> You’re looking at volume here when you need to be looking at efficiency. Players who throw more pass for more yards.

I did that at https://www.reddit.com/r/dataisbeautiful/comments/122c5b2/american_football_starting_quarterback_pass/

Why should I let a player's past greatness affect their current stats? If they have one foot out the door to retirement now and aren't doing so great now then they aren't doing so great. And probably the overall tenor of the game has changed over time. For instance, as overall player height increased in basketball, we see an overall net increase in slam dunks over long three pointers. If you can take two or three steps and have a better chance at making more points then why not do that.

However, i admit a greater volume of data might show something different. Feel free to put that together, I'm excited to see what you come up with. :)