Finger trick notation

[Dane man]

What about this fingertrick used in a G perm?
R2 u' R U' R U R' u R2 B U' B'

For the u move I use the back of my right ring finger on an edge sticker. Does the notation cover that kind of move?

Yes, it does (though I'm not exactly sure which edge you mean). For the front edge:
R2 u' R U' R U R' u(R3) R2 B U' B'
If it's the back edge that you use, then it'd be:
R2 u' R U' R U R' u(R2) R2 B U' B'

Assuming u indicates a double layer turn, then R indicates right ring finger, and 3 (or 2) indicates the 3rd (or 2nd) clockwise corner of the combined u face. Being that you can only push from the edge to cause a double layer turn, you'd be forced to push on the edge to make it work, therefore with double layer turns, the notation will apply to the edge "behind" the corner indicated (if that side was facing you). Another example would be the very common d'(M3).

 

[Renslay]

This all reminds me the history of muscal notations - everything was in chaos, and each country and region (or even school) used different symbols before the modern five-line staff notation spread. (And there are still many different notations as well.)

 

[Dane man]

This all reminds me the history of muscal notations - everything was in chaos, and each country and region (or even school) used different symbols before the modern five-line staff notation spread. (And there are still many different notations as well.)

That happened with cubing notations as well. To this day there are still plenty of people who use d instead of Dw when it comes to the 3x3, even though WCA chose Dw notation. I remember more than ten years ago when people had all kinds of crazy cube notations on the internet. Gladly, one was finally whittled down and standardized, even for bigger cubes.

 

[XTowncuber]

http://youtu.be/J_738z7xYeM?t=2m30s

Can you notate this?

Not saying it's a bad idea, but I don't think this is going to catch on. It's more fun and more practical to just watch a video about it. (you say it's a hassle to record and upload videos about specific fingertricks, but I've done it several times for people. It's not all that bad.)

On a mostly unrelated note, is there any way to embed a video so that it starts at a specific point? (2:30 in this case)

 

[mark49152]

As the guy who started the earlier thread referenced by the OP, I would really like to see a fingertrick notation and I disagree with those who advocate that text or video is good enough. However, as that thread developed I quickly came to the conclusion that there are two main barriers to it happening and thus we're unlikely to see a standard notation any time soon.

First, too many opinions and ideas but no interested and trusted authority (like WCA) to arbitrate and make decisions. That will always lead to fragmentation and thus failure.

Second, no major user or resource to drive adoption. If confined to occasional use on the forums the notation would either fade out or continue to fragment.

I did ask Stachu if he would be interested to include fingertrick notations on algdb.net, but he declined. I think it would really need something like that to take off. If each alg posted on algdb.net could be accompanied by fingertrick hints in a specified notation, that would eventually become a defacto standard.

 

[Dane man]

First, too many opinions and ideas but no interested and trusted authority (like WCA) to arbitrate and make decisions. That will always lead to fragmentation and thus failure.

Second, no major user or resource to drive adoption. If confined to occasional use on the forums the notation would either fade out or continue to fragment.

I did ask Stachu if he would be interested to include fingertrick notations on algdb.net, but he declined. I think it would really need something like that to take off.

This is what I understand to be the problem. Here's the thing, for something to work, there needs to be a consensus among enough people that will use it, so that it'll get used.

It needs to be organized somewhere (I'm thinking the Speedcubing Wiki, as that is now a popular internet resource for cubers, and it won't disturb other sites). When someone uses it (at least from the beginning) there will likely need to be a link to the page where the notation is posted (put the link in your post? in your sig?). And last but not least, we need to have good threads started on the main forums, just to discuss finger-tricks, and have the notation used (I suggest a challenge thread, or a comparison thread. e.g.: "How do you finger-trick <common alg>?", "Who can make the best finger-trick for <difficult alg>?").

Yes, these are challenges, but not impossible. The lack of a notation is the reason it's difficult to get a notation out there. If we just choose a good one and start using it, that barrier will break, and the notation will begin to be used and recognized, perhaps even modified to suit the needs of the community.

http://youtu.be/J_738z7xYeM?t=2m30s
Can you notate this?

Woah, really weird, but okay...
{tB} R2(T3) U(M2) S'(I2) U2(T4) S(I1) U(M1) R2(T1)
...yeah... doable, but really weird. It requires notation on every move. And isn't that bad for your wrist?

