irontwig
Veteran
30 moves is like a million times easier than 20 moves. And 30 moves is not easy without backtracking and plenty of time.
The 20MM (20 Move Method)
- Thread starter Calamity Strike
- Start date
shadowslice e
Legend
This. Just this.
If you want to be able to solve within 20 moves consistently, you're going to have to go a lot more abstract than blockbuilding/NISS/insertions.
/side note
Atm, one way that seems promising to be is to identify stationary points and group cubes in that manner (with symmetries in the same group obviously) then have a look at the inverses of the scrambles and generators to see what patterns there are in them. This would be good for permutation but I'm not sure how orientation would work at this point. I'm currently seeing what happens to this on a 2x2 permutation only solves and if I find a way then I will likely try to extend the basic principles to other cubes.
guysensei1
Legend
In light of recent 'hey look I found cool new method' threads, SHOW US WHAT YOU'VE DONE. PLEASE.
If you simply put up some claims without any actual material to show for it, we will be sceptical by default, if you actually made something up, show us. Perform a walkthouh solve with your method maybe.
If you simply put up some claims without any actual material to show for it, we will be sceptical by default, if you actually made something up, show us. Perform a walkthouh solve with your method maybe.
Ollie
Legend
What have you observed? Why does what you've observed suggest that there is a method that averages 20 moves or less?
I didn't say that, I'm just asking you to clarify what you've found. If it's new, interesting and/or good enough, then myself and others will help you. We don't want to waste our time, either.
Goosly
Veteran
Cool. Make a video and show us please.
Attila
Veteran
Here is, how to do that:
http://www.keptelenseg.hu/keptelenseg/szuper-gyors-rubik-kocka-kirakas-78867
http://www.keptelenseg.hu/keptelenseg/szuper-gyors-rubik-kocka-kirakas-78867
Calamity Strike
Regular
- Joined
- Oct 25, 2015
- Messages
- 61
Ok, what i have found is that, when a computer is solving the cube, it seems to orient certain pieces, then build a block of some sort, almost like ZZ, then, using the oriented pieces, it solves the rest of the cube in 7-10 moves.
Ollie
Legend
A bit of a vague description. But it sounds somewhat like the two-phase algorithm: http://www.cs.brandeis.edu/~storer/JimPuzzles/RUBIK/Rubik3x3x3/READING/KociembaPage.pdf
1. Orient to <U,D,R2,L2,F2,B2>.
2. Solve.
For example:
D' R U F' D2 L F' D2 L U' F B2 U' R' F' B D' U2 R' F R2 D' F' can be solved:
U F2 D2 F U2 B' // orient edges, preserve some blocks (6)
D2 L' B2 D' L' F2 D' F2 D' F2 U2 L2 U B2 D2 R2 // finish (14)
My knowledge of how it works ends there, it's witchcraft to me.
1. Orient to <U,D,R2,L2,F2,B2>.
2. Solve.
For example:
D' R U F' D2 L F' D2 L U' F B2 U' R' F' B D' U2 R' F R2 D' F' can be solved:
U F2 D2 F U2 B' // orient edges, preserve some blocks (6)
D2 L' B2 D' L' F2 D' F2 D' F2 U2 L2 U B2 D2 R2 // finish (14)
My knowledge of how it works ends there, it's witchcraft to me.
Lucas Garron
Premium Member
Good observation!
I highly recommend that you read up on Thistlethwaite (and Human Thistlethwaite), Kociemba (also see my explanation), and Jaap's computer cubing page to see if you've thought of anything that isn't already well understood.
Chances are, there is no easy way for a human to find a short combination of orientation and permutation phases without spending a lot of time searching like a computer (or using a ton of lookup tables).
Berd
Legend
I hope you can pull off what you've described!
Calamity Strike
Regular
- Joined
- Oct 25, 2015
- Messages
- 61
Ok, so I'm not completely crazy. This is actually part of the method.
Was this sarcastic or real?
