This is a leg from the recent Valley Goat --

As a training exercise, I ran most of the long legs 2 or 3 different ways to compare the times. In all cases except this one, the route I expected to be the fastest, was. In this case, however, I expected running in the yellow to be the fastest, but dropping to the trail and climbing up at the end was 18 seconds faster. This was a surprising result to me.

The first thing to note was that I ran this test on consecutive days, rather than the same day. I guess that could invalidate it, but all the other route tests run the same way had the expected results. I tried to run as hard as possible both days.

In any case, whether or not this test is valid, it is interesting to speculate as to what could cause this variability. The possibilities seem to include --

• The trail was genuinely faster, and my intuitions about routes, even after study, are sometimes wrong. This seems likely perhaps in general, but I don't buy it here, but how do you know?

• Weather. It was cold with a strong headwind both days. Dropping down into the trees/trail will reduce the effect of the headwind. Could headwinds have that large of an effect to affect route choice decisions? In the other route tests I ran, the headwind, I think, was basically a constant for the competing routes.

• Variability in nutrition and hydration. Different days will have differing levels of "carb loading" and prehydration. Could differences in these things lead to this much variability. If so (and I think I doubt it), it is an area for me to learn about, and perhaps learn to control for improved performance.

• Warmup and fatigue. Different levels of fitness and fatigue on the different days perhaps is a likely culprit. But I was more fatigued, if anything, on the day I ran faster.

• Sometimes you kick, sometimes you get kicked. Some days I just feel I "have it" (physically and mentally), and some days I don't. I've never gotten to the source of this variability. This variability can be significant in my case, up to 10% on training runs in equal weather. Getting to the source of it is one of my holy grails, but I am clueless about it.

• Experimenter bias. Perhaps I really pushed harder on the unintuitive route so I would get the surprising result, because that is what would make the leg more interesting. If this variability is the sort of thing that is able to be harnessed by willpower, that is a good thing.