Joe Sheehan penned a piece over at Fangraphs that caught my eye. He talks about how the new-school kids of the 1990s have become the old-school adults of the 2010s. Tell you what, it’s true for me. But then again, I spend my time looking backwards at baseball, not forwards.
I am fortunate to work for a company that publishes professional literature for teachers. I’ve learned an awful lot about what comprises best-practice instruction. Good teaching hinges on effective assessment. Assessment is figuring out what students know and can do. Don’t worry, I’m actually headed toward a point here.
There are three basic forms of assessment that go by many names but boil down to these:
- Screening: A benchmarking that determines where each kid starts from and supports instructional planning
- Formative: Monitoring the progress individual students make on a granular, skills-based level
- Summative: Measuring how far a learner has come overall, or “student achievement” in the parlance of our times.
In baseball analysis today these three assessments are mirrored this way:
- Projections: ZIPS, Steamer, PECOTA, and others determine where a player is at prior to the season and what his likely performance trajectory looks like for the next six months
- Statcast, fx stats, heatmaps, etc…: Measurements of skill that indicate how the player’s performance compares to his underlying abilities and his projected performance
- Value-based Stats: WAR, WPA, Win Shares, and whatnot measure a player’s overall season-long “achievement.”
Take a player’s offensive value. At the primary level, offensive wins are what we’re after, and those derive from batting performance against average in most value-stat systems. Secondarily, part of that value resides in, for example, how many singles the player collects. Ground ball/flyball data exists for much of the game’s history. This tertiary level of information doesn’t carry enough meaning to matter when measuring value, no matter how our player got his singles. At a quaternary level, Miller and I have no need for launch angle even though it tells us a great deal about the nature of a player.
This isn’t a fault or bug with radar-based numbers. Not in the least!!! What they do is provide a detailed cross-section of the machinery of baseball’s numerous offense/defense relationships so that we can see minutely how each of its gears operates. But for me and Miller that granularity is superfluous. That and the lack of this information for the other 99% of big-league history makes these stats nearly useless to the mission of the Hall of Miller and Eric. We cannot compare Babe Ruth’s exit velocity to Ted Williams’ to Willie Mays’ to Mike Schmidt’s to Barry Bonds’ to Mike Trout’s. We can only currently compare Trout to his exact contemporaries. Nor are we yet convinced that doing so would be helpful.
So we do what we can. We use summative numbers to compare apples to apples to the best of our abilities. WAR to WAR. Of course, we make lots of adjustments to WAR because of certain beliefs we have about baseball history and about the statistical record. The skill-based big-data numbers can’t really help. Does that mean we ignore them? Well, actually, yes. Not maliciously; they just aren’t terribly meaningful yet for our work. Could they somehow prove helpful to us? Never say never, but it seems unlikely, at least for a couple more decades, by which time the field may have moved onto bio- or psychometrics.
Come to think of it, there’s one place I can see us potentially drawing on Statcast, and that’s for range-based fielding stats. Those amazing plots that show the easy to super difficult plays for an outfielder and how many he converted to outs are pretty snazz. That might be a spot for us, but even then, there’s work to be done about how those charts translate to value.
So we apologize if you come to us hoping for utterly state-of-the-art thinking. It’s not our thing—not that we’re big sticks in the mud. We hope, however, that you’ll stick around because you like the retrospective nature of our quest for a better Hall of Fame. And maybe some of our dumb jokes. Well, my dumb jokes. Miller’s are never dumb. Thanks, as always, for reading.