Problem is, those facts are statistically meaningless

Quote
dzrams
31.9
42.7
47.7

That’s Keenum’s QBR progression. Clearly, it is showing an upward trajectory. That trajectory shows improvement which nearly renders meaningless the 31.9 QBR he posted in his rookie season.

Keenum’s 2015 QBR with the Rams was 47.7. That’s higher than the rating for several notable names: Andrew Luck, Blake Bortles, Sam Bradford, Joe Flacco, Peyton Manning, and Ryan Tannehill. It’s pretty close to the QBR that Derek Carr put up in Oakland.

Oh yeah and it was much higher than Foles’ 30.01 QBR with the Rams.

Also, his 2015 QB rating with the Rams was 87.7. That’s higher than the rating for several notable names: Andrew Luck, Peyton Manning, Sam Bradford, Jameis Winston, Joe Flacco, Brock Osweiler, oh and of course Nick Foles.

Keenum is not great but if he continues to do what he did with the Rams last year, he’s much better than Foles.

The sample sizes are too small and uneven to form any kind of meaningful "trajectory." But the other problem is, I don't think your QBR's are accurate. Are you looking at ESPN here? If you are, it shows the opposite of the trajectory you've listed. It shows this:

2013 = 52.2
2014 = 43.4
2015 = 42.2

But even if those are the numbers and I'm missing something, it's the problem with sample sizes like his 2015 season of only 125 PA's over 5 games where a single great game out of those 5 skews the overall sample significantly.

That trajectory would be based on 8 games in 2013, only 2 games in 2014 and 5 games in 2015 is entirely dependent on a single good game played last year against the Tampa Bay Buccaneers and a single bad game played in 2014 against Baltimore.

No matter how you look at it, he really has no meaningful trajectory at this point. What he has is a full season's worth of NFL starts now in his career (16) and a pedestrian aggregate statistical performance.