nice Z perm that I recommend: M2' U M2' U M' U2 M2' U2 M'
second solve: I saw a U4, which is probably unneeded.
third solve: during f2l, I saw a U and U' back to back. Maybe work on tps for LL?
fourth solve: Mainly lookahead stuff that need a bit of improvement
last solve: Look ahead haha

but I'm no expert, there's many other ways people would advise you to solve. Those are just my recommendations, I doubt that they're very good ones at that. At least I tried? haha.

 

[Milky3bar]

avg12:18.12
18.64, 16.94, 16.53, 19.80, 17.63, 17.19, 17.14, 18.21, DNF(14.36), 17.79, 19.39, 18.46

Pretty average for me now. Switched to CN roux on 18th April(a month plus) when I was average around 17. Any advice would be appreciated! Thanks

 

[Yuxuibbs]

[video]
avg12:18.12
18.64, 16.94, 16.53, 19.80, 17.63, 17.19, 17.14, 18.21, DNF(14.36), 17.79, 19.39, 18.46

Pretty average for me now. Switched to CN roux on 18th April(a month plus) when I was average around 17. Any advice would be appreciated! Thanks

I don't do roux but it looks like first 2 blocks need work, it's not as smooth as it can be and there are pauses. I think LSE and other stuff is good but you might want other opinions.

 

[MarcelP]

Right now it's not bad, but it seems that you've still got a bit of pauses between steps and F2L pairs.

for the first solve, I'd recommend you work on that Z perm. You could have gotten a 20-21 second solve if it was a bit faster. Here's a nice Z perm that I recommend: M2' U M2' U M' U2 M2' U2 M'
second solve: I saw a U4, which is probably unneeded.
third solve: during f2l, I saw a U and U' back to back. Maybe work on tps for LL?
fourth solve: Mainly lookahead stuff that need a bit of improvement
last solve: Look ahead haha

but I'm no expert, there's many other ways people would advise you to solve. Those are just my recommendations, I doubt that they're very good ones at that. At least I tried? haha.

Thanks! I will look at that Z-perm I really don't like Z-perm since it is easily mixed up with the H perm I can get rid of the pauses when I slow down a bit. Unfortunately I have problems getting sub 30 with slow turning. I need to practice that way more I guess.

 

[Milky3bar]

I don't do roux but it looks like first 2 blocks need work, it's not as smooth as it can be and there are pauses. I think LSE and other stuff is good but you might want other opinions.

Thanks! Yeahh after i switched to CN My first block to second block transition became really bad I guess ill do slow solves to improve my lookahead!

 

[Yuxuibbs]

I know for some of the solves, I lost the flow of the solve and couldn't get back on track with look ahead so I completely failed the solve but other than basically forgetting to look ahead on certain solves, what should I work on?

Bump (6 days)

 

[mDiPalma]

Bump (6 days)

1. You only NEED to rotate once for CFOP F2L. You rotate 3-4 times per F2L (some of which are y2).
2. Use the left hand.
3. Look for OLL during last F2L pair.
4. Try yellow cross if it looks easy.

 

[sneaklyfox]

1. You only NEED to rotate once for CFOP F2L. You rotate 3-4 times per F2L (some of which are y2).

Actually, you don't NEED to rotate at all for CFOP F2L. You just need to differentiate between rotating needlessly and rotating advantageously.

 

[Bsmith]

so i'm kind of new (again, i actually used to solve when i was a kid in the early eighties) - picked it back up a few months ago, and gravitated toward roux. so i'm averaging 35-42 seconds usually this last several days, and have dipped below 25 here and there. Anyway, here's a decent solve for me which also I can see where I'm hesitating or not seeing and then being inefficient with easy things like that part with the red/green red/greew/white. anyway, besides obvious things like that wanted to put a vid here to get a general critique and advice on things to work on. this forum rocks, thansk for all the wealth of information!

 

[jayefbe]

Actually, you don't NEED to rotate at all for CFOP F2L. You just need to differentiate between rotating needlessly and rotating advantageously.

This, so much agreement.

 

[uniacto]

Thanks! I will look at that Z-perm I really don't like Z-perm since it is easily mixed up with the H perm I can get rid of the pauses when I slow down a bit. Unfortunately I have problems getting sub 30 with slow turning. I need to practice that way more I guess.

Oh yea, that's true. I guess it might be hard for some people to keep them straight.

practice makes perfect!

 

[JianhanC]

so i'm kind of new (again, i actually used to solve when i was a kid in the early eighties) - picked it back up a few months ago, and gravitated toward roux. so i'm averaging 35-42 seconds usually this last several days, and have dipped below 25 here and there. Anyway, here's a decent solve for me which also I can see where I'm hesitating or not seeing and then being inefficient with easy things like that part with the red/green red/greew/white. anyway, besides obvious things like that wanted to put a vid here to get a general critique and advice on things to work on. this forum rocks, thansk for all the wealth of information!

Your movecount is alright, but you need to react faster, like the last red/green pair. Always lookahead faster than you can turn, and when you're comfortable, increase turning speed and repeat. Also, start developing fingertricks. The T-perm can be done very fluidly if you use your left and right index fingers (experiment yourself; there might be a different way to do it that you like).

 

[KongShou]

Sub 15 ao12(14.16)

Any suggestions?

 

[mDiPalma]

Actually, you don't NEED to rotate at all for CFOP F2L.

Assuming 2-gen pair inserts are used, stock CFOP F2L requires .9375 (1-.5^4) rotations per solve. Correct me if I'm wrong.

You could use fancy pair insertions to correct the orientation of other F2L edges, but this is not easy, nor is it stock CFOP F2L.

You just need to differentiate between rotating needlessly and rotating advantageously.

Coming from a fixed-center solving method, I'll have you know that "rotating advantageously" is oxymoronic.

 

[jayefbe]

Assuming 2-gen pair inserts are used, stock CFOP F2L requires .9375 (1-.5^4) rotations per solve. Correct me if I'm wrong.

You could use fancy pair insertions to correct the orientation of other F2L edges, but this is not easy, nor is it stock CFOP F2L.



Coming from a fixed-center solving method, I'll have you know that "rotating advantageously" is oxymoronic.

You contradict yourself in your own post. Rotating can be advantageous because 2-gen pair inserts can be slow or fast depending on the angle.

I think the point is, F2L can almost always be done with only a single rotation, but that requires knowing the orientation of all unpaired F2L edges in order to determine the most efficient order possible in terms of rotations. Is that what you're saying everyone should be doing? Some people might be able to do that, but I doubt that's feasible for most. I don't see a high percentage of solves with one or zero rotations during F2L, even among very fast cubers.

 

[KongShou]

Why am i not sub1? What do i need to work on?