malcolmocean comments on How to Run a Successful Less Wrong Meetup - Less Wrong

55 Post author: Kaj_Sotala 12 June 2012 09:32PM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (35)

You are viewing a single comment's thread.

Comment author: malcolmocean 14 April 2013 03:16:38PM 1 point [-]

This document is awesome.

Tiny bit of feedback on Fermi scoring. The Science Olympiad rules appear to get logarithmic rounding wrong.

Generally, fives can round either way, depending on if you think your answer is high or low.

When dealing with orders of magnitude, the centerpoint for rounding should be halfway between the two values, except in this case the value is the exponent, not the mantissa/coefficient. This means that halfway is sqrt(10) = 3.16

So in general, values of 3 or lower should be rounded down and values much higher than 3 (even 3.4) should be rounded up.

3.68 x 10^3 ~= 10^4