*1 week ago*

Hello - been a while since I last posted a blog! I thought it would be nice to post some reconstructions/thoughts on my 3x3 solves from last weekend's competition, Cube For Cambodia 2018. The averages from rounds 1 and 2 were decent, but round 3 was a bit below par.

Some things I noticed/have been working on.

- Since worlds finals last year, I've really tried to place more of an emphasis on starting solves safely, and not mis-inspecting or planning pairs incorrectly. Round 2 solve 3 here was probably the only solve I actually made an inspection error, but it didn't hurt the solve too much.
- I made a few serious F2L mistakes in these solves. Perhaps that was a product of concentrating a bit more on solving cross+first pair correctly and less focus on looking ahead. Some of these errors were self-inflicted by missing obvious pairs or panicking a little bit when my lookahead failed me.
- Probably too many lockups for my liking as well. This is reflected in average TPS being below 9, and only 1 solve with >10 TPS. Lockups in competition are essentially just caused by shaking hands which lead to inaccurate turning. I'm generally quite conservative with my fingertricks and algorithm execution in comp, so I just need to keep the nerves under control a little bit better. Before each solve, I spam TPS/algs on an old unstable cube - once I'm not locking up on that cube, that's a sign that my hands are probably ready to go.

Thanks to Brest for doing the reconstructions for the Round 1 solves! :)

**Round 1: 6.26 average**

**Scramble 1: ****B D L F U2 F2 D B U F2 D' B' U2 F B L2 D2 B2**

The blue cross on this scramble is very straightforward, and the yellow & red are pretty easy too. I think I took a while to decide which one I'd solve in my inspection. It was fairly simple to see the orange/white pair would form during the cross, but to solve it afterwards would require a rotation.

Inspection: x y'

Cross: F R U L2'

1^{st} pair: U y' U R U' R'

The orange/yellow pair would have required another rotation, so instead of that I did the rather inefficient white/red pair. Both options were probably just as good.

2^{nd} pair: R' U R U L' U L U' L' U L

At this point, all edges are oriented on the cube after a y rotation, which is very handy.

3^{rd} pair: y R' U R U' R U R'

4^{th }pair: U R' U' R

Technically, I know how to recognise this ZBLL (or EPLL skip), and am working on the pi 2GLL set, so skipping EP here is only a little bit lucky :p

COLL: U2' R U2' R2' U' R2 U' R2' U2' R U

**4.92 seconds, 40 moves (STM), 8.13 TPS**

**Scramble 2: ****R' D2 L' D2 F2 R' D2 U2 B2 F D' L2 R2 D2 B U R2 B' R**

I think I missed the easy blue cross on this solve, and so ended up doing a 9 move cross. Not great.

Inspection: y x

Cross: D R' D' R U D2 R' F R

1^{st} pair: y L' U L R U' R'

I solved pair 2 and effectively did a y2 rotation to place the solved pairs in the back.

2^{nd} pair: y' d R' U' R

3^{rd }pair: D' L' U L (U D)

4^{th} pair + VLS: R U R' U U R U' R' U R U' R' U' R' F R F'

Slight mistake in PLL recognition (thought it was a U perm). This solve had a few lockups/pauses as well.

PLL: U R' U2' R' D' R U' R' D R U R U' R' U' R U

**6.97 seconds, 60 moves (STM), 8.61 TPS**

**Scramble 3: ****R2 L' F2 D2 F' D L2 B' D L U B2 U B2 D2 L2 D' F2 D**

This was a nice one. Keyhole extended cross was straightforward, the only issue might have been the rotation. I was also able to see that the green/red pair would end up as a simple insertion case, so made sure to do a u' at the end of the X-Cross to avoid a rotation.

X Cross: F' R' D' R y R U' R' u'

2^{nd} pair: U' R U R'

Here, the green/orange pair can be solved without a rotation, yet I rotated to solve the blue/orange one. This was most likely because I saw the blue/orange F2L pieces during the first two pairs, and perhaps because my thumb would have been blocking the green/orange corner piece.

3^{rd} pair: y' L' U2 L U' L' U L

4^{th} pair: d U R' U' R U R' U' R

OLL: U' R U2' R' R' F R F' R U2' R'

Brest said the OLL lockup cost about 0.26 seconds lol.

