Samuel Wanjiru New World Record Holder

[joeltetreault]

It is a little misleading to look at the frequency distribution of the fast times (vs Temp) only without looking at the frequency distribution of the top races run at those temperatures. [For example, assume that one one top race was run when temps were 1-3C at start and that race was in the top marks (e.g., WR); it would only have one entry but the conditional frequency would be 1. I suspect that if you looked at Fukuoka, which had some cool temps and plotted times, adjusted maybe for the runners PRs etc. you could get some insight.

I am not disputing the physiology comments and think 45-55 is best for start and 50-60 for finish, also depends on sun, wind, etc.

I totally agree with all of the above...one runs into all sorts of independence issues when simply looking at weather. Quite frankly, there are a lot of other factors that go into a race that can neutralize or exacerbate whatever weather effects that may exist. When I first posted the weather conditions for the top 11 marathons, I also gave a similar disclaimer afterwards in the same thread:

11 data points is way too small to draw any conclusions, and to really look at some trends, one would want to get stats on all the sub2:09 or sub2:10 marathons to do more rigorous statistics. Of course, it all depends on what you want to look at/prove. My feeling is that there are a lot of other confounding variables that make looking at the effects of temperature very hard: including the points above, there is quality of field, presence of Olympics that year, prize money, planned pace for the rabbits, etc. I think if you could get all this data for all the major races, and then normalize the times to the WR at the time, one could investigate the effects; of course, getting all this data would be very very hard. Well I haven't put too much thought into this...I suppose one could assume each individual marathon is fairly consistent from year to year and then look at the effects of temperature that year, and repeat for each of the top marathons. That "study" would be a bit easier to pull off.

http://www.trackandfieldnews.com/discus ... torder=asc

[gh]

Data point discussions aside, I would at least hope that most of you would now agree that yesterday's temperatures, regardless of where they might be on the optimal scale, were in no great way detrimental, and instead seemed to have some sort of salutary effect, at least empirically.

According to Tilastopaja stats, of the top 30 men, only 3 didn't get either a seasonal best or a PR (places 1, 3, 6). Of the other 27, 20 got lifetime bests.

On the women's side, 5 places (1, 2, 4, 5, 7) didn't get a best. Of the 25 who did, 20 got PRs.

So combined, 40 of the first 60 got PRs, only 8 didn't get a PR or SB.

[Gebfan2]

Data point discussions aside, I would at least hope that most of you would now agree that yesterday's temperatures, regardless of where they might be on the optimal scale, were in no great way detrimental, and instead seemed to have some sort of salutary effect, at least empirically.

According to Tilastopaja stats, of the top 30 men, only 3 didn't get either a seasonal best or a PR (places 1, 3, 6). Of the other 27, 20 got lifetime bests.

On the women's side, 5 places (1, 2, 4, 5, 7) didn't get a best. Of the 25 who did, 20 got PRs.

So combined, 40 of the first 60 got PRs, only 8 didn't get a PR or SB.

Wouldn't have guessed this. Suggests that the ideal temp may be lower than the mid 40s F. I would have guessed that the cold temps cost Wanjiru at least 30 seconds, maybe more. But all those PRs suggest otherwise.

Joel and 26.235:

Good points. Excellent work on your previous post Joel. But he fact that none of the fastest marathons has a starting temp over 55F says something.

Take the 2007 Chicago marathon as an example of how heat affects performance. 2007 was a statistical outlier for temps in Chicago with a starting temp over 70F, and a finish over 80F. Adere won the 2006 marathon in 2:20:42, and the 2007 marathon in 2:33:49. Patrick IVuti won the 2007 men's race in 2:11:11 (edging out Gharib), but ha placed 5th in 2005 in 2:07:46.

But how much does heat affect distance running performance at different distances, and which temps are ideal? If your point is that there needs to be a lot more study of the data, I agree.

One more observation. Wanjiru seemed to skip or drink very little at every water stop yesterday, and it didn't seem to affect him. He drank lots of water at the Olympics last year. He seemed to know exactly the amount of fluids he needed for the temp he was running in yesterday...

[malmo]

Data point discussions aside, I would at least hope that most of you would now agree that yesterday's temperatures, regardless of where they might be on the optimal scale, were in no great way detrimental, and instead seemed to have some sort of salutary effect, at least empirically.

According to Tilastopaja stats, of the top 30 men, only 3 didn't get either a seasonal best or a PR (places 1, 3, 6). Of the other 27, 20 got lifetime bests.

On the women's side, 5 places (1, 2, 4, 5, 7) didn't get a best. Of the 25 who did, 20 got PRs.

So combined, 40 of the first 60 got PRs, only 8 didn't get a PR or SB.

Wouldn't have guessed this. Suggests that the ideal temp may be lower than the mid 40s F. I would have guessed that the cold temps cost Wanjiru at least 30 seconds, maybe more. But all those PRs suggest otherwise.
..

Exactly how do those PR's suggest otherwise?

[26mi235]

Citing SB in the 100/200/400/800/1500/ and 5000 makes the point about being good conditions (or at least not bad) but how many marathons have they run this year 0, 1, 2? This is a fast, flat course and if they ran Berlin in the summer or another marathon in not great conditions, a SB means very little. I got one every time I ran.