Running Nerd Returns

Neither sun, nor heat, nor humidity, nor more heat . . . .

July 2016 was the hottest month on record (that record being my running log) where I've dutifully recorded the temperature of nearly every run I've done in the last few years. Initially my main goal was to develop a calculator that would help me remember what to wear at what temperatures every fall when it starts to cool down. That was complete a few years ago, and at this point I think I could actually get along just fine without it - but it served its purpose.

But all this heat has got me thinking: can I blame the heat for being so slow? After a very nice prolonged drop in average pace, I've gotten slow in the last few months:

After a multi year trend of getting faster, that darn blue line is headed in the wrong direction. But, who is to blame? The pink bar, or the green line? Let's see what the data says . . .


Suspect 1: The pink line.
I've certainly been running more, and it seems logical that being tired might be slowing me down. I can't exactly do a comparison between these recent 130+ mile months and previous 130 mile months, because there aren't any. There are barely even any 100 mile months. But I can compare against a few other things that are related.

There's two ways to get more miles: run longer, and run more often. I've been doing them both. In the last 3 months I've logged 68 runs. My highest 3 months from last year only totaled 54 runs. Running more often means fewer rest days, and perhaps that lack of rest makes me slow. Here's a chart that compares my running pace today to the pace the last time I went running subdivided up based on the number of days since I last ran.

 

And that pretty well sinks the idea of rest making me faster. Whether it is one, two, three or 4 days of rest, I apparently only speed up or slow down one or two seconds per mile. Not enough to explain my 20 seconds of slowing. Apparently after 5 days, I do get suddenly faster, but then that effect goes away when I rest a 6th day. (Or we can chalk that one up to only having 4 data points. 

I've also averaged 6.4 miles per run over the last 3 months, whereas the previous 12 months ranged from 5.2 to 6.0 miles per run. The longer a run the slow I go, right?

Wrong. From 4 miles out to 15 miles, the blue line shows that I only slow down (on average) less than 15 seconds per mile. So increasing my average run length by 0.5 to 1.0 miles isn't enough to explain another second or two of slowing. Further, you can see from the big spread of red dots stacked up on the 4.5 and 5 mile runs that I do over and over, that I have a large spread (more than a minute per mile) on these runs. There is actually less spread over that cluster of runs around 11 miles. Let's move on to the next suspect.

Suspect 2: The green line. (a.k.a. "Waaaah! It's hot!")
Summer time means it's hot and, in this part of the country, humid. And I never get out to run as early as I'd like, so I'm not helping myself out. So, here's what my running speed looks like against the temperature:

I've got a parabolic fit line in there, which shows that when it's hot, I run slower, and when it gets really cold, I start to run a bit slower, too. I've also got the against the "perceived temperature" which is the wind chill when it's cold, and the heat index when it's hot:

The difference here isn't as significant as I thought. The highs and lows are stretched out a bit more, but those parabolas are really quite close to the same. If I use the fit parabolas to predict a pace at 20 or 80 degrees, they only vary by 2 or 3 seconds per mile. But they do predict that at 20 degrees I'll be around 7:22/mi, and at 80 degrees I'll be at 7:45/mi.

These can be used to calculate the temperature at which I am fastest though, which they do to pretty decent agreement. At an actual temperature of 36 degrees (or wind chill of 33) I should run everything at 7:20/mile. And going back to the original chart up at the top, my fastest months were run when the average temperature was near optimal: March and April both averaged 37 degrees for my runs, and I ran them at 7:08/mile. June and July averaged 70 degrees, and I ran those at 7:28/mile, which can be explained easily enough on temperature alone.

So, there's your conclusion for the day. Remember, Running Nerd always gets his man . . . er . . . trend line.