[Help Thread] F2L Discussion and Help

[aboeglin]

NEVER LOOK AT THE TIMER DURING A SOLVE!!!

Thanks for the tip, I have to admit I do it quite often myself. I'm sure I could gain 2sec on average. Because looking at the timer isn't making you slower at the moment you look, but you don't keep track of the pieces that you'll want to place next while looking at the timer, so after looking at the timer I often find myself stuck in front of my cube for a good second.

This being said, I also like to know how long my F2L takes, so that I can have an idea where stands my F2L and where stands my LL. F2L being the most precious progress to me. Knowing that LL will easily go down with enough practice.

 

[mark49152]

Many timers have an option to just show "running" or something instead of the time.

 

[carbon131]

i now got to sub 28 avrages i did 4 sessions in a row and all of the no higher that 28 seconds

thanks alot pipkicksass and ragecuber

 

[pipkiksass]

i now got to sub 28 avrages i did 4 sessions in a row and all of the no higher that 28 seconds

thanks alot pipkicksass and ragecuber

Good to hear. I need to listen to my own advice, and concentrate on fluidity. I've been stuck at 19-20 seconds for a couple of months. The way forward is smoother F2L with inserts in the back/left and better/more consistent lookahead. I believe that's true for all cubers between sup-60 and sub-6!!

 

[TheNextFeliks]

I average like 16.50. My cross takes about 2.80 and f2l is about 2.10 per pair or about 8.50 total. Any general tips to improve f2l? I can post a video this weekend if necessary.

 

[TDM]

I average like 16.50. My cross takes about 2.80 and f2l is about 2.10 per pair or about 8.50 total. Any general tips to improve f2l? I can post a video this weekend if necessary.

Your cross needs improving. Almost three seconds for a cross at that speed is nearly as bad as me From the splits in this thread, you should be aiming for about two. I don't know what the average movecount for a cross is, but I think it's about 6 (5.81 iirc and 4.81 for CN, but I'm too lazy to check), which means your TPS during cross is only around 2 - if you can easily plan the cross in inspection (and if you can't, try BLD cross), try to look for other solutions with better fingertricks, even if they're a few moves longer. Stuff like R u R' [u'] to insert the edge on RU to DR. Also, remember that the cross doesn't have to be done by putting every edge in place first - you can solve them relative to each other (e.g. a D2 away from being solved), then solve them after. Or you could do some like this, then solve them, then continue with the others. If efficiency during cross is the problem, you could always look at something like JARCS, which can do optimal cross solving, and you can find tricks from there that could be helpful.
It's difficult to say which parts of the F2L after that need improvement without a video. One would be very helpful.

 

[kcl]

I average like 16.50. My cross takes about 2.80 and f2l is about 2.10 per pair or about 8.50 total. Any general tips to improve f2l? I can post a video this weekend if necessary.

Do you mean your cross+f2l is 8.50? If so, your last layer sucks harder than mine. Focus on that XD

If not, definitely get that cross sub 2, and try to make the entire cross and F2l sub 10. Then just try and get your ll to about 4ish..

 

[TheNextFeliks]

Your cross needs improving. Almost three seconds for a cross at that speed is nearly as bad as me From the splits in this thread, you should be aiming for about two. I don't know what the average movecount for a cross is, but I think it's about 6 (5.81 iirc and 4.81 for CN, but I'm too lazy to check), which means your TPS during cross is only around 2 - if you can easily plan the cross in inspection (and if you can't, try BLD cross), try to look for other solutions with better fingertricks, even if they're a few moves longer. Stuff like R u R' [u'] to insert the edge on RU to DR. Also, remember that the cross doesn't have to be done by putting every edge in place first - you can solve them relative to each other (e.g. a D2 away from being solved), then solve them after. Or you could do some like this, then solve them, then continue with the others. If efficiency during cross is the problem, you could always look at something like JARCS, which can do optimal cross solving, and you can find tricks from there that could be helpful.
It's difficult to say which parts of the F2L after that need improvement without a video. One would be very helpful.

Ok. Yeah my cross is bad. Most times I can plan the cross. I need to practice it.

Do you mean your cross+f2l is 8.50? If so, your last layer sucks harder than mine. Focus on that XD

If not, definitely get that cross sub 2, and try to make the entire cross and F2l sub 10. Then just try and get your ll to about 4ish..

