The Lottery

[robvalue]

Hmm... Well, if the balls are not all equally likely, that would suggest some bias in either the selection process or the physical nature of the balls themselves. For example, if one ball was heavier than the others, you would probably see it come up more often.

However, you first have to determine whether the variance you see in the data compared an expected probability distribution is significant. That could be done, I imagine it must have been done somewhere. For example, if I throw a coin 100 times and get 60 heads, is that significant evidence that the coin is biased? There are statistical methods for testing this (often using the normal distribution as an approximation to other distributions, in this case binomial).

If the difference is actually significant (which depends on your chosen level of significance, there are two types of error that can occur and which cannot be eliminated, only minimised) then you could build a model based on the data you have.

I would highly doubt the differences are significant, although I've not looked up any studies. It would suggest a big problem with the equipment, which I presume they would have already addressed.

Like most gambling games based only on chance, there is no system. In the case of the lottery, all you can do is try and pick a combination which the least number of other people are likely to pick; so if you win, you share the jackpot with the least number of people.

Maths.

---

Sorry if I cocked any of that up, not done this stuff for a while