# Random permutations of Z_n

In http://www.springerlink.com/content/y19u81675243r237/fulltext.pdf, the author states the following without proof (equation 3.1):

Consider a random permutation $\pi$ of $\mathbb{Z}_n$. What is the probability that $\pi(i+1)-\pi(i) \pmod{n} <n/2$ for all $i$?

The claim is that this is $(2+o(1))^{-n}$, which makes sense and seems like it should be a standard argument. Does anyone have a formal proof?

-
The maximum $\pi(i+1)- \pi(i)$ can be is $n-1$ in $\mathbb{Z}_n$. –  PEV Jan 31 '11 at 23:09
I think you mean (2 + o(1))^{-n}? –  Qiaochu Yuan Jan 31 '11 at 23:36
Good call, Qiaochu. Fixed. –  Jeremy Hurwitz Feb 1 '11 at 6:21

This question was answered at MathOverflow: http://mathoverflow.net/questions/54051/random-permutations-of-z-n

-