The paper in the OP cites this paper which has been my benchmark for a while on the level of “science” going into studies around Thursday night football and injuries.

Somebody posted a critique of that article where he highlighted two issues. The author replied to that critique basically saying “yeah you’re totally right, my bad” but I don’t think ever corrected the original paper? Anyways, the two issues were:

• He miscalculated a statistical ratio. This is kind of damning for a study purporting to be statistical in nature, but whatever that’s kind of boring to most people.

• For the second, much more interesting issue, I will just quote the critique:

Second, it is also worth noting that there is an explanation for why we might observe a lower reported injury rate on Thursdays even though it is implausible that shorter rest exerts a protective effect. It is a quirk of the way the timing of NFL injury reports work [Citation2]. Teams typically issue their first injury reports on Wednesday of each week: 3 days after a Sunday game but 6 days after a Thursday game. Any injury that resolves to the point the player can fully practice in that 3–6 day period would not be reported if it occurred on a Thursday but would be reported if it occurred on a Sunday. If 20% of reportable injuries resolve in that period, that would fully explain the lower Thursday injury rate.