**4.22 seconds, 38 moves (STM), 9.00 TPS**

** **

**Scramble 4: ****R U L U D L B U' F' R' U F' B D2 R2 D2 R2 F2 B**

Planned cross and first pair in inspection for this one, but it wasn't a particularly easy start.

Inspection: z

Cross: r' (U' D') R' F R2 u'

1^{st} pair: R U2' R' U2 L' U' L

This is where it all went downhill, I mis-inserted the second pair and then used a couple of rotations to fix it up :(

2^{nd} pair: R' U' R y U R U' R' y' U' L U L'

3^{rd} pair: y U' L' U L U L' U' L

Nice OLS case saved the solve from being a complete disaster, thankfully!

4^{th} pair + OLS: U' R' F R F' R U R'

PLL: R' U' F' R U R' U' R' F R2 U' R' U' R U R' U R U

**7.27 seconds, 64 moves (STM), 8.80 TPS**

** **

**Scramble 5: ****R2 B2 R2 F' U L2 F' L' U2 B' R' D' B2 R2 U' L2 D L2 B2 U2**

I spent like 5 seconds trying to find a good cross solution, to no avail.

Inspection: x2 y

Cross: D' L' R' F R D' R' D2

First two pairs were nice and easy, and left me with the remaining two F2L edges oriented. Overall, it was the lockups in F2L and OLL which slowed this solve down.

1^{st} pair: R U R2' U' R

2^{nd} pair: y' L' U L U' L' U' L

3^{rd} pair: R U R' U2 R U' R'

4^{th} pair: U R' U R U2' R' U R

LL was very standard as well.

OLL: U2' R U B' U' R' U R B R'

PLL: R U R' F' R U R' U' R' F R2 U' R' U'

**6.89 seconds, 63 moves (STM), 9.14 TPS**

** **

**Round 2: 6.32 average**

**Scramble 1: ****U2 L U2 R' D2 R' B2 F2 U2 L' R' U R' B D2 F' D B2 L' D**

Yeah, there's hardly anything on this scramble, I went with the only cross with a piece solved, which was red, and tried to plan a decent solution and first pair. The pair I saw in inspection was the blue/yellow pair, as the edge didn't move during the cross solution, and the corner went to the top layer. Unfortunately the first pair required a rotation, but I was able to solve it into the back.

Inspection: y2 z'

Cross: F' R' F D F D2

1^{st} pair: U y' R' F U' F' R

The blue/white pair would have been a better choice for pair 2, but I saw the green/yellow one first, and it wasn't too bad either.

2^{nd} pair: U' R U' R' U y' L' U L

I have no idea why I rotated in the y' direction for the last two pairs, but somehow I managed to execute them quite well with my left hand. The rotation directions were definitely a mistake though. The only benefit I can think of is that (y' U) is probably slightly faster to execute in one motion than (y U).

3^{rd} pair: y' U L U L' U' L U L'

4^{th }pair: y' U L U2 L' U' L U L'

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

PLL: U y x R2' D2 R U R' D2 R U' R

**6.54 seconds, 57 moves (STM), 8.72 TPS**

**Scramble 2: ****D L F2 U' R U2 B' D2 F' L' D R2 L2 D2 R F2 L F2 R2 D2**

Again, only one cross colour with solved edges, so I was immediately drawn to orange. However in hindsight, white would probably have been a better choice. After the first two moves of the cross (R D), it's better to do F L2 than L2 F to finish off the cross, because it moves the green/yellow edge into the top layer and sets up an easy rotationless first pair.

Inspection: z'

Cross: R D x U L2' x' (D2 U')

1^{st} pair: R U R' U R U' R'

Always a bit risky to solve two diagonal pairs to start F2L, but the blue/white pair was too easy to pass up.

2^{nd} pair: L' U L U' L U L'

After this, I used keyhole to solve pair 3, and whilst restoring the D layer, did a U' to begin solving the final pair.

3^{rd }pair: D2 L' U L (D2 U')

4^{th} pair: y R U' R' U' R U' R' U R U' R'

I used a different OLL alg to avoid the AUF move. I also think it flows a bit more nicely than the standard alg.

