What does this algorithm actually do?

[MagicVince]

Hello,

If you are wondering what this or that OLL actually does (in terms of permutations for instance), this website can help you:

Formula Analyzer

Rubik's cube Formula Analyzer with graphic output of the permutations and orientations effects on corner edge and center cubelets and an algorithm database.

cube.vgreco.fr

When I introduced myself two and half years ago I told you I was working on a website that would analyze the effects of an algorithm. For some unknown reasons, I did not post the URL of my website when it was in alpha stage, some months later (lucky you, you missed the ugly earlier versions). But I recently updated it and I consider that it is now in beta stage.
I am taking this opportunity to show it to you.

I could speak for hours about this website, but I will try to be brief. Basically you type (or copy and paste) any formula into the input section, say an OLL. An image is generated that shows the effect of the algorithm (more exactly it shows what situation it can solve) with colored arrows. In this way you can see the desired effects (on orientation) but also the side effects (on permutations). There is a summary window that gives you some extra informations (list of the permutations/orientations, how many times you should repeat the algorithm to cancel it, length in QTM etc.). Then, you can search the mini-database (less than 150 algorithms, but it has all the OLL and all the PLL) for a similar algorithm that would leave the corners unchanged, for instance. By clicking on its formula, you can now explore it and so on.
Does such a website already exist? Do you find it useful?
Any feedback is welcome.
Thanks,

MagicVince.

 

[Deleted member 36445]

this is amazing, thank you for your hard work!

Spoiler

 

[GodCubing]

There is a CMLL visualizer that does similiar things for CMLLs instead of all algs

 

[MagicVince]

To solve the cube, just follow the arrows' directions...

I am joking, but I confess I already tried to get closer to the solution only by looking at the permutation:

Both the image and the permutations show the complexity to reach the solution. If one could mesure this distance to the solution through this complexity, then God algorithm would be within reach...

There is a CMLL visualizer that does similiar things for CMLLs instead of all algs

Interesting. Thank you for the information, I did not know this website.
For U2 R U2 R' U2 R' F R F', it gives this image.

Mine will output by default this one:

But you can disable arrows to show only edge permutations and/or flips, or at the contrary corner permutations and/or twists.

Of course if the edges go to the bottom layer, my result will be rather messy (it was not designed with Roux in mind).

Hello again

I am working on two small improvements:

1. In the input field, you can now select the part of the formula you want to be analyzed. I am not totally satisfied with the behavior (the selection disappear if you click outside the input field although still active, if you click again inside the input field, it still does not appear but is still active if you click inside the former selection whereas it is OK if you click out of the former selection)
2. You can now choose your color scheme (Western or Japanese) and your cube orientation. This one is working properly. I don't know if need to have a custom color scheme or if these two presets are enough.

For instance, if you look at this Y-Perm (FRU'R'U'RUR'F')(RUR'U')(R'FRF') you can select the first couple of parentheses, click on "Find Similar" and see an OLL 37, or the two last couples of parentheses, click on "Find Similar" and get an OLL 33.
These modifications are now officially published here: https://cube.vgreco.fr/
What do you think ?

 

[Deleted member 36445]

hi

firstly, good job!

secondly, can you make the visualization more standard?

for instance, this is yperm on your system

and this is the standard visualization

as you can see, swaps are symbolized by a single arrow, not two
it's just a detail and I can agree if you decide to leave as it is.

 

[MagicVince]

Thanks Felipe.
Actually this visualization is closely linked to the permutation/orientation notation that appears in the Effect window.
This allows me to easily do this:

and go from this output to the previous one (if I decide to remove the orientations for instance). The single arrows would not be compatible with the edge flips and corner twists.
More precisely, the effect is noted like this: (urf, ulb-)(ufl+)(ul, ub+)(uf+)(u+) meaning the URF corner takes the place of the ULB corner, the ULB corner is twisted anti-clockwise and then replaces the URF corner etc. The arrows are following the same logic and thus it is important for me that the curved blue arrow for the twist is linked to the longer arrow for the swap.
I hope this makes sense.
PS: you can right-click on an image of my website and copy it to paste it here (or elsewhere).

 

[NONOGamer12]

bruh i just realized y perm has been lying to us

@MagicVince @Abram Grimsley literally there is an f and f' at start and end when i was looking up how to solve blind it gave me the same alg exept without those f moves

 

[Abram Grimsley]

@MagicVince @Abram Grimsley literally there is an f and f' at start and end when i was looking up how to solve blind it gave me the same alg exept without those f moves

I know. But your post about y perm lying to us is still funny to me. Because, it's true. The "y" perm has been lying to us

 

[NONOGamer12]

I know. But your post about y perm lying to us is still funny to me. Because, it's true. The "y" perm has been lying to us

oh lol

 

[Abram Grimsley]

Hey, we should see if we can find a better alg for y perm. So that we don't have to worry about "f" or "f'" moves

 

[cuberswoop]

Hey, we should see if we can find a better alg for y perm. So that we don't have to worry about "f" or "f'" moves

R2' D' R2 U R2' U' R2 D R2' U' R2 U R2' U R2
I don't know how good it is because my hands are too sweaty from geoguessr but it looks good.

 

[Abram Grimsley]

R2' D' R2 U R2' U' R2 D R2' U' R2 U R2' U R2
I don't know how good it is because my hands are too sweaty from geoguessr but it looks good.

This is great. Though, there are way to many R2 moves, but other than that, it looks like it could be a good one

It's almost like a reverse v perm. Kinda

 

[ruffleduck]

Hey, we should see if we can find a better alg for y perm. So that we don't have to worry about "f" or "f'" moves

F moves are necessary for good y perms, I'm afraid.

 

[Abram Grimsley]

Not necessarily. There has to be a way to get a good y perm without f moves

 

[ruffleduck]

Not necessarily. There has to be a way to get a good y perm without f moves

I assure you--There are none. Try generating the algs. The best you're going to get are long RUL abominations or sq1-like algs that have too many half turns.

 

[Abram Grimsley]

Noooooooooooooooooooooooooooooo! Lol
Your probably right. We'll, it's the thought that counts.

 

[MagicVince]

What would be a good algorithm for you?
For sure, there is no 2-gen Y-Perm (unless you allow cube rotations).
The only thing you can do is perhaps minimize the use of the 3rd face.
What do you think of R2 U' R2 U' R2 U R' F' R U R2 U' R' F R?
Only 2 Fs compared to the initial algorithm I mentioned and less R2s than the one from cuberswoop...

 

[bulkocuber]

I don't understand what's wrong with F moves.

 

[NONOGamer12]

Hey, we should see if we can find a better alg for y perm. So that we don't have to worry about "f" or "f'" moves

lol yea

F moves are necessary for good y perms, I'm afraid.

No Never!!!

I don't understand what's wrong with F moves.

cuz it is like the lied 2 us cuz the f moves make a totally different alg

 

[MagicVince]

I made some changes so that you can now have different kinds of images with a lot of parameters to fit your needs.
So now, for F(RUR'U')F'(U) you can get those images :
default:

Straight arrows (for permutations only):

No space between arrows:

No cubbies shown on border if not yellow:

You can also change sizes, thicknesses, colors, spacing etc.
I also fixed some minor issues and changed a bit the overall look and feel of the website.

@Filipe Teixeira, now Y-Perm can look like this (unset "curved arrows", set "hide grey on borders", set "spacing of arrows" to "none"):

which is more standard, I agree...

Is there anything else you would like to control in the display of the result?