[Help Thread] F2L Intuitive vs. Algorithm Solving

[JL58]

My summary of this is that intuitive F2L (pair insertion, really) is about:
- using empty slots (you might want to read Dr Pochmann's thesis, he can't even figure out every time what the optimum solution is - Stefan, I read it 3 times already, great job!)
- dealing with edges and corners stuck in the wrong slot
- performing insertion from any angle (saves costly grip-changes)
- trading cube virtuosity for anticipation (to go fast you have to slow down, they say

 

[4weeksAndImSub60]

Yeah, I tried learning F2L algorithmically and I gave up on it after the 2nd day. Far too many algs and you won't understand how your solving it. Personally, i like a good challenge, and if that's what you want.. go with intuitive F2L. You'll get a lot better solving F2L, I'd even say you'd get better faster with intuitive over algorithm. Plus, with algorithms, you only learn a specific case. Meaning, you won't learn how to solve F2L pairs from any side of the cube. So, with algorithms you'll come across cases where you have to rotate the cube to set up that case and then proceed by mindlessly executing an algorithm.

 

[cubekid]

Just like pretty much everyone here has said, learn intuitive F2L first. In all honesty, it took me maybe 2 hours to understand what the basic idea is, and then about 2 weeks to get the feel of it. Yes, the solving initially wasn't optimal. Yes, it's hard. But all in all, if you're going to continue using CFOP, understanding F2L is absolutely necessary because this is where the largest time drain occurs, and thus it is the most important step.

My suggestion is that after you go through some tutorial for intuitive F2L and have become comfortable with the process, then go and start looking for better algorithms. But even then, don't even think of learning algorithms for F2L until you are sub 1minute.

It's all about prioritizing what will give you the most value for your time. When you're around 1:30, learning ZBLL won't help you nearly as much as just learning 4look LL (just cause it will take so much longer to learn ZBLL, while 4look LL takes maybe a week)

anyway, just my $0.02. You'll thank yourself when you get in the 20s if you learn intuitive F2L, since it's much easier then to learn just a few algorithms then. Plus, you'll feel better about yourself

 

[4weeksAndImSub60]

One other thing you may want to know is that intuitive will help you with advanced empty slot knowledge and solving the cube from all angles. Like, there are some cases that I know how to solve using like 4 different algs, and I choose which ever one is faster depending on where the slot I must put my pair is located. So, yeah.

 

[mmiddymaster27]

Im a mediocre cuber. Im relatively new to cubing. About one month. I'm down to sub :50 with 2 look OLL, Friedrich Intuitively F2L, and PLL. I am good at memorizing the algorithms, rather than performing them quickly. Do you think it would benefit me in learning algorithmic f2l? I already know intuitive.

 

[josmil1]

Im a mediocre cuber. Im relatively new to cubing. About one month. I'm down to sub :50 with 2 look OLL, Friedrich Intuitively F2L, and PLL. I am good at memorizing the algorithms, rather than performing them quickly. Do you think it would benefit me in learning algorithmic f2l? I already know intuitive.

stick with intuitive f2l...but learn the algs for any cases that you think are difficult. You can get much lower to sub 20 with 2 look oll, pll and intuitive f2l

 

[CuBeOrDiE]

Personally, I also think solving without algorithms simply gives you more self-satisfaction. But I agree with everything above-said.

 

[Joël]

I just noticed this thread, and I'd like to add my take on this.

First off all, the fact that you realise the difference between learning algs and executing them brainlessly and learning the logic behind algorithms shows me you are on the right track. I think both elements are very important in cubing.

I think you should teach your hands the algorithms, and your brain the logic behind them. In other words, you should be able to execute the algorithms without thinking... This is because you should execute an algorithm while also performing another task: looking ahead for the next pair. But, you should also understand how each and every single move of the F2L algorithms is needed in order to solve the pairs. This is because there are not just 41 cases (at least that's not how I see it), there are all kinds of special cases where a piece is in a different slot, or some F2L pairs haven't been solved, allowing for quick shortcuts, and you need to use the same basic idea of the algorithms, but make a slightly different version... Understanding the logic of the F2L algs will help you to be flexible enough to minimize the move count in those situations.

 

[cmhardw]

Here are my two cents.

I learned my F2L by memorizing all the algorithms, and over time I developed an understanding of how each and every one of the algs works. I currently use empty slots frequently, and I allow my solving order to be very organic (Xcross as well as partial cross solves are frequent for my solving).

I don't buy into this idea that you must learn intuitive F2L first, and then later on learn algorithms for the more complex cases. If the goal of F2L is to be able to execute a situation mindlessly enough as to allow lookahead, then you need for every case you use to be deeply rooted in muscle memory. So if you learn an intuitive alg that is a "bad" alg (high move count, does not utilize the optimal empty slot shortcut, is not 2-gen when a fast one exists, etc.) then you will deeply root this "bad" alg into your muscle memory. Later on, once you discover what will become your best speedcubing alg, you must unlearn your old case and begin the process of committing the new alg to muscle memory. It is very difficult to change your brain's perception of a case from your old alg to a new alg, trust me on this.

I'll give you an example. When I first learned to sport stack I learned the cycle routine by watching Emily Fox's then World Record solve video in slow motion. I copied her hand motions exactly (only reflected for left hand) in every way I could. I did this so that I would begin implanting this "good" method for stacking into my muscle memory from the very beginning. I have never done another stack routine other than this one. I feel cubing should be handled the same way, but again that is only my opinion.

