Avocado's Number wrote:
One subject that I haven't seen much discussion of is the effect of pacing on neuromuscular fatigue and running economy even after controlling for air resistance. When I was working long, stressful hours and getting little sleep, I often did my most intense training on treadmills, in part because I was (and still am) convinced that the task of running at a constant distance behind an object (whether the front panel of a treadmill or the body of a human pacer) is, as a neuromuscular matter, less complex and less fatiguing than the task of running at a specified speed without any assistance or cues from pacers, training partners, or similar sources of feedback.
Of course, none of this is to suggest that there aren't other "pschological" reasons that people tend to be able to produce faster times when they have special incentives to do so.
My wonky $.02:
The progression of neuromuscular fatigue is by definition partly psychological. Voluntary muscle activation, the one subject to central
NMF, is mediated in part by serotonin, dopamine, and noradrenaline. Since the latter is affected by the rate of perceived exertion, I would argue that the motivation and sense of competition involved in running behind a human pacer will almost always be greater than what you’ll feel on a treadmill. Unless it’s a treadmill race... Which, btw, is a study I’d love to see.
My anecdotal experience: I’ve done my share of treadmill hours for the same reason as you. Even when training for races, I’ve never clocked my best times on a treadmill - and I’m pathologically competitive, though not an elite runner.