Probability Thread

If you knew that the first layer permutation would be solved then you can't really call it a PBL skip. In which case, it's just a LL skip which is 1/162

But I think what he's asking is what are the chances of both layers being fully oriented and permuted after you make a face.

i keep getting PLL skips on 7x7... while none on smaller size

I might be wrong but I think this is just chance.
 
Last edited:
So, if you really just build one side and totally ignore the permutation of the bottom layer, then of course the chance of a skip after that layer is 1/6*1/162 = 1/972 which is actually not that rare.
 
If I force edges to be solved at the end of F2L so that I have a maximum of 4 corners left, what the probability that:
a) At least one of them will be solved
b) It's a 3-cycle left (3 twisted is not allowed)

I came out with the answers 31/108 and 31/162 respectively (~0.287 and ~0.191) but I'm really not sure whether I made an error somewhere so I'd appreciate it if someone else had a go.
 
If I force edges to be solved at the end of F2L so that I have a maximum of 4 corners left, what the probability that:
a) At least one of them will be solved
b) It's a 3-cycle left (3 twisted is not allowed)

I came out with the answers 31/108 and 31/162 respectively (~0.287 and ~0.191) but I'm really not sure whether I made an error somewhere so I'd appreciate it if someone else had a go.

I got the same thing for part a)

For part b) I got 2/9

4*(2*3^2) = 72

1) 4 choices for the solved (permuted and oriented) corner.
2) 2 possible derangement permutations of 3 remaining corners
3) 3^2 possible orientations of the 3 unsolved corners

4*3*3^3 = 324 total possible positions for F2L and LL cross solved.

72/324 = 2/9
 
What is the probability of getting a random scramble of a 7x7 with one center fully solved? 4x4?
I think all you need to do to calculate this for the nxnxn is just take out the positions that one composite center contributes to the total number of positions on the nxnxn. To see how I got the following piece (in the denominator of the fraction below) of the formula for the number of positions on the regular nxnxn, see this post.

\( \frac{20!^{\left\lfloor \left( \frac{n-2}{2} \right)^{2} \right\rfloor }\left( \frac{1}{4!^{5}} \right)^{\left\lfloor \left( \frac{n-2}{2} \right)^{2} \right\rfloor }}{24!^{\left\lfloor \left( \frac{n-2}{2} \right)^{2} \right\rfloor }\left( \frac{1}{4!^{6}} \right)^{\left\lfloor \left( \frac{n-2}{2} \right)^{2} \right\rfloor }} \)

There are \( \left\lfloor \left( \frac{n-2}{2} \right)^{2} \right\rfloor \) orbits (sets) of 24 non-fixed center pieces. 4 are in each face. Therefore we change 24! to 20! (24-4=20). For the other factor, we change 4!^6 to 4!^5 because we are pretending to have 5 composite centers instead of 6.

\( =\left( \frac{\frac{20!}{4!^{5}}}{\frac{24!}{4!^{6}}} \right)^{\left\lfloor \left( \frac{n-2}{2} \right)^{2} \right\rfloor }=\left( \frac{20!}{4!^{5}}\frac{4!^{6}}{24!} \right)^{\left\lfloor \left( \frac{n-2}{2} \right)^{2} \right\rfloor } \)\( =\left( \frac{20!}{24\left( 23 \right)\left( 22 \right)\left( 21 \right)20!}\frac{4!^{6}}{4!^{5}} \right)^{\left\lfloor \left( \frac{n-2}{2} \right)^{2} \right\rfloor }=\left( \frac{1}{\left( 23 \right)\left( 22 \right)\left( 21 \right)} \right)^{\left\lfloor \left( \frac{n-2}{2} \right)^{2} \right\rfloor }=\frac{1}{10626^{\left\lfloor \left( \frac{n-2}{2} \right)^{2} \right\rfloor }} \)

All you have to do is substitute a value for n, corresponding to the cube size you want. See this link for the probability of a random composite center for cube sizes 2 through 11 (you can change the value of n for larger cube sizes).
 
True story,

Today I got 3 E-Perms in a row. I thought "what are the chances of that?" Later, I got two PLL skips in a row. Can anyone tell me what were the chances of getting both of these things happening on the same day. Considering there are 365 days in a year.
 
True story,

Today I got 3 E-Perms in a row. I thought "what are the chances of that?" Later, I got two PLL skips in a row. Can anyone tell me what were the chances of getting both of these things happening on the same day. Considering there are 365 days in a year.
How many days there are in a year doesn't matter. However, How many solves you did today does.
 
(chances of Eperm)^3 x (chances of pll skip)^2 x 100
i think its 1/24186470400
not entirely though
You have to consider that the 3 consecutive E perms and 2 consecutive PLL skips can start and end anywhere in the 100 solves. Also it's possible to have more than 3 or 2 in a row (implied from the question, unless he meant exactly 3 and exactly 2).

True story,

Today I got 3 E-Perms in a row. I thought "what are the chances of that?" Later, I got two PLL skips in a row. Can anyone tell me what were the chances of getting both of these things happening on the same day. Considering there are 365 days in a year.
For 100 solves in a day:

P (>2 E perms in a row in 100 solves)=
\( 1- (35/36)^{100} \)\( - 100(99/\binom {100}{2})(1/36)(35/36)^{99} \)\( - \binom {100}{2} (98/\binom {100}{3}) (1/36)^2(35/36)^{98} \)
P (>1 PLL skip in a row in 100 solves)=
\( 1- (71/72)^{100} - 100(71/\binom {100}{2})(1/72)(71/72)^{99} \)

The probability of getting at least 3 E perms in a row and 2 PLL skips in a row on the same day will be close to the sum of the two above probabilities. Since you can't have an E perm and a PLL skip in the same solve, there will be a bit of error (anyone know how to get it without error?).
 
Last edited:
Am I correct that the probability of getting a cll skip if using cll ell for last layer is 1/162?
162 = 4!*(3^3)/4
4! permutations (can be odd permutation for corners because edges are unsolved)
3^3 orientations (3 orientations per corner except for the 4th whose orientation is dependent on the other 3)
/4 for AUF
 
If you did just one average then...

Chances of PLL skip ==> 1/72
chances getting X 3/5 times ==> 5 C 3 = 10

so 3 PLL skips in of one avg 5 neglecting order = (1/72) / 10 = 1/720
 
I did an AO5 recently, and I got 3 PLL skips. What are the chances of that?

P (exactly 3 PLL skips in 5 solves)
= \( \binom {5}{3} (1/72)^3(71/72)^2 \)
= \( 25205/967458816 \)

About 0.00002605%, assuming all the PLL skips were 1/72 (no CxLL, OLLCP or other ways of forcing easier PLL cases used during OLL).

edit: If you meant >2 PLL skips in 5 solves it would be
= \( \binom {5}{3} (1/72)^3(71/72)^2 + \) \( \binom {5}{4} (1/72)^4(71/72)^1 + (1/72)^5 \)
= \( 25205/967458816 + 355/1934917632 + 1/1934917632 \)
= \( 8461/322486272 \)

Which is about 0.00002624%
 
Last edited:
If you did just one average then...

Chances of PLL skip ==> 1/72
chances getting X 3/5 times ==> 5 C 3 = 10

so 3 PLL skips in of one avg 5 neglecting order = (1/72) / 10 = 1/720

I'm sorry, but that's not even close.

Odds of getting a skip : 1/72
Odds of not getting a skip : 71/72
Solves with a skip : 3/5

1/72^3 * 71/72^2 * 5/3 = 1/226,845

I would think that's how you would do it, correct?
 
Last edited:
Back
Top