OLL: R' U' R' F R F' R U' R' U2 R

PLL: U2' R2 U R' U R' U' R U' R2 (D U') R' U R (D' U2')

**6.46 seconds, 65 moves (STM), 10.06 TPS**

**Scramble 3: ****U2 B' D2 L2 D2 B2 U2 B' D2 R' B D U B R B2 D F' U'**

After the first two scrambles, I was kind of spoiled for choice here, and chose yellow instead of red. However, the yellow cross solution was 2 gen and I was able to see a rotationless first pair pretty easily, so I stuck with it.

Inspection: y

Cross: R D2 R D' R'

1^{st} pair: U R U' R' L' U' L

Big mistake here not seeing or tracking the blue/orange pair. I believe I mis-inspected, and thought the green/orange corner would end up in its slot, so ended up doing a bunch of moves which didn't solve an F2L pair.

Do Nothing: y U' R' U' R U R' U' R

2^{nd} pair: y' R U' R2' U' R

After this, I realised I'd made a mistake and would need to fix it.

3^{rd} pair: D2 U' R U R' D2

4^{th} pair + Summer Variation: R U R' U2' R U2' R' U R U' R' U R U2' R'

PLL: y' x' R2 D2 R' U' R D2 R' U R'

**5.97 seconds, 55 moves (STM), 9.21 TPS**

** **

**Scramble 4: ****R2 F2 L U2 F2 L F2 U2 R D L2 F D B' F2 L' U F' R2 B**

This solve was a bit of a disaster. Locked up a bunch in cross and F2L and then made a pair mistake as well, which basically killed the solve.

Inspection: y

Cross: R2' D' r2 U' r' u'

1^{st} pair: U y' R U2' R' U R U' R'

2^{nd} pair: y U2' R U R' U R U' R'

After the first two pairs, I did a y rotation for some reason, perhaps because I wasn't looking ahead very well and didn't notice that the blue/white pair could have been solved into the back-left slot without a rotation. Also, because the two solved pairs are occupying the front two slots, I know I'm likely going to have to rotate in the y direction to finish F2L anyway, but I started to do it one pair too early, and then realised as soon as I started.

For some reason, I also forgot how to solve F2L pairs here.

Do Nothing: y y' U' U2 L U2 L' U L U' L'

3^{rd} pair: y U R U2' R'

(The reason for doing U R U2' R' instead of R U' R' is probably an attempt to get a better OLL case... it was way too late for that.)

4^{th} pair: y R U' R' U R U R' U R U' R'

OLL: U2' R' U' R U' R' U2 R F R U R' U' F'

PLL: U R' U2' R U2' R' F R U R' U' R' F' R2 U'

**10.39 seconds, 75 moves (STM), 7.24 TPS**

** **

**Scramble 5: ****R' F' U F R2 F L B2 R2 D R' F2 D2 R' U2 R U2 R2 U2 L2**

Orange cross is by far the best option - the only tricky part is figuring out which pairs to solve first.

- The blue/white pair is definitely not a good choice.
- Blue/yellow is together, but would require a rotation after the cross solution
- Green/yellow is decent but will also require a rotation.
- Green/white will be easy, and will transform the green/yellow into a nicer solution as well.

I ended up ignoring the blue/yellow in inspection, and planned to solve the green/white, followed by the green/yellow

Inspection: y'

Cross: r' R2 (D2 U')

1^{st} pair: R' U2 R U' R U R'

2^{nd} pair: y' U R U' R' U' L U L'

Whilst executing the second pair, it was fairly easy to track the blue/yellow pair and insert it quickly right afterwards.

3^{rd} pair: y' U' R U R'

At this point, all the corners are solved. Sadly, I'm not quite pro enough to do two 3-edge cycles in a 3x3 speedsolve.

4^{th} pair: y U R U' R' d' L' U L

OLL: U' R U R' U' R' F R F'

PLL: U' R U' R U R U R U' R' U' R2 U2'

**5.43 seconds, 53 moves (STM), 9.76 TPS**

**Final Round: 6.98 average**

**Scramble 1: ****D B2 D R2 F2 U2 F' D B U2 F2 R2 D' L2 B' F2 L U2**

I knew the start of this solution was going to be quite awkward to execute, so I tried to do it quite safely to ensure I didn't make any mistakes. Starting with an X-Cross with the pair solved in the back is quite nice.

