First of all, it’s easier to get a defense to play cohesively with less preparation than an offense. This hypothesis should engender an expectation that point totals would be lower on Thursdays than they would be for a game between the same two teams playing on a Sunday of the same week. But is that accounted for correctly in the over/unders?

With this in mind, the season-opening Thursday game week one shouldn’t necessarily be taken into account because there’s no shortage of time for teams to prepare for it. And after a certain point mid-season, the continuity of repetition should begin to outweigh any perceptual bias.

This brings us to another very important and misunderstood point. There is a pervasive misconception of how betting lines are set which is explained here.

The point where the cutoff should be for public expectation to begin matching reality is somewhat arbitrary for any given season. But for consistency’s sake, we’ll put it immediately before week nine.

If a person had bet the under on Thursday games from week two through week eight for the past five-plus seasons beginning in 2019, they’d be 25-14. Anyone betting overs from week nine onward across that time span would be 28-21, for a combined total record of 53-35.

Someone wagering $110 per bet with traditional juice ($110 stake earns $100) would be about $1500 ahead.

Statistically speaking these are small data sets. But odds/results history going back to 1952 (over/unders back to 1977) can be found here:
SportsOddsHistory.com

—@PoisonPill4