Understanding the F2L well enough to solve organically using empty slots, solving pairs before finishing the cross, doing something advanced like occasional Xcross or multi-slotting, using 2 gen cases when they're faster, etc. is extremely important. But as a beginner/intermediate level solver do you think that you will intuitively discover the optimal/best alg for each and every situation right from the start? If you can, that is excellent! But if you are like us regular humans then why risk learning and fully committing some, or possibly many, sub-optimal algorithms into your muscle memory? Then you must spend effort later on to unlearn the bad habits you have developed right from the very beginning. In my opinion the effort spent unlearning the bad habits you developed right from the beginning of your F2L is MUCH more than the effort of initially studying the F2L cases from websites and learning what others do right from the start. How do I know this? I learned many very bad habits with my turning style (wristing, very little 2-gen) from the very beginning and I have spent years unlearning them. I have not even fully unlearned all of my bad habits, but I am making continual progress on this. Why spend months, if not longer, unlearning bad habits when you can just learn it correctly right from the start?

Again, just my opinion. Don't waste your time with studying only intuitively F2L at the start, and adding the algs to that later on. Just learn the algs right from the start now, but also study them so you can develop an understanding of how they work.

Chris

 

[frogmanson]

hmm i didnt know what was intuitive f2l was when i first started so i learned all the algs. but imo both will be the same. I eventually figured out how to do every algorithm i learned from every angle etc. but from what I see from newer cubers using intuitive f2l some are 35 40+, while I was never higher than around 27 ( i learned f2l pll then oll fyi). so yeah intuitive you "learn" less but it'll be hard to recognize cases but algorithms you'll "learn" more but you will recognize faster + you will eventually learn to do them intuitively.

also, some really nice cases are very difficult to figure out intuitively especially with pieces in wrong slots. hmm only 1 alg pops into mind atm because its the only 1 I can write down xD. F U r U r' F' oh and this 1 too since its 1 extra move xD F U r U r' U' F'

 

[Rayne]

Once you get good at F2L intuitively, once you see a case you can execute it as if it were an algorithm, because you've done it so many times it's in your muscle memory.

 

[iasimp1997]

Don't learn F2L yet. Wait 'til your sub 1 minute. Then start learning. You should really wait until your really good at Beginner.

 

[FatBoyXPC]

When people refer to beginner, are they referring to LBL or Keyhole?

I somewhat disagree simp, because it is quite frustrating when you are doing your method, learn a new method and tack 10 seconds onto your time. Granted when I started doing Fridrich F2L I was mid 40s average already, didn't break my 32 PB at all, in fact, rarely came close. I also didn't know anything about look ahead, or try inserting from different angles. I solved with Keyhole the other day with different angles and it was mid 30s, but a lot has improved since then (such as 2LOLL had worse algs back when, and I didn't know the E perm at the time to fix opposite corners, I just did A/Ab twice to fix it).

I think if they start learning intuitive or alg F2L now. However, I'm not sure I agree with Chris about that 100%. I think you should learn how to do it intuitively, but don't commit it to memory. Just learn the different cases on your own, and learn the algs for the others. I mean most cases aren't hard (RUR' and it's mirror, etc). I'd actually recommend just looking at the F2L algs page, setting up that case, then doing it intuitively, so you can see how it works. I suggest you then learn Alg F2L. I started looking at how I do my F2L cases, and realized a lot of them have a lot of moves (might explain my high F2L times) and realized there were quicker ways to do them. I now often look at my algs page just to see if I can make it any better.

 

[Chapuunka]

The way I see it, the algorithms make it easier at the beginning (assuming you're good with algs, although I'm not), but one intuitive F2L clicks, it'll be faster than the algs. It just takes more practice...eventually you'll figure out more optimal cases, and you'll be able to do them from any angle without learning new algorithms. Not to mention you'll figure out what works best for you and your style.

And all this could easily be achieved through the algorithms, but you'll have to go through and actually understand and dissect them, not mindlessly execute. But then again, I never learned the algs.

 

[averyb]

- I can't lookahead using intuitive


that is wahat happened to me about two weeks ago, during thos two weeks i was practicing for a tournament tomorrow and we had to use storeboughts in it so i was using them and they moved so slow that i was able to look ahead. then i found a super fast storebought and i could look ahead now. it worked for me so it might work for you

 

[y3k9]

Algorithmic f2l???

I couldn't find any similar threads so I decided to make this one... So what I want to know is if I should learn all the algorithms for f2l? Does it help? I would be willing to learn all of them, it's just I want to make sure they'll help.

 

[MichaelP.]

Ugh, the reason you can't find similar threads is because it belongs in the OAQT and your answer can be found on the wiki.

Fridrich F2L
There are many ways to solve the 'F2L' on a cube. A common system is using the Fridrich method first two layer approach. After solving the cross, a corner-edge pair is paired up, and then inserted into the correct slot. A total of four corner edge (or 'CE') pairs are made and inserted to solve the first two layers.
The concept of pairing up four corner/edge pairs was first proposed by Ren Schoof in 1981.
Most people agree that the Fridrich F2L should be done intuitively, but when you are starting out it can be instructive to see some optimal solutions for each F2L pair. Below is a collection of such optimal solutions.

- Taken from the wiki's F2L page.

 

[Zarxrax]

Yea this has been covered lots of times before, but here's the answer:
Learn intuitively, then after you are very comfortable with f2l, go through the algorithms and analyze them.
For many cases, your intuitive solution will probably be the same as the algorithmic solution.
For those where your own solution is different, look at the alg, try to understand how it works and whats going on, and then decide if it is faster than your own method.

 

[Cubezz]

I say learn it intuitively. You will learn a little more if you take the extra time to try intuitive, imo.

 

[y3k9]

Already now intuitive backwards and front. Link me to a similar thread?