A problem from probability

1. Jul 27, 2005

philosophking

Hey guys, I'm pretty sure this problem is from probability, but I have no idea how to approach it.

The other day my friend gave me this problem:

Say you have the cartesian plane (R^2), and you take some infinite line that divides the plane up into two half-planes. Say I now randomly throw a point down on the plane: what is the probability that the point will land on one side of the plane?

My friend said I'd be surprised at the answer because it's not 1/2, but then my question is, what would be the answer? It sounds like it would be something from measure theoretic probability, but I've never studied that (nor basic probability, actually).

The reason I thought to post this was that there was a question similar to this one that was posted in this forum a while back. Would the answer to this problem be zero, because there are infinitely many points?

What kind of assumptions would have to be made to make this problem solvable, if it is not solvable already?

Thanks for the help.

2. Jul 27, 2005

mathman

The basic problem with your example is that throwing a point on the plane cannot be done from a uniform distribution - you can't define one for an infinite plane, or an infinite line in one dimension. Therefore you must define a legitimate distribution function. The probability for each half plane will depend on the relationship between the line and the distribution function.

3. Jul 27, 2005

philosophking

Okay, thanks a lot!