 

[teller]

And isn't that bad for your wrist?

Dude...repetition is bad for your wrist. Novel movements like this are actually quite therapeutic.

 

[Arkwell]

~snip~

Good idea but if you go over to Noahaha's '13 Cubers one scramble' and watch all the best cubers solve you'll see each uses different fingertricks for the same alg. Watching 'Example Solves' with different cubers you see the same thing. For me asking better cubers about a specific alg I'm having trouble with worked much better. A lot of times I go to 'Request an Alg' and describe how I do it and they will always help(Thanks Teller & Anton!)and even film themselves which will beat notation any time.

 

[Dane man]

Good idea but if you go over to Noahaha's '13 Cubers one scramble' and watch all the best cubers solve you'll see each uses different fingertricks for the same alg. Watching 'Example Solves' with different cubers you see the same thing. For me asking better cubers about a specific alg I'm having trouble with worked much better. A lot of times I go to 'Request an Alg' and describe how I do it and they will always help(Thanks Teller & Anton!)and even film themselves which will beat notation any time.

The fact that all cubers use different finger-tricks is the reason for creating a notation. And as you can see from this thread, many would prefer a notation to descriptions or video because it's faster to read, simpler to produce, and able to be universally understood.

We're not saying that video or words are bad or don't work, because they do. But we are saying that there is a more efficient way for these things to be expressed for the purposes of conversation and consistency.

 

[Kirjava]

Take guysenei1's challenge.

 

[Dane man]

Well, I guess it could be but we'd have to make some additions to the grip notation. Yours uses the same grip principle as XTownCuber's, so I guess that because that is also a popular trick grip, we should modify the grip notation. Thanks for showing that.

The normal grip notation currently only expresses the normal positions of the left and right hands. Piece specific gripping would need a few more things to express that. Because DeeDubb's notation doesn't talk about the specific grip involved, it can still express the fingertrick, though it'll be done without the grip expression (This is assuming that white is F):
{Grip notation} (R2'(M2) U2(M2) L2'(M4) B2(M2))*3

So the fault is in the grip notation. And the grip notation needs to be kept from getting overly complicated, while at the same time being specific about which fingers, where, and for how long to hold on (if entire alg, or until the finger-tricks require otherwise, or until the next grip notation). Thanks for pointing that out.


Anyone have ideas on how we can improve the grip notation? (I've got to head out for a bit. I'll come back with something as well.)

 

[Kirjava]

{Grip notation} (R2'(M2) U2(M2) L2'(M4) B2(M2))*3

There's no way you'll derive what I did in the video from that string of information. Poor show.

Anyone have ideas on how we can improve the grip notation? (I've got to head out for a bit. I'll come back with something as well.)

The fault is the grip notation for now, but extend your notation for this weird exception and I'll just find something else you'll need to account for. This will keep happening, as I explained.

 

[Dane man]

There's no way you'll derive what I did in the video from that string of information. Poor show.

Interesting thing about that, if the Grip notation was expressed correctly, then there would be no need for DeeDubb's notation in the alg. It would be {Hold onto _ Pieces with _ fingers} R2' U2 L2' B2. A possibility would be {T-R1, M-R2, t-Fc, m-Bc} R2' U2 L2' B2.

The fault is the grip notation for now, but extend your notation for this weird exception and I'll just find something else you'll need to account for. This will keep happening, as I explained.

There is only so much you can do with the cube and your fingers. If you start using your tongue, then I'll stop.

Everything has it's limits, but that doesn't make it useless.

 

[mark49152]

extend your notation for this weird exception and I'll just find something else you'll need to account for. This will keep happening, as I explained.

I don't think a finger trick notation should account for every weird exception. That would add complexity for minimal extra benefit. Instead it should carry just enough info to provide hints as to how to execute "typical" algs. Videos will always be better for showing weird stuff.

 

[Dane man]

I don't think a finger trick notation should account for every weird exception. That would add complexity for minimal extra benefit. Instead it should carry just enough info to provide hints as to how to execute "typical" algs. Videos will always be better for showing weird stuff.

Agreed, hence the desire for an assumptive finger-trick notation, instead of a fully-fledged finger-trick notation. There would be too much to add if we wanted to add everything, especially for the things that... well, basically no one does. But judging by what both XTownCuber and Kirjava have presented, that is something that would be common rather than exception, and so, we make a simple to understand grip notation, and beyond that, be assumptive.