Ok. I meant the second way f2l without cross is 8.50. My LL is around 4.50-5.00

 

[kcl]

Ok. Yeah my cross is bad. Most times I can plan the cross. I need to practice it.



Ok. I meant the second way f2l without cross is 8.50. My LL is around 4.50-5.00

I would focus on a few things.
1. Get cross sub 2.
2. Make f2l sub 10
3. Maybe try for sub 4 LL. You can do this by learning better finger tricks for LL algs and just drilling them. I can sub 1.2 all but like 3 PLLs, and all my OLLs are sub 1.5. Including my reaction, my LL is usually high 3/low 4. Hopefully this helps!

 

[TheNextFeliks]

I would focus on a few things.
1. Get cross sub 2.
2. Make f2l sub 10
3. Maybe try for sub 4 LL. You can do this by learning better finger tricks for LL algs and just drilling them. I can sub 1.2 all but like 3 PLLs, and all my OLLs are sub 1.5. Including my reaction, my LL is usually high 3/low 4. Hopefully this helps!

Ok. Thanks. I average like 1.5 for plls and can sub-2 all of them. Not very easily for the bad n-perm but I can. Idk as far as OLLs go.

 

[NoobyCuber]

Any tips on planning out the cross from the inspection? That seems to be one of the weakest points of my solve.

Also, my biggest weak point in F2L is finding the pairs. Any tips on that?

 

[jeff081692]

Any tips on planning out the cross from the inspection? That seems to be one of the weakest points of my solve.

Also, my biggest weak point in F2L is finding the pairs. Any tips on that?

For the cross badmephisto has an excellent video https://www.youtube.com/watch?v=hY_FbpdNRCQ
One of the things that really helps with seeing the whole cross is learning more ways to insert two cross pieces together instead of one by one.
An example I recently learned was that if you set up the cube with F’ R u’ R U with cross color on bottom, then the inverse of that U' R' u R' F is very fast and for me it is faster than other ways of solving those remaining 2 cross pieces. And since then I have had solves where I could easily insert the first 2 cross pieces and get this similar pattern and while still in inspection I can tell myself to put that last cross piece in the back and do R' u R' F.


For finding pairs in F2L it really just comes from practice. The more you practice the more you can recognize pairs quickly and you will eventually be able to subconsciously know where you should be looking for the pairs.

When I play around with the idea of being color neutral I have a hard time finding pairs. Much more than with my normal cross color. I know where I should be looking but I just can't recognize the pairs as fast as I can with what I am used to. The only difference is that I have had more practice solving on green and blue than other colors. So I say just keep practicing and focus on trying to find the pairs and one day you will notice you are getting better and better and that is when you start beating your personal bests.

One thing I noticed is that I subconsciously block out pieces with blue when I solve a green cross. When solving a green cross I see a red and yellow edge and know that that is one of the pieces I am looking for. When I solve on a white cross (which is unfamiliar to me) I can't tell the difference between the importance of a green-orange edge and a green-yellow edge while in a normal solve. The only way to recognize the difference quicker is to get more experience with it.
So one thing you can try is to mentally block out pieces that are opposite of the cross you are solving and you might be able to better filter where the F2L pieces are.

So in short the more you practice the more you will be familiar with these parts of the solve overtime.

 

[NoobyCuber]

For the cross badmephisto has an excellent video https://www.youtube.com/watch?v=hY_FbpdNRCQ
One of the things that really helps with seeing the whole cross is learning more ways to insert two cross pieces together instead of one by one.
An example I recently learned was that if you set up the cube with F’ R u’ R U with cross color on bottom, then the inverse of that U' R' u R' F is very fast and for me it is faster than other ways of solving those remaining 2 cross pieces. And since then I have had solves where I could easily insert the first 2 cross pieces and get this similar pattern and while still in inspection I can tell myself to put that last cross piece in the back and do R' u R' F.


For finding pairs in F2L it really just comes from practice. The more you practice the more you can recognize pairs quickly and you will eventually be able to subconsciously know where you should be looking for the pairs.

When I play around with the idea of being color neutral I have a hard time finding pairs. Much more than with my normal cross color. I know where I should be looking but I just can't recognize the pairs as fast as I can with what I am used to. The only difference is that I have had more practice solving on green and blue than other colors. So I say just keep practicing and focus on trying to find the pairs and one day you will notice you are getting better and better and that is when you start beating your personal bests.