Inspection: x2 y'

Cross: U' F (U D') F R' U' L U L' D

Before inserting that green/red edge to make the extended cross, I saw the blue/orange pair and was able to track it very easily. However, I went straight into solving that pair, and missed the solved green/orange one.

2^{nd} pair: U R U R' U2 R U' R'

This was doubly bad as it left a really awkward case for pair 3, which I solved using two rotations.

3^{rd} pair: U y' R U' R' U2 y' R' U' R

4^{th }pair: L' U L U L' U' L

OLL: U R U R' U R U' R' U' R' F R F'

A very small detail here is that I did the AUF before PLL in the anti-clockwise direction so that I could simultaneously do the regrip that was required for the V perm.

PLL: U2' R' U R' U' y R' F' R2 U' R' U R' F R F

**6.64 seconds, 62 moves (STM), 9.34 TPS**

**Scramble 2: ****F2 L' U2 L' B2 U2 L' F2 L' B' L B' U' L2 D R2 B' L' F R'**

It probably would have been a nice idea to use the 1x1x3 block in a better fashion, but I didn't see anything particularly easy on white or orange, so I decided just to use it as an F2L pair.

Inspection: x'

Partial cross: U' R2' F D2

1^{st} pair: U L' U' L

Finish cross: U' r U' r'

There was a slight regrip mistake before pair 2.

2^{nd} pair: U' R U2' R' U2 R U' R'

At this point, it gets a little bit tricky, as I don't see my first pair. This is a result of solving my first two pairs into the front slots on the cube.

3^{rd }pair: y R U2' R' y R U R'

4^{th} pair: y' U R U' R'

Whilst the 3rd and 4th pairs are reasonably move-efficient, the rotations were quite costly and hurt the flow of the solve.

OLL: U2 R U B' U' R' U R B R'

PLL: U' R2 U R' U R' U' R U' R2 (D U') R' U R (D' U) (Missed the last U move here on the first try, lol)

**6.97 seconds, 57 moves (STM), 8.18 TPS**

**Scramble 3: ****L2 B D2 L2 B R2 F' U2 F' D2 L D' U2 L' U2 R2 U R D2 L2**

A few good options on this scramble, I ended up going with green and inserting the first pair before the final cross edge.

Inspection: y2 x

X-Cross: L F' L' U' L' U L U' L' R2

I was tracking the orange/yellow pair during the extended cross, so don't really mind too much that I missed the easier red/yellow pair after the R2.

2^{nd} pair: L' U L U R U R'

Huge mistakes here, panicked a little bit after the fast start and did a bunch of rotations when they were unnecessary.***

3^{rd} pair: y U R R' U R U2' R' U y' R' U' R

For the 4th pair, I wanted to avoid a dot case, so did an alternative solution.

4^{th} pair: y' R U' R' U' R U R' U' R U R'

Definitely should have done OLLCP here...

OLL: U r U' r2' U r2 U r2' U' r

PLL: U x' R U' R' D R U R' D' R U R' D R U' R' D'

**7.21 seconds, 66 moves (STM), 9.15 TPS**

***A much better (multislotting) solution for the last two pairs would have been U y' R U2' R' U y' R' U' R, U L' U L, which leaves a nice ZBLL finish.

**Scramble 4: ****R' U2 R D2 L2 B2 R' U2 F2 R D' R2 F2 L2 B U2 F L D F'**

The most obvious cross is on orange, but the green/yellow pair was a little bit hard to deal with. Luckily, I was able to inspect enough to see that the first two pairs would actually become a really nice multislotting case.

Inspection: z'

Cross: D R D' R' U' L' (U2' D2')

1^{st} and 2^{nd} pair: R U' R2' U' R2 U' R'

3^{rd} pair: y' U' R U' R' U' F U' F'

The solve is going very well, and the last pair is already set up. This is where everything goes downhill - I mess up the last pair insertion really badly, which messes with my OLL recognition, and so I then recognised the OLL case incorrectly. With half decent execution for the rest, it would have been a pretty straightforward sub 7, but it ends up being an 8.

