Speedcubing is an Art & a Science

As far as getting faster is concerned, there are certainly many “rules” that need to be followed. That said, at a certain point it WILL make sense to “break the rules”

And no, I'm not talking about the WCA rules, those ones you have to follow... Even the ones like turning your phone camera so it faces away so you don't decide to cheat by… looking at a small screen to see the cube in a mirrored state instead of looking at the actual cube???

Sigh

But I digress

What I'm referring to here are of course the "rules" of what makes for good and bad decision making within a solve

When to stick to the fundamentals and general principals that work more often than not

And when to go against the grain (intelligently, of course, as opposed to smearing poo on the walls of your schools bathroom, which I'm sure in most places would be considered against the grain)

A couple of *mostly* useful 3x3 CFOP rules that immediately come to mind for me are...

• 1. Never y2, you can just solve from the back instead

• 2. Don't inspect something obscure unless the 4 cross edges are solved after executing what was inspected

• 3. When you can see and choose between a rotationless pair and a pair involving a rotation, avoid the rotation

• 4. When choosing between a free pair (3-4 move insert) and a non free pair, choose the free pair (same general logic also applies to "any efficient pair" vs "any inefficient pair", basically just choose the more move efficient pair, M'kay...)

I could go on and on but I don't want this to drag on any longer than it has to, so lets just get to the thesis (don't worry, I'll be showing more examples of breaking "rules" near the end of this email)

Cubing is both an ART and a SCIENCE

I remember first seeing this quote when I read Ian Schefflers "Cracking the Cube" ages ago (Highly recommend giving it a read, guarantee you'll learn something about cubing you didn't already know)

For whatever reason, this quote stuck with me and recently I figured out why

When it comes to getting faster at cubing, there are pieces of pragmatic advice that exist because they're more often than not the best piece of "one size fits all" advice 

This would be the science of cubing as it were

However, just like any general principal that exists anywhere in the world, there are always exceptions to the rule 

Whatever it may be, exceptions to any and all universal rules do exist 

BUT, they don't negate the rule as a whole and the rule still exists and is universal for a reason

You can't think outside the box unless you know what's inside the box to begin with 

This is why I get frustrated with people who ask if they should learn full ZBLL or switch to ZZCT when they don't even know a full basic speedsolving method down to its fundamental steps (just fix your damn OLL PLL and F2L algs before worrying about the viability of VLS from the left slot you damn egg)

Anyways, back to the ART/SCIENCE stuff, because I’m not lecturing you on the importance of the fundamentals today

Take this scramble for example

D' R2 U2 B F2 U2 B D2 L2 B D R U B2 D' B2 L' R D

x2

D R' F R2

3/4 cross, breaking the rule of "not having cross solved before end of inspection. However this isn't where the inspection ends, it ends after solving the 2 red pairs, which are relatively easy to track

U' R' U2 R U' R' U R

Notice how because there's no L move done, the way the steps flow together is much nicer on the hands in terms of regrips done 

Doing the L after the first D would make the first pair much shittier in comparison, and doing it after the R2 just makes for very jagged flow in terms of fingertrick requirements

U2 R U' R' y' U R' U R

Here is where inspection ends.

The other rationale behind leaving the cross edge until later was in the hope that it would set up the 3rd pair nicely, especially since it wasn't at all apparent that doing standard cross was going to set anything up

U2 F'

And voila, cross is now done post inspection! Definitely a major breach of rule #2, however there's a lot of rationale behind this as you can tell

U R U R2' F R U F'

Breaking rule #4 because we still preserve the free pair anyways 

Plus having a built pair ready for the FR slot makes VLS available which is quite nice if you know VLS. If you don't...

...I also don't think any of the 3 reasonable F2L cases you would have gotten by inserting

blue orange sensibly would beat R U R2' F R U F'

For fun, let's pretend that in the solve, before this line starts, that we add a y rotation 

The green orange pair can still be solved rotationless (according to ZZ law, which is the most useful definition in this context) 

However this would absolutely destroy the free pair

Here we already have a conflict between rules 3 and 4, since one contradicts the other. Literally no matter what you do, a rule HAS to be broken here!

