[Help Thread] F2L Intuitive vs. Algorithm Solving

[Litz]

Sorry for late reply.
So i have been practising Intuitive f2l for around a week. Evceryone recommends turning slowly with look ahead. But no matter how I try, I cant look ahead.The peice iam looking out for changes its position as soon as i insert the first pair and it is difficult for me to track 2 pieces at once. I have very insignificant improvement sice i started learning it. Some hints please ? ( "practise practis practise"- this i will do but iam not sure whether it would do any good.)

You just need to practice. Your problem is simple really. You say you started using intuitive f2l one week ago. Odds are you still need to pay attention to the pieces your are solving because your brain is still thinking of how to solve it. With practice, you start using your brain less when solving pieces because your fingers just know what to do instantly (every case basically turns into an algorithm). Once that happens, your brain is free to focus on the remaining pieces and think how they will end up after solving a pair, as solving the pairs will be on auto-pilot.

Don't worry too much about looking ahead right now. Focus on recognizing the cases and figuring out how to solve them as fast as possible without thinking.

 

[jayefbe]

Just backing up what Litz said, you can't look ahead until your F2L cases are so ingrained that they're practically automatic. It sounds like right now you should be focusing on F2L efficiency (every case should either be extremely finger trick friendly or ~8 moves max), and then getting to know the cases so well you can do them in your sleep. One thing I found helpful was doing F2L blind. You find a corner edge pair, close your eyes, then pair them and insert them. Open your eyes again, find another pair and repeat. If you can do that quickly and consistently, then you can start finding and tracking other corner edge pairs when solving the previous pair.

 

[crazycubes]

Thnx for the advice

My times are improving so is my lookahead. I have hit sub 1 min and am looking forward to get sub 30.

 

[PokeCuber]

I personally think that it is best if we learn to do F2l intuitively (practice and experience of >200 solves) and then move onto the algorithmic f2l and correlate the both by learning 1 algorithm per day. If you get Excited and try to learn more per day(like I tried to do) ,you will endup learning ZERO. So keep practicing and good luck to all of you.

 

[pipkiksass]

I personally think that it is best if we learn to do F2l intuitively (practice and experience of >200 solves) and then move onto the algorithmic f2l and correlate the both by learning 1 algorithm per day. If you get Excited and try to learn more per day(like I tried to do) ,you will endup learning ZERO. So keep practicing and good luck to all of you.

I didn't learn any F2L algorithms at all until I'd done a few thousand solves. I still only use two algorithms, both for cases where the pieces are in the slot, but the corner is not correctly aligned. The rest is all intuitive, as it should be.

 

[guysensei1]

Do intuitive for every case except for the ones you find tough.

 

[TDM]

Do intuitive for every case except for the ones you find tough.

This. Although sometimes, looking at algorithms can be very useful, for example I was using [U'] R U2 R' U2 R U R' for a while before I looked at the wiki. I thought it was a good case, and I didn't know there was a better alg for it. But definitely, if there's a case you find particularly slow, look for another alg e.g. setup with R U R' U F' U' F. You can solve that with fast and rotationless algs such as l F' R U' R' U l'.
I also found looking at reconstructions helpful. I've found two algs for cases from looking at Feliks' reconstructions. I still can't find any reconstructions with two particularly bad cases, but then again the best people probably know how to avoid them...

 

[EMI]

I learned F2L completely with algorithms. That has never been a problem, I still always know what each case is doing. I don't believe it is important how you learn F2L, except maybe in the very beginning.

 

[mark49152]

e.g. setup with R U R' U F' U' F. You can solve that with fast and rotationless algs such as l F' R U' R' U l'.

Or U' (R' F R F') R U' R', one of my favourites. There are lots of great F2L tricks, many of which could be intuitive if I was smart enough to see them. Algorithmic and intuitive isn't black and white. It's always worth looking at algorithms as examples and seeing what new tricks you can absorb into your repertoire, but always try to understand how it works so it becomes as intuitive as possible.