One thing I noticed is that I subconsciously block out pieces with blue when I solve a green cross. When solving a green cross I see a red and yellow edge and know that that is one of the pieces I am looking for. When I solve on a white cross (which is unfamiliar to me) I can't tell the difference between the importance of a green-orange edge and a green-yellow edge while in a normal solve. The only way to recognize the difference quicker is to get more experience with it.
So one thing you can try is to mentally block out pieces that are opposite of the cross you are solving and you might be able to better filter where the F2L pieces are.

So in short the more you practice the more you will be familiar with these parts of the solve overtime.

Thanks for the video and tips! I'll definitely try out some of these tactics in practice. I'm sure it will help a lot. Much thanks forever!

 

[aboeglin]

I have learned full PLL lately and improved my F2L which now leads me to about 32 ao100 on good days, my PB ao5 being 27.48sec. This morning I did an average on my F2L, it is barely below 23sec ( including the cross ) at the moment, which is a little slow for my global average ( I should be average about 19sec on my F2L ). I think my cross is quite good and I can sub4 it most of the time. While doing the F2L average I realized one big factor, I do a lot of cube rotations during F2L pairs, sometimes even 2 to solve one pair. For example, I might do a y or y' to get a piece out of a slot, and then another one to insert the pair. I never use any F or B moves, I do sometimes do a d or d' move if I see that my pair insertion is already ready but not on top of the correct slot. I just learned today another way of inserting which is R'FRF', but I'm not sure yet how it could help me reduce cube rotations.

I think these cube rotations and d moves are keeping me from progressing. I still gain perhaps half a second per day since I'm still working on improving my PLL algs and get much better at recog, but my F2L isn't improving much lately. Cube rotations make me loose track of empty slots and sometimes I track an edge for my next pair, but I have to do a cube rotation during the previous pair, and then I loose the corner, I need to look behind the cube again to make sure it was my corner. I think I now got quite confident with solving pairs, I know how I'll solve it once I see it, but I think the way I do this could be improved.

If anyone got tips on avoiding cube rotations I'd appreciate it greatly !

Happy cubing everyone.

 

[Dapianokid]

I have learned full PLL lately and improved my F2L which now leads me to about 32 ao100 on good days, my PB ao5 being 27.48sec. This morning I did an average on my F2L, it is barely below 23sec ( including the cross ) at the moment, which is a little slow for my global average ( I should be average about 19sec on my F2L ). I think my cross is quite good and I can sub4 it most of the time. While doing the F2L average I realized one big factor, I do a lot of cube rotations during F2L pairs, sometimes even 2 to solve one pair. For example, I might do a y or y' to get a piece out of a slot, and then another one to insert the pair. I never use any F or B moves, I do sometimes do a d or d' move if I see that my pair insertion is already ready but not on top of the correct slot. I just learned today another way of inserting which is R'FRF', but I'm not sure yet how it could help me reduce cube rotations.

I think these cube rotations and d moves are keeping me from progressing. I still gain perhaps half a second per day since I'm still working on improving my PLL algs and get much better at recog, but my F2L isn't improving much lately. Cube rotations make me loose track of empty slots and sometimes I track an edge for my next pair, but I have to do a cube rotation during the previous pair, and then I loose the corner, I need to look behind the cube again to make sure it was my corner. I think I now got quite confident with solving pairs, I know how I'll solve it once I see it, but I think the way I do this could be improved.

If anyone got tips on avoiding cube rotations I'd appreciate it greatly !

Happy cubing everyone.

The only thing preventing you from progressing is yourself. :3 It's just a matter of practice, but not mindless practice: Practice for a GOAL! Get better at a specific thing, not just at speedsolving the cube itself. Just wanting to be sub-20 avg won't get you there, you have to know what that takes.

Sledgehammer is an excellent resource for two things: edge control, and fluidity/reducing rotations. If you are about to rotate, using the right insertions correctly will make that seamless and the rotation won't be a hinderance. The MAX number of rotations it should take I think is 4 to solve F2L, 0 for LL, and 1 for cross. I think that 2 rotations is all you usually need for F2L, though.
One thing people at your level don't grasp right away (it's basically a secret of cubing but not really a secret ) is that you can insert a pair that is built into the back slots without rotating. I'm assuming you can already insert on the left. Well, you can also insert the back pairs. Do R'U'R on a solved cube to see what I mean.
Basically, right now, it's just about learning the shortcuts and familiarizing yourself with how to do things in a more finger friendly way.

