So, for those who want it in full, here is the joke I referred to earlier.

A geneticist, a physicist, and a statistician are all asked by a gambler to advise him on which horse to place his money in the Melbourne Cup.

The geneticist goes away, and analyzes the breeding lines of all the horses in the race, but after several days of effort, he reports back saying "it’s just too difficult – I can’t tell which horse has the best chance from this alone."

The statistician has a go. He looks up the results of every race all the horses have ever been in, places them into his favourite statistical analysis software, and has it churn out more statistical data than you thought possible. But it is all too variable, and he comes back to the gambler saying "I’m sorry, but the standard deviations are just too large – you’re taking too much risk".

The physicist then receives his instructions. "No problem", he says, and a day later he is back waving a page full of diagrams and mathematical equations and with the instruction that the bet should go on horse number 4, Delightful Dobbin. The gambler rushes out and places the bet, and come Cup day watches excitedly as Delighful Dobbin trails in a distant last.

The physicist gets a summons to see the disgruntled gambler, who, understandably, is not in the best of moods. After listening quietly to the complaint, he offers his explanation. "Oh, I’m sorry, but didn’t I tell you?; my calculations assume a spherical horse rolling through a vacuum."