Originally Posted by noisebeam
For example if you had data that showed death during sleep happens to 20% of people, do you assume you have a 20% chance of dying while sleeping?
Yes, and no.
If I don't have the additional data, then yes - because that's all I know.
Consider the odds of picking the winning numbers in a lottery. Say it's one in 100 million.
Now say you have additional data... from your crystal ball which tells you the winning numbers in advance. Now your odds are looking much better.
The point is that for 99,999,999 combinations, the odds of winning are actually 0, and the odds for the winning number winning is 1. But when we don't know which is the winning number, we just say the odds are all the same for all the numbers... 1 in 100,000,000.