And if there's a tie for the first place, I recommend using 7-7 questions from both winners (5-5 if there are three winners). Thus, the tied winner could complere half a (one third) round as well.
That's too much of a mess.
I'd also recommend using 0.5-0.5 (0.33-0.33-0.33) points if there are multiple best estimators for a question (this should be applied for all previous rounds as well).
Yes, this is obvious. I don't see why some questions should be worth twice as much as the others. The game is about coming closest to the answer. "Closest" means closer than everyone else. If you and another person have the same answer, you were not closer than the others. You were equally close. You weren't as good as someone who was closer than everyone else.
An easy example to demonstrate that the current system is nonsense and unfair:
How long is blablablablabla? Answer: 24 m
A: 19
B: 19
C: 19
D: 19
E: 19
F: 19
G: 18
Now A, B, C, D, E and F each get 1 point, but G gets 0 points. This means that G loses a lot in the overall standings (because not only one person, but several people are awarded more points than him), which he would not have lost if the others had different guesses, even if their guesses had been closer to the actual answer. (for example: 24-23-22-21-20-19).
But anyway, it's only a small issue because the actual effect is minimal. However, I can't see how anyone could think the current system is better in principle.
