So after 16 pages, how far have we gone, and how much further can we go?
On one end of the spectrum, Renato declares the wind effect to be not more than 40 seconds.
On the other end, we have several "scientific" approaches which say that such high winds should generate 3-4 minutes advantage.
Seemingly in the middle, we have Ken Young (ARRS) who somehow statistically computed a race time bias of 1:37 (for the men), for a course that is usually 13.5 seconds slow. This means the effect of wind/weather is a net 1:50.5.
Even a letsrun analysis (point 7 in the week that was) seems to suggest a 2:00 minute advantage over 2010 (also wind aided).
Ken Young seems to have created controversy/support for saying it's "statistially valid", and not "excessively aided", but looking at the ARRS website, and you see that the times are also not "record quality". They will not be considered an "ARRS record". The times are included, and coupled with the bias, will be used to calculate world rankings and marathon comparisons.
It may not seem apparent, with all these numbers floating around, but the ARRS calculations seem to support Renato. The 2011 bias gives Mutai a "normalized" time of 2:04:37, and Mosop 2:04:41. On a course like Berlin, with an average bias of 68.5 seconds faster, these times become 2:03:28.5 and 2:03:32.5. Even London would have given Mutai a 2:03:58.2