2-gen contest

For all the people that think 2-gen solving is fun, but easy, here is a nice variation/challenge:

Scramble: Just like the rest of this contest, you can only scramble with RU moves, BUT during that scrambling you have to do 1 L' move and also 1 L move (in that order).
Example: RUR2U2 L' R'UR2 L U2R2UR'
Result: You might or might not be able to solve this with only RU moves, but you can always solve it with only RU moves, 1 L' move and 1 L move
Beginners challenge: Solve with only RUL moves
Regular challenge: Solve with only RU moves and a maximum of 1 L' and 1 L move
Additional challenge: Find a way to determine of L moves are actually necessary for solving the cube.

Now if anyone wants to try this for speed, just insert the L' move after the 5th turn and the L move after the 11th turn in the provided scrambles.

P.S. I realize this is not just "keeping it to discussing times , algs, etc."
 
But how would that algo help in 2-gen BLD solving? Well it is 2-gen but neither of the 2 sides being turned are the U face if you're preserving U/D sticker orientation.
Click to expand...
1. Memorize orientations differently using front/back colors, or
2. Do <F,R> instead.
 
AvGalen said:
P.S. I realize this is not just "keeping it to discussing times , algs, etc."
Click to expand...
That's fine. :p
I'll try to figure out what I can learn from the J-perm.

StefanPochmann said:
I'm simply talking about orientation keeping the FB stickers on the FB sides. Instead of your UD version.
Click to expand...
That's what I meant with F2B2, sorry for not correcting that. (Should've said R2U2?)
Anyhow, could you give an example solve with sane setups?

ch_ts said:
1. Memorize orientations differently using front/back colors, or
2. Do <F,R> instead.
Click to expand...
We're presuming <U,R> for the easy alg.

EDIT: I have a general solution to Arnaud's challenge. CP is the only issue. Essential solution: Solve as many corners as possible, you'll need either a 2-cycle or a 3-cycle to place DFL and finish CP (CO and EP are trivial while the L is up - Sune and Allan (make sure L' doesn't hide FL).
3-cycle: L' U R2' U' R U L (hand-found, based on an obvious comm idea)
2-cycle: (U2) L' U' R' U' L (found by getting a gazillion algs in CE, then realizing I could severely chop this one down)

I'll do an average tomorrow.

(But: First AvG-2.5-gen solve: 32.60+ [with CLS :)])
 
Last edited:
FU said:
But how would that algo help in 2-gen BLD solving? Well it is 2-gen but neither of the 2 sides being turned are the U face if you're preserving U/D sticker orientation.
Click to expand...
Oh gosh... pure ignorance dripping here. I so need to write my tutorial about this stuff.

Lucas Garron said:
That's what I meant with F2B2, sorry for not correcting that. (Should've said R2U2?)
Click to expand...
I suspected. And yes, should've been R2U2.

Lucas Garron said:
Anyhow, could you give an example solve with sane setups?
Click to expand...
Will do some tomorrow.
 
Last edited:
Lucas Garron said:
AvGalen said:
P.S. I realize this is not just "keeping it to discussing times , algs, etc."
Click to expand...
That's fine. :p
I'll try to figure out what I can learn from the J-perm.

EDIT: I have a general solution to Arnaud's challenge. CP is the only issue. Essential solution: Solve as many corners as possible, you'll need either a 2-cycle or a 3-cycle to place DFL and finish CP (CO and EP are trivial while the L is up - Sune and Allan (make sure L' doesn't hide FL).
3-cycle: L' U R2' U' R U L (hand-found, based on an obvious comm idea)
2-cycle: (U2) L' U' R' U' L (found by getting a gazillion algs in CE, then realizing I could severely chop this one down)

I'll do an average tomorrow.

(But: First AvG-2.5-gen solve: 32.60+ [with CLS :)])
Click to expand...
You have the same approach as I do Lucas. The problem is in the CP, so you solve that first and continue with regular 2-gen. (and I also used RUL-J-Perm to figure it out)But it is much more fun to try it without doing the corners first. That is why I gave the challenge to determine whether or not you would actually need the L/L' moves.

I think challenges like this is why there is such a strong correlation between mathematicians, programmers, cubers. We handle big and impossible problems by dividing them into smaller problems which are (preferrably) independent of each other.
 
