Let $X$ and $Y$ be any random variables defined in the same sample space. What do we mean exactly by $X>Y$?

Also, what do we mean $X>c$, where $c$ is a constant?

Thank you very much!

  • 1
    $\begingroup$ How are you defining a random variable? $\endgroup$ – anomaly Feb 11 '17 at 13:00
  • $\begingroup$ Thank you. Let $X$ and $Y$ be any random variables defined in the same sample space. $\endgroup$ – Probability is wonderful Feb 11 '17 at 13:16
  • $\begingroup$ I was asking whether you're defining a random variable as something vague like an "event" (which is standard for probability courses that don't talk about measure theory), or as a measurable function $f:\Omega \to \mathbb{R}$ (or $\mathbb{C}$, etc.) as below. $\endgroup$ – anomaly Feb 11 '17 at 13:40
  • $\begingroup$ I see. Sorry for my inaccuracy. I'd like to define a random variable as a measure function, like what you just said $\endgroup$ – Probability is wonderful Feb 11 '17 at 14:08

Random variables are measurable functions from a measurable space $(\Omega, \mathcal A)$ to another measurable space $(G, \mathcal G)$. Usually the latter space is $(\mathbb R, \mathcal B(\mathbb R)$. So saying $X > Y$ means that, for every $\omega \in \Omega$, we have $$X(\omega) > Y(\omega)$$

Of course since $X(\omega), Y(\omega) \in G$, it is implied that the relation "$>$" is defined in $G$

  • $\begingroup$ I see. Thank you very much! So, $X>Y$ implies that $P(X>Y)=1$ , right? $\endgroup$ – Probability is wonderful Feb 11 '17 at 14:10
  • $\begingroup$ @Probabilityiswonderful yes :) $\endgroup$ – Ant Feb 11 '17 at 14:12

While @Ant's answer is correct there can be another meaning. "$X>Y$" can refer to some event, which can be formalized as $$ \{\omega:X(\omega)>Y(\omega)\}. $$ In which case $P(X>Y)$ can have any value.

You can understand $Y>c$ in the same way by considering $c$ a a (constant) random variable $(c(\omega)=c\ \forall \omega)$.

A simple example: throw two dice, observe the scores $X$ and $Y$. You can compute $P(X>Y)=5/12$.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.