Anyways, the blue orange pair requires a U F U' F' insert, which although not bad, leaves a crappy last pair and BL as the last slot which is bad pair prioritisation since BL is generally *the* blindspot for right hand dominant cubers

U2 x U' r' U' r U2 x' R U R' U' F' 

One thing I want to point out here is to do with rule #1 about avoiding y2. A great example of when you should y2 is if you were to have a U L U' L' insert as your last pair and you know VLS

If you replace the U2 at the start of the line with Dw2, looking at this scenario, if you knew VLS do you really think a simple y2 that takes maybe .5 tops is going to outweigh saving potentially up to a second of more off your total solve time?

This particular VLS doesn't save that much time compared to the standard OLL, but again, a y2 when you run the numbers really ain't bad at all (There's a clip on MentalBlockTV somewhere of Aiden Bartlett doing a y2 in the 3x3 final of the pre US nats 2018 SnoCo competition in like .2x, which is about the fastest I can get while practising on a stackmat)

As an added aside, I wouldn't be surprised if some 7-12 move BL F2L cases could be done faster with a y2 followed by an FR solution (though if you have your LUgen turning under control this is kind of a stretch)

U2 Gperm done

Now let's show an example of a solve breaking even more rules and conventional wisdom, while still being a "good" idea and leading to a good result

A while ago I broke my on camera stackmat PB with a 4.26

Anyone reading this who's seen the video or the reconstruction in CF knows just how ridiculous this solve is on so many levels

Not only was the skip ridiculously lucky, but the way I actually got it was very, shall we say, unconventional

D F2 U' B2 D R2 B2 U2 B2 F2 U' R2 B' R B L B F2 L2 R F2

x'

R' U' R' U R U R' U L2 - That's right, instead of solving a cross like a normal person would (and should), I did an XXcross with two of the edges placed opposite of where they should be

Truth be told, using M2' U2' M2' cross as a legitimate technique is so much worse than people realize. Most of the time, solving the cross like normal wouldn't amount to more time being wasted than you would waste by doing M2' U2' M2' later in the solve. This was a very rare exception

Still, M2' U2' M2' does have the added benefit of being able to manipulate your F2L, OLL and PLL cases a great deal. During F2L you can turn ugly cases into nice 7-8 movers very often, and during PLL you can basically do "PLL parity" in the sense that you do it from an angle that forces the better PLL using your 4x4 solving knowledge

You can also turn some bad OLL cases into great ones, like turning R U R' U R U2' R' F R U R' U' F' into f R U R' U' f'

Bottom line, M2' U2' M2' isn't the worst technique, but you have to be able to justify doing it for more than just a slightly easier cross, or because you think it will make a case in your solve better (and this doesn't happen every time)

L' U2 L2 U L' - I did see this pair in inspection, but somewhat second guessed myself at the time which is the main reason why I didn't cancel

Looking back on it however, I don't think cancelling would have been the optimal play, as silly as it might sound

Executing L' U2 L2 U L' from home grip is insanely easy, whereas cancelling into it would have definitely lead to at least 2 ugly ass regrips instead of one

Not the most "conventional" line of thinking, but I think it makes a *lot* of sense here, even though generally cancelling moves will be the better thing to do

y' M2' U2' M2' - At this point in the solve, I noted that I had a special case where I could turn a 7-8 mover into a 3-4 mover, so I did

Part of me probably also just wanted to fix the damn cross already, but I've definitely had solves where I left it alone until LL since using M2' U2' M2' earlier wasn't going to really give me a "boost"

U' R U2 R' - This was just lucky, not gonna lie

Hopefully with that, you get the idea now of just how ridiculous and deep the art of cubing can get

...along with how useful it can be as well

However, let me make this absolutely clear. The science of cubing should always come first, and any attempt to meaningfully explore the art of cubing will lead you astray and form bad habits

At the very least you should try to work towards high level competence, if not near mastery of the science of cubing.…

I can certainly teach the science of cubing, which is what I do with my online course business aka Speedcubing Solution. The art, however, I cannot as easily teach. Nor do I try to in the rare event that I think it's helpful to a client of mine

Instead, I prefer to inspire some ideas in my clients who I think are ready for more artistic exploration! However, only they can do the exploring, since no two cubers solve the same way or have the same style

If anything, learning from other cubers and their artistic choices often expands my horizons more than attempting to teach them my artistic ideas (which can definitely go against a persons preferred solving style)

Art cannot be rushed, nor can it be easily taught beyond the scientific framework that it may or may not fit within

I can only encourage you to think outside the box and not limit yourself to following dogma that, although can be very helpful with cleaning up your solves, will only limit you in the long run if you want to be truly great

...But of course, if you want to think outside of the box, know what's inside the box ;)