4^{th} pair: U' R' F R F'

OLL attempt 1: U' r U' r' r U r'

OLL: R U2 R' U' R U' R' F R U R' U' F'

PLL: U' R2' F2 R U2 R U2' R' F R U R' U' R' F R2

**8.03 seconds, 64 moves (STM), 7.97 TPS**

** **

**Scramble 5: ****U2 R2 U2 R' F2 L F2 D2 R2 D' F L' R' D' B' R2 D B F U'**

In inspection I was able to see that the white/red and the orange/yellow pairs would end up in the top layer, and so those would be the pairs I'd solve first. Important here is to solve them with only 1 rotation.

Cross + 1^{st} pair: r' R D R' D2 R2 U' R' U y' R' U' R

2^{nd} pair: U L' U' L U' L' U L

The next pair is decent, but at this point my lookahead to the last pair wasn't going very well - I think I semi-guessed that the yellow sticker in the BDL position meant that the back-left corner was solved and I would need to just flip the BL edge, which was an incorrect guess - which lead to the rotation error.

3^{rd} pair: R U R' U2' R U R'

4^{th} pair + Winter Variation: y y' U L U L' y (U' U) R' U R U2' R' U R U2' R' U' R

Luckily I got an A perm - I like these :)

PLL: x R2' D2 R U R' D2 R U' R B'

**6.76 seconds, 54 moves (STM), 7.99 TPS**

Become a free member to post a comment about this blog.

## Showdown baxa

Posted 1 week agoCongratulations for the world record!

## Sofia Nguyen

Posted 1 week agoYay! Waiting for this for a while. Always love to see you critique your own solves :)

## Sam Durocher

Posted 1 week agoSecond on the blog! Congrats on the world record. You freakin' shattered it!

Question: Is the Angstrom Air SM your main now?

## William Yang

Posted 1 week agoAwww would have been sub 4 if you didn't get the lock up would have been even harder to beat then:P Also do you think it is possible to get sub 3? I believe you are the closest to it with 3.01 don't know if anyone has anything quicker.

## Max Chan

Posted 1 week agoI was blow away by the 2 sub 5s in your first average.

## Scented Wipes

Posted 1 week agoGOOD JOB FELIKS!

## Eliezer Cho

Posted 1 week agonice

## William Yang

Posted 1 week agoThe anticipation for the extra video is unreal

## Cuberious The SpeedCuber

Posted 1 week agoYES ANOTHER BLOG!!! I CAN FINALLY COMMENT!!!

Feliks you are known to have possibly the best lookahead in the world. You mentioned that most of your practice (95%) is subconscious/timed solves.

How did you build up your lookahead?Did you do a bunch of speed solves and developed a superhuman level of tracking multiple pieces ("tracking") or do deliberate solves (walkthrough solves by yourself) and figured out "if I solve this pair, then this other pair ends up right here" ("knowing"). In other words, it seems like doing deliberate solves can skip the "tracking" stage because you'llknowwhere the pieces will go. I know there's a blog page, but I want to study these theories further.Did you watch lookahead tutorials from Joseph Skyler (jskyler91)?

## Tianyuan Li

Posted 1 week agoHi, Faz.Firstly, congratulation to the world record for only 4.22 seconds.I am a cuber from China who is your faithful follower.I always wondered how to be a color neutral cuber .It is the color neutral to help you make the world record.So I am eager to know how to practice to be a color neutral cuber.

## Abhinav Akash

Posted 1 week agoHow to be color netrul

## Abhinav Akash

Posted 1 week ago"neutral"

## Abhinav Akash

Posted 1 week agoFeliks,congratulation for your new WR btw

## Abhinav Akash

Posted 1 week ago"Question" - Are you doing tutorials for bilndsolving ?

## Ethan Varn

Posted 1 week agoDo Nothing. Lol

## William Yang

Posted 1 week agoFeliks what is your average TPS at home? Is it any quicker than you competition TPS?

## Cuberious The SpeedCuber

Posted 1 week agoAccording to Brest's analysis of his 6.44 average of 100, his average TPS is above 9

## William Yang

Posted 1 week agoThnx

## Simon Cummings

Posted 1 week agoThat 4.22 solve was really nice, gg as always. Rip on the lockup tho lol

## Feliks Zemdegs from CubeSkills

Posted 1 week agoWilliam - Probably around 9 or so, as Cuberious has mentioned. It's probably very slightly quicker than competition TPS, probably due to using a stackmat.

