Originally Posted by
UlleyMiller
I love these threads.
Just offering my learnings:
After 10 heads with a fair coin, the guy with the Aphex Twin song for a username is right, the next flip could be heads or tails, and the prior streak has no bearing on the next flip.
The 'law of averages' is often confused with the principles behind the law of large numbers (in a similar way to how karma gets watered down to make people think if they hold doors open, the same will happen for them). If you're playing with astronomical sample sizes, the logic will hold true, but when you don't properly consider the relatively tiny number in the sample size and the slow rate at which a new draw is added to the sample, you're applying maths to blindfolded darts.
There are claims that 53% of prizes north of £1m have been won by lucky dips. This in itself is also meaningless. If 99% of all tickets are lucky dips, that suggests they're rubbish, if 3% of tickets are LDs it suggests they're epic. The split of lucky dips to non-LDs matters too... although it's not the only consideration.
Lucky dips are not random. They're generated by a computer, and it's not possible to get a truly random set of numbers from a computer (not that I've ever seen evidence for, or even seen anyone proposing evidence for).
Some 'systems' will increase your chances if you are a subscriber to conventional probability theory, but the increases are absolutely miniscule.