AvGalen said:
You have the same approach as I do Lucas. The problem is in the CP, so you solve that first and continue with regular 2-gen. (and I also used RUL-J-Perm to figure it out)But it is much more fun to try it without doing the corners first.
Click to expand...
Actually, I didn't end up using the J-perm. But I forgot to mention that if DFL gets replaced in the scramble, you can to use a Niklas (CP of the J-perm: L'URU'L) if you have 2 corners to swap.


AvGalen said:
That is why I gave the challenge to determine whether or not you would actually need the L/L' moves.
Click to expand...
Aha. So that "of" was "if." :p
If you know about CP in 2-gen, and remembered that the J/Niklas uses only L' and L once, I thought that this was obvious.

I thought the regular challenge was just to solve it once (or twice), and the additional challenge was to know what you're doing. :p

AvGalen said:
I think challenges like this is why there is such a strong correlation between mathematicians, programmers, cubers. We handle big and impossible problems by dividing them into smaller problems which are (preferably) independent of each other.
Click to expand...
Isn't that life?
Anyhow, here we're being analytic.
I used to dislike the scientific method, mostly because I didn't want to accept it as the logical way to deduce just because they teach it to us. Now, I like it, and even though it doesn't work quite as well in math, it's implicit in problem-solving like this.
 
20.91, 28.12, 20.77, (20.67), (DNF), 27.28, 23.62, 48.17, 22.64, 25.34, 26.17, 24.74 > 26.78 avg

1. R2 U R' U R2 U' L' R2 U2 R U2 R2 U' L R' U' R2 U' R2 U
2. R U2 R U2 R U L' U' R' U' R' U2 R' L R' U' R' U R' U
3. U' R2 U' R2 U R L' U2 R2 U R' U2 R L R' U' R U2 R2 U'
4. R U' R2 U2 R2 U' L' U2 R' U2 R U' R' L U R' U2 R2 U2 R
5. R2 U2 R2 U R U2 L' U R' U' R' U' R' L U R U' R2 U2 R
6. R' U' R' U' R U' L' R' U R U2 R U2 L U R' U' R' U' R
7. R U R' U2 R2 U' L' R2 U R' U R' U2 L U' R U R U R
8. U' R U' R2 U' R L' R' U R2 U' R2 U' L R' U R U' R U'
9. U2 R U' R2 U2 R' L' U R U2 R' U R L U' R U R2 U' R2
10. U2 R2 U' R' U R2 L' U' R U' R2 U R L U R' U' R' U R2
11. U2 R2 U2 R' U' R L' R2 U R' U2 R2 U' L U' R' U' R2 U' R
12. U R' U R U2 R' L' U2 R2 U' R2 U' R' L R U R U R2 U

I was being really careful because my first avg had 2 DNFs and I got another DNF as I was rolling out the first of the 2. :(

And I'm still not happy with the randomness of those scrambles.
 
Your first few scrambles and ways to solve the corners with FB orientation. Not necessarily optimal, and with more algs one can of course do more. Here I only considered using these:

double swap: (R U' R' U)*3
double swap: (R' U R U')*3
"X-perm": R2 U2 R U2 R2 U2 R2 U2 R U2 R2 U2
"U-turn": U R U2 R' U' R U R' U' R U' R'
"U'-turn": R U R' U R U' R' U R U2 R' U'

Scramble 1:
[scramble] R2 U R2 U2 R U' R U' R' U' R' U2 R' U' R U'
[orient corners] R2 (R' U' R U' R' U R U' R2 U' R' U' R' U R U R U' R) R2
[solve BDR] (R' U R U')*3
[solve FRD] U2 (R U' R' U)*3 U2
[solve rest] R2 U2 R U2 R2 U2 R2 U2 R U2 R2 U2

Scramble 2:
[scramble] R' U R' U2 R2 U2 R' U' R2 U R2 U2 R' U2 R U'
[orient corners] R2 U2 R' U' R U' R2 U' R U' R2 U' R U' R' U2 R
[solve FDR] R2 (R U' R' U)*3 R2
[solve BRD] (R' U R U')*3
[solve rest] R U R' U R U' R' U R U2 R' U'

Scramble 3:
[scramble] R' U2 R' U2 R U R U2 R U2 R' U' R U2 R' U'
[orient corners] R (R' U R' U' R' U' R U R U R2 U R' U' R U R' U R) R'
... Darn this must be like the worst case. I give up for now.