 

[Lucas Garron]

I pretty much agree that we shouldn't need to account for Kirjava's execution. I sometimes also do fun stuff that no one should ever need to notate, but for nearly all my solves and algs I use a standard set of fingertricks.
I still think a decent set of tricks should cover most of what every Fridrich solver uses, and possibly work well other methods, too.

I suppose I could be convinced otherwise if someone shows me lots of common algs that many cubers execute weirdly. Unless there are a lot I'm not thinking of, a decent fingertrick notation would still be quite useful.

 

[goodatthis]

I guess that this is pretty much a "My PLLs" video in text form.

 

[Kirjava]

I pretty much agree that we shouldn't need to account for Kirjava's execution. I sometimes also do fun stuff that no one should ever need to notate, but for nearly all my solves and algs I use a standard set of fingertricks

You seem to be unaware of the popular A perm that was executed with a similar fingertrick to that video. This kind of stuff is going to crop up.

 

[Lucas Garron]

You seem to be unaware of the popular A perm that was executed with a similar fingertrick to that video. This kind of stuff is going to crop up.

Doesn't immediately come to mind, but I may have seen it. So that's at least one.
(Got a link to a video of someone using it in a speedsolve?)

 

[blade740]

I think a notation that covers MOST (certainly not all) fingertricks is possible, but the more complete you make it, the less precise it will be.

First of all, define a standardized grip system: I think defining which face each thumb is resting (or which sticker, if you want to be extra precise) is sufficient. The other fingers freely float, though you could define which fingers hold the other side of the cube for completeness.

I have proposed a system of ordered pairs in which the face each thumb rests on is represented by a number: 0 = D, 1 = F, 2 = U, etc. The reason for using numbers rather than simply the letters is because you can have grips like [1, 3] and [1,-1], in which the right thumb is on different faces, but the wrist is rotated 360 degrees.



Next you define the X axis moves: R, L, l, r, and x. These moves are all done with the wrists (unless otherwise stated, but we'll get to that later). Wristing these turns results in a change to the grip. For example, starting in [1, 0] grip and performing a wrist R turn, you end up in [1, 1] grip.

An X on its own stands for a cube rotation with both wrists (e.g., going from a [0, 0] to a [1, 1] grip). To define grip changes, you use the same ordered pair system: a [0, +1] grip shift moves your right wrist clockwise one place (similar to doing an R move). Y and Z rotations get much more difficult to define concisely, so for now I'm going to just leave them alone. In general, a rotation along the same axis as the face where the thumb is is done with a "slide", and one that goes perpendicular to the thumb (for example, a y rotation from [1, 1] grip) needs a regrip of that hand. This certainly doesn't cover all cases but it's close enough for now.

Any moves that are not performed with your wrist are described with the letter of the move, then (within brackets to signify a fingertrick), a letter signifying the finger used (T-I-M-R-P for thumb index middle ring pinky, lower case t-i-m-r-p for left hand), then (optionally) a sticker where the finger begins the move, and (again, optionally), either "pull" or "push".

Thus, starting from a 1,1 grip, the move written:
U[I, BRU, pull] is the simple pull of the right index finger to perform a U. However, if you're performing the move from the sticker opposite the thumb of that hand, you can omit the starting sticker. By "sticker opposite the thumb", I mean the usual resting place for an index finger: BRU for U from [1, 1] grip, UFR for F from [1, 0] grip, etc. This applies for similar moves with the ring/pinky fingers: BRD for D from [1, 1] grip, as well as things like E and S slice turns, I suppose. Similarly, assume most moves are "pull" moves, unless specified otherwise. I think allowing for "optional" notation keeps things more concise for the majority of written algs.

Note that in my system, a "U2" generally signifies a single move. For the "double trigger" fingertrick, you'd generally write U U[M]

As an example, here are two different ways to perform a T-perm:

[1, 1] R U[I, BRU] R' U' R' F R2 U' R' U' R U[I, BRU] R' F'[T, DFR]

[1, 0] R U R' [0, +1] U' R' F R2 U' R' U' [0, -1] R U R' F'[T, DFR]

Anyway, I hope this makes as much sense to you all as it does to me. This system allows you to flexibly describe most fingertricks in use. Granted, kirjava's example doesn't work (although I have a few algs in this modified "column" grip and I think it could be separately defined), and I'm sure there are a million other exceptions, but it mostly works. Let me know if anything needs more explaining.