To expand on what I already posted... well the first two games having bad grades are really no mystery.
QBR accounts for context and leverage and horsefeathers like that, so there's probably something about the 4th game that the formula didn't love.
I'm not vouching for it or saying it's a great or terrible stat either way (and, like with most football advanced analytics, given the nature of the game and the sample sizes involved, you need to take it within proper context and/or with a grain of salt), but it's probably no worse than the arbitrarily designed and old passer rating.
There's a decent breakdown here - https://en.wikipedia.org/wiki/Total_quarterback_rating#:~:text=The QBR calculation accounts for,expected points added" per play.
EDIT - Because I'm a dummy and lost track of the schedule. Said 3rd game when I meant 4th lol.