Franklin said:
The error margin on this particular excercise (101 sample, 100 tosses) is less than 0.00 => As someone said, if this happens on the street, what is your bet?
Thank you providing me another example. Let me be clear: I am not talking here about mathematics or physics of coin toss, I am pointing how people create (and mostly unitentionally biases).
a) Man is flipping a coin;
b) Street hustler is flipping a coin;
c) Honest scientist in his laboratory is flipping a coin.
Do you see difference?
If we are talking mathematics, there is no difference. But if we are talking about biases and how people perceive what is happening, then there is difference. At first you made your answer conditional, now you endorse another bias creating mechanism. If I am not mistaken in terminology, psychologists are calling it "priming". You are add one small bit of new information (in this case street hustler) and it changes how we judge situation (including probability).
Franklin said:
Again, you should know that sample size counts. In this case we have a large sample and extremely heavy bias. And that's just GT winner and dope. Then we look at dirty doctors, lieing managers, prior palmares. Every spread we have is heavily against Froome.
Actually we do not have large sample size. We have opinions, interpretations, we have some facts (but even these facts are biased to interpretation and sometimes interpretations are presented as facts), but mostly we have huge caps of relevant information our i.
I am not saying that Froome is clean, if someone ask me to make a bet, I´d decline (I do not have enough information). If I am forced to make a bet, then I guess, I´d bet against Froome (but even in latter case I am not sure is my decision influenced by "real facts" against Froome, or influenced by dislike of Froome). And anyway, I am not here estimating Froome, I a talking in more general terms: I dont think we can make confident estimates in these situations.