 

[Lazy Einstein]

For me I used intuitive F2L because I wanted to learn 1LLL first.

So I did intuitive F2L to learn how the cube moves while trying to increase my look-ahead by training with the added difficulty of considering how to pair and insert F2L at the same time as tracking and looking ahead. It makes the look-ahead training more intense but it forces me to always think about look ahead and try to track pairs while doing F2L intuitively when stressed or overwhelmed. Hurts the brain at first but seems like with practice it will be extremely beneficial for look ahead(especially since I solve CN). Then when you get to LL you have 78 LL algorithms to try to remember.

By the time you learn LL algs you should have good look-ahead experience and can start learning F2L algs if you wish. Although I'd probably start working on VHLS, WVLS, COLL, PLL Skip Cases, increasing PLL Recognition for minimal pause time and rotations, and learn multi angle PLL and OLL algs to prevent AUF for transition of previous stage.

If you do that you should only have been cubing 6 months to a year with incredible LL understanding, control, and speed. Plus you will have had good look-ahead training so when you start F2L algs you will already be trained to always be looking ahead when thinking about F2L algs.

I am still a nub myself but that is how I have been doing it. Now I am 2 1/2 months in cubing, know 78/78 last layer. PB is only 24.xy but I have good CN recognition for F2L pairs from any cross. If I did F2L algs I doubt I would be at this point with last layer and Recognition.

I am just going to work on look ahead now. Instead of starting F2L algs I am going to work on optimizing my last layer algs(i.e. changing the ones the don't feel good to me, and doing the mirrors to stop AUFs).

I think you have to pick what makes it most fun for you because that is what will make you keep training.

 

[aboeglin]

For me I used intuitive F2L because I wanted to learn 1LLL first.

So I did intuitive F2L to learn how the cube moves while trying to increase my look-ahead by training with the added difficulty of considering how to pair and insert F2L at the same time as tracking and looking ahead. It makes the look-ahead training more intense but it forces me to always think about look ahead and try to track pairs while doing F2L intuitively when stressed or overwhelmed. Hurts the brain at first but seems like with practice it will be extremely beneficial for look ahead(especially since I solve CN). Then when you get to LL you have 78 LL algorithms to try to remember.

By the time you learn LL algs you should have good look-ahead experience and can start learning F2L algs if you wish. Although I'd probably start working on VHLS, WVLS, COLL, PLL Skip Cases, increasing PLL Recognition for minimal pause time and rotations, and learn multi angle PLL and OLL algs to prevent AUF for transition of previous stage.

If you do that you should only have been cubing 6 months to a year with incredible LL understanding, control, and speed. Plus you will have had good look-ahead training so when you start F2L algs you will already be trained to always be looking ahead when thinking about F2L algs.

I am still a nub myself but that is how I have been doing it. Now I am 2 1/2 months in cubing, know 78/78 last layer. PB is only 24.xy but I have good CN recognition for F2L pairs from any cross. If I did F2L algs I doubt I would be at this point with last layer and Recognition.

I am just going to work on look ahead now. Instead of starting F2L algs I am going to work on optimizing my last layer algs(i.e. changing the ones the don't feel good to me, and doing the mirrors to stop AUFs).

I think you have to pick what makes it most fun for you because that is what will make you keep training.

This is funny, I did it the exact other way. Not that I learned any alg for F2L ( yet ), but I rather focused on F2L than LL. I always heard "LL is something you have time to work out, what keeps you back from doing good times is F2L". And I believe it myself as the most progress I did was when I improved my F2L. I spent a few days of this week learning the 15 algs of PLL. I was averaging 37sec ao100 on Monday, I am now at 39 since I'm not as fluent at PLL than I was at 2-look PLL ( recognition isn't as good at the moment and I don't perform the 15 freshly learned algs as the 6 I already knew ). Back on Monday I was averaging 22-23sec for my F2L, which means my LL was about 15sec which was still a lot, I thought I could probably gain 3 to 4 sec learning full PLL, which will probably pay off in a few days once I trained my recognition, but still, all the way I'll keep focus on F2L, PLL will come by itself.