ACube found an alright algorithm which 3-cycles the front corners as well as the back corners, exactly what I would've needed in the first solve. So that would be:

Scramble 1:
[scramble] R2 U R2 U2 R U' R U' R' U' R' U2 R' U' R U'
[orient corners] R2 (R' U' R U' R' U R U' R2 U' R' U' R' U R U R U' R) R2
[solve all corners(R' U2 R2 U2 R2 U) (R U')*3 (R U)*3

There are only 120 possible corner permutations. I certainly wouldn't want to learn 120 algorithms for one-step CP without setup moves, but I estimate 15 algs to be enough for a two-step CP without setup moves, and that would be reasonable to learn.
 
While we're at devising a complete system:

I think UL/UF make nice buffers for edges:
U-turn: R U2 R' U' R U R' U' R U' R' U (trigger last move with ring/pinky)
U'-turn: U' R U R' U R U' R' U R U2 R'

(How about "Pseudo U-turn"? PsU ("syoo") and PsU' ("syoo prime"))

I suggest EF, so that the cycle can always start at UF. Otherwise, a PsU / PsU' involves + (see below).

Buffer: UF/UL
Memo: Normal, like Old Pochmann or M2, starting from UF (main buffer).

Let's call the U-perms C (clockwise: R2 U R U R' U' R' U' R' U R') and W (ccw inverse).
Just to represent how blind this can go, initialize M=W, and let + mean (switch the value of M between C and W)

Execution:
R-slice edges: Set up to UR with ARF (:p), M +, undo setup.
UB: U' M + U
UL: +


EDIT: I just realized that those PsU turns don't work. So it's either something like UR'U2RUR'U'RUR'URU2 (and inverse), or Stefan's alg with an extra edge to shoot in case of parity.

And yes, I know this is slow/inefficient. You can easily resolve 3-cycles with easy setups. But it's so simple.
 
Last edited:
StefanPochmann said:
There are only five targets, no need for setup moves. Here are two nice algs:

U' R' U' R U R U R U' R'
R' U' R' U' R U R U R U'
Click to expand...
I know. I actually setup to DR more often (not UR), and I can easily do 3OP-style. But I wanted to get down a full 2-cycle method.

And sorry about those algs, I edited in a note, but submitted the edit way too late because the tab got forgotten...
 
12: 00:10.45 x
11: 00:12.42 x
10: 00:14.35 x
9: 00:07.06 x
8: 00:13.45 x
7: 00:15.48 x
6: 00:12.71 x
5: 00:11.00 x
4: 00:11.50 x
3: 00:19.81 x
2: 00:13.95 x
1: 00:11.32 x
avg:12.67

not that good
 
(25.02), 20.63, 17.84, 21.59, 17.75, 22.40, 19.05, 19.66, 16.42, (8.56), 18.64, 17.99

I know my OH sucks, so this is good practice I guess?

PLL skip on 8.56 (and I beat a couple of guys who used 2 hands! :D), others non-lucky.

Edit: My 2H results: 7.45, 10.62, 8.14, 8.79, 12.30, 12.06, 8.60, 11.15, (12.31), 7.04, 12.14, (6.02). 3 PLL skips this time for some strange reason :D
 
Last edited:
2H average : 7.44 :)
8.55, (5.32), 8.25, 7.15, (8.92), 7.45, 7.12, 7.11, 8.64, 7.10, 7.21, 5.81

I've been messing around with a siamese 3x3x3 all day so I guess that helped me get used to 2-gen :p
 
Last edited:
Average: 16.04 seconds
Individual Times: 12.16, 19.83, (21.58), 17.50, 18.55, 11.55, 13.53, 17.17, 13.72, 19.14, 17.30, (10.20)

Not very good. I was thinking more like a 13s average, but a couple of the OLLs and the z perm threw me off :/
 
You must log in or register to reply here.

Similar threads

C
Calling all Roux users (and other...) <M.U> 2-Gen Contest
Replies
12
Views
2,417
Athefre
Athefre
unirox13
Looking to collect the decorative sticker from the Next Gen Rubik's 3x3 boxes
Replies
0
Views
161
unirox13
unirox13
Ordway Persyn
How best to clean a speedstacks mat? (Gen 5)
Replies
3
Views
255
1x1 wr holder!
1
iguana
What is God's number for a 3-gen scramble?
Replies
8
Views
611
Bruce MacKenzie
Bruce MacKenzie
t0xiC
4-gen or 6-gen on megaminx ZZ?
Replies
7
Views
443
Fred
Fred
Back
Top