 

[aboeglin]

I already do inserts in the back slots, and left slots of course, I do practice a lot and that's exactly why I originally posted, I want to improve my F2L. And sledgehammer seems to be a good thing to dig into.
So basically, the advice would be to get better back slot inserts and play a bit with sledgehammer ? It's actually funny because even though I used this trigger a lot in various LL algs, I never heard about it for F2L before today, I just found out about it on a 5 video series that someone posted on the forums. I start to understand the idea of edge orientation and what sledgehammer is good for. I just don't seem to get a good way to perform the F' at the end, it's like I sort of need to change my hands' position before performing the whole trigger if I want it to be good.

 

[jeff081692]

If you want less rotations you can try looking at F2L algs on the wiki. As long as you stick to at a maximum 1 rotation per pair you can still get very fast.
I personally have a list of all my F2L cases and I just counted around 20/42 cases where I insert into the front right slot has a rotation or d.

For the cases where you use 2 rotations can you post the exact set up moves for it and what you did to solve it? We might be able to optimize it better for speed by making it a 1 rotation solution and maybe even rotation-less. My advice is if you are doing something wrong look at how other people handled the same problem. So either look up cases, try to figure out better solutions on your own, or ask us about specific cases. You have to tackle this problem one by one until you get the amount of rotations down to where you desire.

 

[aboeglin]

I've been playing around a little tonight, only slow F2L while watching tv. I tried things with sledgehammer as well, I kind of understand the edge orientation flip to allow me to solve it only with R, U, L, even though sometimes it leads me to not optimal setups, so I guess I need to learn when to use it or actually rotate the cube. But one thing that edge orientation learnt me, is that I can more easily recognize if I'll need a rotation or not, and it helps me to get a better solution than what I had before.

The pairs where I used two rotations were the ones where I usually had a piece in a slot, so one rotation to take it out with RUR' for example, and then I had to do another one usually to insert the pair ( which I would now call a bad edge case ). At least it helped me to understand it a step further I guess. Like now I would totally take an edge out using F or B move if it allows me to get it as a good edge, because then I know that I avoid a cube rotation and I can solve it with R, U, L. Before, I would have done this cube rotation, take my edge out, and it would have been a bad edge, I think that's the bad cases I was loosing a lot of time with. And, I could finally get slow solves down to 0-1 cube rotations, without using d or d'.

So, I suppose I'll take the next step with my F2L and get more into tracking down bad cases and reading / testing more with sledgehammer and edge orientation. Thanks for the guidelines and hints guys !

 

[jeff081692]

Instead of rotating then doing R U R' to take out a piece use either L' U' L for taking a piece out of the FL slot. Use R' U R to take a piece out of the BR slot and L U' L' for BL pieces. This would eliminate all your unnecessary rotations and leave you with just the 1 rotation if it's one of those cases where it makes sense to rotate.

 

[aboeglin]

Yes, I can now do F2L without rotation on slow solves, using sledgehammer and hedgeslammer. But my F2L got much slower. I really try to force myself to solve the pairs only using R,U,L, when my edge is correctly oriented, otherwise I either would do a cube rotation ( I now know that one is enough since it'd translate a bad edge to a good edge, making it possible to solve only with R,U,L ). The problem I have using R,U,L is that in some cases, especially when solving the last two pairs, I have some cases where I find the move count with R,U,L quite high, or said otherwise, there would be an easier approach probably by pairing the pieces with the edge being a bad edge and doing a cube rotation to insert and/or use sledgehammer to insert ( which can orient OLL edges depending on the situation ).

I'm just not really good at recognizing F2L cases and find it hard to look cases up as I really learned it all intuitively so far. But I at least spotted one case that's been driving me nut for some time now and the solution is : y' (R' U2 R) U (R' U' R), the case can be found here : http://badmephisto.com/badmephisto-speedcubing-method.pdf ( 9th case from the top on the left side of the page but right column : corner with white on top, misoriented edge matching its center ). That's the exact case where I think the y' at the beginning makes it much easier than a sledgehammer.

So, I'll try to keep solving with ways that are better than what I was doing so far, at the moment my F2L got totally messed up with the timer ( cannot get it done in less than 30sec anymore ), but I believe it's because I have to relearn some cases. Once I'll have the new moves in muscle memory I'll be back to where I was with better tools for look ahead and further progress.