Last edited by a moderator:
Chree
Veteran
- Joined
- Jun 7, 2013
- Messages
- 1,226
- Location
- Portland, OR, USA
- WCA ID
- 2013BROT01
- YouTube
- Visit Channel
Lucas is the realest.
Ollie
Legend
Can't say that for certain!
But seriously, yeah it is.
If you're interested, I left Cube Explorer running and the first optimal it found for that scramble is 18 moves:
U F' B2 L D' L' U2 R' F2 R' U2 L2 // make lots of pairs and blocks , but not quite at the EO stage yet (12)
F // orients edges (1)
D R B2 U' R' // witchcraft (5)
To me, this makes even less sense than the previous 20 move solution. The first solution is somewhat like ZZ, but the second solution orients all the edges towards end. At this point, it can solve the rest of the cube in 5 moves.
If you read either Lucas's or Herbert's explanations, this will make a bit more sense. But from a speedcubing point of view, if there is an 'easy' explanation to why the optimal and the 20 move solution are so different, then that would be crazy.
Edit: It's based off the inverse solution. Fun! Now go and read FMC stuff.
Edit2: Might be a NISS thing? The ending of the scramble is very similar to the optimal solution.
Last edited:
irontwig
Veteran
Does anybody know how much searching (on average) CE needs to find a 20 move solution? Should be some orders of magnitude too many for humans.
mjm
Regular
What's the Skype group?
Calamity Strike
Regular
- Joined
- Oct 25, 2015
- Messages
- 61
Its just called cubers, you can add me on skype by clicking on the skype icon under my name, and ill put you in the group.
DGCubes
Veteran
Very interesting. I always like to keep an open mind, and if something does exist, it could really bring cubing to a new level. The hardest part (for the speedsolving aspect) would be getting inspection sub-15, but I won't say it can't happen.
Petro Leum
Veteran
has anyone tried looking at optimal solutions by computer programs and analyzing all the edge and corner cycles and looking if there are patterns of similar cycles/similar moves in the solution that repeat themselves?
i know i know, i'm just throwing some weird thought out of my twisted mind.
i know i know, i'm just throwing some weird thought out of my twisted mind.
IRNjuggle28
Veteran
I often wish I was more like you, Calamity Strike. I'm the type to be logical to a fault, often at the expense of emotions and optimism, and to not chase whims or uncertainties. If I were in your shoes, I'm sure I'd assume that someone would've already found evidence pointing towards the possibility of a 20 move method if one was possible to find, or even accuse myself of being arrogant for thinking that, as a newer cuber, there was a chance of me understanding something about the cube that more experienced cubers hadn't. Assuming that I didn't immediately dismiss the idea, I'd still be tentative to post a thread about it, fearing being wrong in a way that embarrassed me, or being ridiculed. Like many others on here, I'm skeptical that such a method exists, but I am completely sincere in saying that the fact that you're hopeful enough to think there's a solution and brave enough to post a thread about it inspires me.
One question I have about the possibility of this method is why you think it would take 20 moves? It doesn't make sense that a 20 move scramble would be solvable in 20 moves, but a 15 move scramble would also be solvable in 20 moves and not necessarily less. I think that if there's a method that leads to consistent <20 move solutions that's achievable by a human, it's most likely a method that leads to consistent optimal solutions. I would name your concept "optimal solution method" rather than 20MM. Your choice, though. Good luck with your search and with cubing in general.
One question I have about the possibility of this method is why you think it would take 20 moves? It doesn't make sense that a 20 move scramble would be solvable in 20 moves, but a 15 move scramble would also be solvable in 20 moves and not necessarily less. I think that if there's a method that leads to consistent <20 move solutions that's achievable by a human, it's most likely a method that leads to consistent optimal solutions. I would name your concept "optimal solution method" rather than 20MM. Your choice, though. Good luck with your search and with cubing in general.
shadowslice e
Legend
Well, that's essentially what in trying to do with permutation for corners using stationary points at the moment.
Basically, I'm grouping scrambles according to stationary points and some other factors (like what cycle the other corners are in) and trying to find patterns in the generators and matrices but I can't spend much time doing it atm cause of college.
Similar threads
