Reply To: Could someone explain it to a non American please…?

@CTLAWYER

@huju

@Avi K

@jackk

@Reb Eliezer

Thank you @CTLAWYER for the clear answers.
Re your point:
“The AP does NOT decide the election results, election officials in each state do. However, the AP and other news organizations call or make predictions. They are considered reliable by many, but not all.
… It is not over until the votes are certified by authorities in each state”.
Great, that’s exactly what I was asking about. The media is treating it as though the results are finalised and official. It’s fair to make highly likely *predictions*, but journalistic integrity demands that they refer to the results as “predictions”, or to the candidate-elect as “projected”, “presumptive” or whatnot. It is dangerous to treat the current numbers as final results. (Especially when those results are so razor-edge and being contended legally.)

Re:
“The ballots to be counted are being predicted to follow the similar percentages as those already counted”, or as @Reb Eliezer said “check … what is the statistical trend in the area”.
If so, then there should be little difference between calling it when there are 5% of votes left to count and a variance 1%, compared to when 50% of the votes remain uncounted and there is a variance of 10% between candidates. Ok, a statistician can rightly disagree with me on that. But my point is that, being that it is mathematically possible for the result to shift sides, it is irresponsible to not count all the way before calling a result. Especially when there is SO much hanging on these razor edges. (And especially when a change is not just possible but has a fair likelihood, and especially when we see different pockets of ballots have very clear leanings, and don’t just reflect state-wide averages.) And, on that last bracketed point, apparently some of the ballots still uncounted are from counties which are indeed predominantly right-leaning, even if the state is more left.