For me, only option (2) of the original proposed three works. The opposite of proving a hypothesis is to

**disprove** it, or to prove it wrong. EM's "refute" would also work. "Disconfirm" strikes me as unidiomatic, and although providing a counterexample is

__one way of__ proving a hypothesis wrong, I wouldn't really call it

__the opposite__ of proving it right.

If he's only just coming up with the hypothesis, how can you already have the experimental evidence that proves it wrong?

Perhaps because you had already come up with a similar hypothesis yourself, and done your own experiments that show it to be wrong, but he is not aware of your work.

You might have observed that the 3rd, 5th, and 7th Fibonacci numbers are 2, 5, and 13. This might have led you to hypothesize that if N is prime (and greater than 2), then the Nth Fibonacci number is also prime. To confirm this, you looked at the next few and found that the 11th, 13th and 17th numbers are 89, 233 and 1597, which are indeed also prime. Just before trumpeting your discovery to all your colleagues, you looked at the 19th number, which is 4181, and found to your dismay that it can be factorized as 37*113. So you kept quiet.

Then, one day, someone else comes up with the same idea. Now you can smugly point out that he's wasting his time.