Defensive stats and 8-track tapes

About 15 years ago, a film editor named Dean Goodhill invented a new motion picture format that was a vast improvement on existing film technology. Using conventional 35-millimeter film stock, the new format, dubbed Maxivision, boasted much higher resolution and a smoother image, thanks in part to a doubling of the standard 24fps frame rate. Maxivision projectors promised to reduce film wear and scratches, too. Among the format's champions were Martin Scorsese and Roger Ebert.

Maxivision never really took off, because it was almost immediately supplanted by digital video projection, now found in nearly every multiplex in the Western Hemisphere. Like vinyl records, Maxivision still has its adherents, but its moment, if it ever had one, has long since passed.

This phenomenon, where an existing technology is meaningfully improved just before it becomes obsolete, happens all the time. I call them tweener technologies.

Most of you are probably too young to remember this (hell, I'm too young to remember this), but 8-track tapes had one fatal flaw:

... a mechanical click that occurred four times during the play of an 8-track tape. Most of the time the click would be placed between tracks (songs), but sometimes they would place the click right in the middle of a track. That's right, at some point right in the middle of a song, the volume would fade out, then a moment of silence followed by a loud click-click, then a little more silence and the volume fades back in.

You can hear the click sound here.

When the 8-track format was in its death throes, about to be supplanted by the compact cassette tape, a solution to the clicking problem was discovered. At least, that's what my older brother tells me. Assuming he's not lying,* the click fix came too late to save the 8-track tape. The market for 8-tracks is strictly nostalgia now. Like vinyl, but with more hipster cred.

* One time he told me to always wash my hands with cold water, because it "freezes the germs."

Every so often I read about "smart" traffic lights that monitor approaching cars and adjust to optimize the flow of traffic. This can't come soon enough as far as I'm concerned, but I fear smart lights will be obviated by self-driving cars. That's the frustrating thing about tweener technologies: Though they represent a real improvement to an existing technology, and can be often be brought to bear more quickly or cheaply* than the superseding technology, they don't get implemented or are underutilized because the new new thing is just around the corner.

* Maxivision projectors were much cheaper than early digital-video projectors.

So, what does any of this have to do with baseball? Fielding statistics. The best defensive data available to the public, in my opinion, are the Plus/Minus and Defensive Runs Saved stats from Baseball Info Solutions. You probably know how it works by now: A team of video "stringers" watch every game and record the location and speed of every batted ball, so that each play can be compared to a database of batted balls hit to the same location at the same speed. So let's say Justin Upton catches a hard-hit liner in section "G," 255 feet from home plate (I'm making this up), and let's say major league left fielders make that play 35 percent of the time, Upton would be credited with 0.65 plays made above average (1 minus 0.35). Something like that.

The knock against the BIS data is that it's "subjective"; that the stringers have to make judgments about whether a batted ball was a fly ball or a line drive, and whether it was hit hard or soft or somewhere in between. Another criticism is that it's often difficult to tell precisely where a fielder is standing from a TV broadcast, relative to the bases or other landmarks. As well, Michael Humphreys, in an article for The Hardball Times Baseball Annual 2012, demonstrated that batted balls, as recorded by BIS stringers, tended to cluster around the fielding positions, whereas you'd expect a more or less equal distribution from foul pole to foul pole.

John Dewan, the owner of BIS, addressed these concerns in The Fielding Bible Volume III. The subjective elements -- Was that a liner or fly ball? Was it hit hard or only medium hard? -- have been fixed by the use of a timer, and the problem of balls clustering around the fielders' positions has been solved with better data (including HD broadcasts, which weren't as common when the "clustering" data was amassed), improved techniques, and more experience. This is real progress, I think.

But something even better might already be here. FIELDf/x technology, in which cameras precisely record the position and motion of fielders and batted balls in real time, might someday render Defensive Runs Saved (and its cousin Ultimate Zone Rating) obsolete. So is the new, improved Defensive Runs Saved a tweener technology?

Not just yet. FIELDf/x data, in contradistinction to Defensive Runs Saved, Plus/Minus, and Ultimate Zone Rating, isn't publicly available. Presumably most FIELDf/x clients have the words "Major League Baseball Club" in their name, and that makes all the difference. If DVDs were available only to film industry insiders, Super VHS (a classic tweener!) might have become the dominant consumer video format.

Smart teams (and almost all of them are smart these days) are no doubt making use of FIELDf/x data, but you and I cannot. As long as that's the case, defensive data recorded by video stringers, even with its inherent imprecision, will be relevant and useful, and might therefore avoid -- or at least postpone -- the sad fate of 8-track tapes.