I'm buried in numbers on a regular basis. Whether it's from those I chart for Reception Perception, nuggets dug up from scouring game logs or even passed along from the NFL's Next Gen Stats tracking data, I'm not short on stats. Some of my co-workers even assigned me the mockingly authoritative "captain spreadsheet" nickname.
However, I'm not a fan of all football stats. To be quite honest with you, there are more than few that I actively dislike, or find flat-out terrible and outright deceptive in the pursuit of understanding what matters on the football field.
The statistics that I tend to gravitate toward as items that are consequential and descriptive of the reality of the game are those that can actually be well-explained in a typical football sense. This is why last offseason, when I read Brian Malone's discovery that running quarterbacks throw to their running backs less often than the average signal-caller, it made so much sense. It's rare that a runner out of the backfield is the primary read on a play call, more often they are a check down option and third or later in the progression. The running quarterback is less likely to hit that third or fourth progression before their instincts take over and they break the pocket to run. Therefore, running backs playing alongside a mobile passer are less likely to see as many targets as their counterparts teamed with traditional pocket players.
Malone's findings backed up that intuitive "football-based" conclusion. From 2000 to 2015, among quarterbacks (10 total) with both 250 pass attempts and a run-pass ratio of at least 1:5, only pre-conversion Terrelle Pryor of the Oakland Raiders threw to his running backs at a rate above the 20.9% target share league average.