Statistics are funny in that the past is not predictive. Consider this. If you toss a coin into the air a thousand times it will generally come up head half the time. Wo you might think that if it has come up tails 100 time in a row the next toss has to be heads?
Not at all, it is only the same 50% if the coin isn't rigged.
The coin doesn't care about the prior tosses and the next one is brand new.
This would seem to contradict the fact that in the long term the number of each instance should end up abut the same (heads or tails). It doesn't, but it seems like it should.
Of course this sort of logic can seem persuasive so in the recent election we saw a lot of trash statistics thrown out to prove nothing really.
For instance some places had voted for the presidential winner most of the time but not this time. This means absolutely nothing.
Still a lot of nonsense data is still out there nd is still believed by some.
Think snake oil salesman. The product is phony and they jst want your money or your vote.
Tar and feathers anyone?
No comments:
Post a Comment