Just how realistic are the bounds in Chebyshev's inequality?

66 Views Asked by At

I was reading about Chebyshev's Inequality

$$ P(|X-\mu|>k \sigma)<\frac{1}{k^2} $$

and the following came to mind: We know that it is rather conservative. It's a crude estimate for all the common probability distributions. My question is can we find a distribution for which the inequality is strict ? Or perhaps to within a constant factor ?

Thanks in advance!