I started cubing the 5th of December 2013, my F2L still improved last week, I see more and more 17-19sec F2L coming up, so I suppose I'm soon to get sub-30 ao100, and only then I'll start learning OLL algs, dots mostly at first since they add a few moves when doing 2-look OLL.

And I totally agree with your last statement, the most important is to keep the fun and the motivation. Whatever you feel like practicing is better than no practice anyway. But all in all I don't think LL optimization should be a focus before sub25. Considering you know full OLL and PLL I'd say your LL might be anywhere between 6 and 10sec including recognition, which means you still take 30sec in average for you F2L. So if you want to improve your times I'd rather go for cross and F2L pairs practice rather than adding algs to your LL.

 

[LaraMJ]

Actually I think intuituve is the best way, you can see how the cube actually works, I personally recommend to just learn specific ones, by example, a misoriented edge that is in it's slot or the same with a corner or even both. Using those and their simetrics and/or reverse I am able to do some OLL skips.

 

[RustyCube]

Is initiative the way?

I used to have a rubik's cube about 3 years back and learnt to solve it with the beginners method but using the keyhole method for the F2L. anyway 3 years down the line I picked up a new cube yesterday for the first time and I'm on about 1:20 average. I looked at some of the methods but they are just too much of a wall of info at the moment, so I'm considering trying to learn an initiative version of Fridrich. Would that be beneficial to me? Or am I just dooming myself to never go sub 1:00.

*Just tried solving it using that type of method and got 1:19 which is the fastest time I've had since getting my cube yesterday*

 

[Jaycee]

I believe the word that you're looking for is "intuitive".

Fast-ish times can be achieved with keyhole, but the next step is intuitive F2L. It's one of the hardest things to master, but it's also one of the most important, because for CFOP(Fridrich) solvers, F2L takes the majority of the solve. I recommend practicing intuitive F2L whenever you can during a solve, and if there is a case that you can't figure out on your own, you should look on the Wiki here for algs for the case. Just make sure you understand why the algs work (you're going for intuitive, after all).

 

[Rocky0701]

I don't know what you mean by "Initiative", but if you are looking to get fast, i recommend Fridrich, it is by far the most used method and most fast 3x3 solvers use it. You can still get really fast with Roux however, it is less algorithmic and more intuitive. You will have to ask others on the forum about Roux, but as for Fridrich, i would get your beginners method down to about a minute, then learn 2 look OLL and then 2 look PLL. After you can get your average with that to about 50 seconds, learn intuitive F2L.

 

[mDiPalma]

Yes, take initiative. Learn ZB.

 

[Eva]

I think Fridrichs is a good way,because there are a lot of steps witch can make you faster (not only f2l).
My f2l sucks and I'm around 40.
A good guide to learn the Fridrichs method is http://www.speedsolving.com/forum/showthread.php?6085-How-to-Get-Faster-using-the-Fridrich-CFOP-Method.
It go's step by step,but it doesn't begin with F2L,that's step 3.

For F2L itself I would use the intuitive way of the Fridrichs method:http://www.youtube.com/watch?v=k-xbcAMfWwM witch is also used in the guide above.
If you're gonna learn intuive F2l your times will temperately get a lot higher (so above now),but then they jump down.
Although I wanna suggest that you learn 2 look OLL/PLL first and if you've mastered that you're definitely sub 1 minute.
In the guide above there're good links for tutorials and stuff.

Good luck

 

[Rocky0701]

Pretty much every tutorial by Badmephisto will help you loads.

 

[manstrong]

Just spend an hour trying to get the first two layers. Don't waste time looking at someone else's algorithms when you can instantly memorize your own.

 

[AFatTick]

F2L algs are just so short... I don't see how they can be no intuitive