Quick Thoughts on Grading

Working through our annual position-by-position recap, I thought of something I really haven’t considered before: we don’t really do grades for players.

Sure, I make a ruling on whether or not a player has “met expectations” or not in a given year, but we really don’t go deeper than that. It’s always struck me as the intuitive way to do things, but I’ve never really tried to articulate why that is. Why do things that way instead of assigning one player a B+ and another player a C- or what have you?

For starters, I think if you try to come up with an established grading system, you’re going to end up spending more time on that system than the grades themselves. Think about how hard it is to encapsulate everything that goes into how a player performs on the field. We all know stats don’t tell the whole story, but even a meticulous review of game film can leave a whole bunch of things out. 

How do you encapsulate the difference in statistical production from one year to the next when the scheme a player plays in changes? How do you account for the performances of his teammates? What about injuries? What about opponents? There are so many variables that defining what to do in a given circumstance would almost always result in using customized grading scales for every player, which defeats the whole purpose.

You could, of course, devise some kind of grading scheme where you’re comparing the player to the theoretical best at his position, but that, too, seems fraught with issues. How do you determine what’s best? Is it the best player of all-time for that position? How do you determine that?

And then there’s the very issue of publicizing your scale. Some people are very protective of their grades and the process by which they come to them. But if we don’t know how you’re grading, of what value are the grades?

So I’m left thinking that the only way to grade a player is to come up with some kind of definition for what we expect for that player before the season, then return after the fact and see how they measured up. There’s so much we can’t know about players, and so much of what we do know is context dependent. I think the only way to grade is to compare how players did to how we thought they’d do and call it good from there. If nothing else, then we’re grading players by what we thought of them to begin with.