## Ryan Ng

Posted 1 week agoHey there, Feliks! How much of ZBLL and OLS do you know? And btw, what do you do to inspect where the F2L pairs go?

## Cuberious The SpeedCuber

Posted 1 week agoFeliks, in example solves, will you actually use those F2L fingertricks in a real speed Solve?

## Idris Sentjurc

Posted 1 week agoProbably the most inconsistent average from Feliks. With a little higher tps, you could probably already break the WR on the first solve.

## Jesse Steele

Posted 1 week agojust got really into cubeskills, thanks Feliks for the inspiration and help for kids like me learning how to get faster. this whole website is helping me a lot. You're the best Feliks.

## Feliks Zemdegs from CubeSkills

Posted 1 week agoRyan - Probably between 100 and 150 ZBLL algorithms. For inspecting F2L pairs, I generally look for pieces which won't be heavily affected by the moves used to solve the cross, or which will remain/end up in the top layer.

Cuberious - Not necessarily the exact fingertricks, example solves are more solution-focused.

Idris - perhaps, although it's always hard to get super high TPS on solves with PLL skips, because PLL is generally the highest TPS part of a solve, and bumps up the average TPS.

Jesse - Cheers for the kind words! :)

## Naga Advaith Kommuri

Posted 1 week agoGood job for the 4.22 WR, the execution was easy but there was a lot of turning.

## Abhinav Akash

Posted 6 days agoWell done Feliks,now get back the 4x4 single and average

## Ryan Ng

Posted 6 days agoOK, thanks Feliks! Will you make a tutorial on ZBLL? I know COLL is technically part of it but I want to know about some ZBLL algs you recommend.

## William Yang

Posted 6 days agoFeliks what things help you plan out your cross and first pair so quickly? Is there something you do to help you recognize a good pair?

## Cuberious The SpeedCuber

Posted 6 days agoWhy did you stop Friday Fours?

## William Yang

Posted 6 days agoJust got a 100/100 on the PLL recognition trainer on Color Neutral. Biggest achievement of my life but doesn't help me too much when I have 4 look last layer

## William Yang

Posted 6 days agoAlso Feliks what happened to the extras video did you post it?

## Ethan Jan

Posted 6 days agoIf you had an A perm with a U pre-AUF and then U' post-AUF, would is be better to do Y A perm or U A perm U'?

## Jacob Shum

Posted 5 days ago@Feliks How do I become sub-12? I'm currently averaging mid-13's with my best ao5 as a 12.77. I know all of COLL, and some WV, SV, MW, BLE, VLS, OLLCP, ZBLL, ELL, and CLS. Really long list of algs xD.

I average about 1.5 for Cross, 7.5 for F2L, 1.5 for OLL, and 1.5 max for PLL. My main downsides are the transitions between Cross-F2L, and OLL-PLL. F2L-OLL isn't really a problem for me.

## William Yang

Posted 5 days agoJacob Shum Did you happen to check out Feliks' F2L algs not the useful ones but the different slot position ones that decreased my time down to sub 12 my cross is 1-2s F2L is 6-7s LL is 2-3s I get mostly Sub 12 but sometimes get sub 10. Your F2L always have room for improvement as it takes up the bulk of your time. Feliks has >1s cross 3-5s F2L and >2s LL

## Mealo Guo

Posted 4 days agoCOOL!## Cuberious The SpeedCuber

Posted 4 days agoWilliam, It's on the YouTube Channel.

## William Yang

Posted 4 days agoOh ok thank Cuberious I didn't bother to look because usually puts it on recently posted tab and I didn't see it there so I assumed that he didn't post it yet

## Kieren Marsh

Posted 2 days ago"so ended up doing a bunch of moves which didn't solve an F2L pair.Do Nothing: y U' R' U' R U R' U' R"lmao!!!

The "Do Nothing" is my most proficient subset.

## William Yang

Posted 1 day agoIf you were panicking Feliks you are really good at hiding it I couldn't tell that you were in a rush even if I knew.