May 17, 2009

Prospectus Idol Entry

Ken Funck's Initial Submission

by Ken Funck

Printer-friendly

Contact Author

Bio: My name is Ken Funck, and I live with my wife and two children in Madison, Wisconsin, a wonderful place I didn't have the heart to leave after getting my English/Creative Writing degree from the University of Wisconsin. I'm a third-generation fan of the Chicago Cubs, but I'm not one of those Cubs fans; I keep score when I go to Wrigley, and I know who Mick Kelleher is. I've never written about baseball professionally, I don't blog, and my journalism experience is limited to (a) selling story ideas to The Onion for $10 a pop in the summer of 1994; (b) editing my high school newspaper; (c) composing analytical pieces for my Strat league; and (d) writing reactionary letters to the editor of our local alternative newsweekly under the pseudonym Alvin P. Blatherson, Sr. (a serial comedy that I like to call "Blather, Rinse, Repeat").

So, why do I think I have the chops to win the BP Idol contest? First of all, during the day I manage large-scale data warehouse applications for the State of Wisconsin, so I know which end of a statistical database is up. Secondly, I'm a long-time BP subscriber, so I have instant access to the sharpest, most insightful and entertaining baseball analysis in the business. And most importantly, I may be the only entrant who has actually already won a BP writing competition: Derek Jacques' 2008 Rock The Vote contest. That submission, a wordy treatise on a faux HoF candidacy metric called CHEERS, managed to win me an autographed copy of BP2K8. Derek was also kind enough to forward me a fan letter sent to the BP Inbox, a goosebump-raising experience I think you can all identify with. I'm more of a generalist than a specialist, so the concept of writing something clever and unique each week on a topic of your choosing is right in my wheelhouse-in Idol terms, I guess that means I'd be comfortable during both "Country Week" and "Pretend I Care About Quentin Tarantino's Performance Tips Week." My attached submission is pretty representative of my writing style, which I believe matches well with BP's general tone-but that's up to you to decide.

Entry: Seeing Is Believing

When Gary Huckabay boldly declared that baseball analysis is dead, I suspect he hoped to never again see a chart like this:

Most Patient      P/PA  OBP  SLG  OPS   TGF
Nick Swisher      4.53 .332 .410  742  .638
Jack Cust         4.38 .375 .476  851  .618
Adam Dunn         4.32 .386 .513  899  .627
Jason Giambi      4.30 .373 .502  875  .641
Kosuke Fukudome   4.29 .359 .379  738  .630
Bobby Abreu       4.29 .371 .471  842  .644
Curtis Granderson 4.26 .365 .494  859  .638
Mark Reynolds     4.24 .320 .458  778  .619
Mike Cameron      4.24 .331 .477  808  .586
Fred Lewis        4.23 .351 .440  791  .606
Top 10 Avg.       4.31 .356 .462  818  .625

Least Patient        P/PA   OBP  SLG  OPS  TGF
Yuniesky Betancourt  3.15  .300 .392  692 .592
Bengie Molina        3.18  .322 .445  767 .625
Cristian Guzman      3.21  .345 .440  785 .616
Alexei Ramirez       3.27  .317 .475  792 .541
A.J. Pierzynski      3.34  .312 .416  728 .588
Robinson Cano        3.35  .305 .410  715 .548
Vladimir Guerrero    3.37  .365 .521  886 .580
Garret Anderson      3.43  .325 .433  758 .606
Carlos Gomez         3.44  .296 .360  656 .592
Corey Hart           3.47  .300 .459  759 .580
Bottom 10 Avg.       3.32  .319 .435  754 .587

Here we see the most and least patient hitters in the 2008 MLB season, as measured by pitches seen per plate appearance. There's nothing groundbreaking here - the correlation between plate discipline and performance is well-established, so no one should be surprised that the Jobian approach exhibited by those in the left-hand column resulted in better rate stats than the Plastic Hacktastic Band on the right. Anyone with a browser, a spreadsheet and the dexterity required to hit Control-C could have assembled this chart, which was exactly Gary's point: after years of slicing and dicing the same ingredients, analysts have teased all the flavor out of historical player performance data, leaving only a nasty gruel.

What's needed to spice things up is data from outside the normal realm of performance, like that found in the final column of the chart above: TGF. TGF isn't based on standard performance metrics, but it correlates quite nicely to batter performance - 8 of the top 9 TGF scores in our (admittedly small) sample belong to patient, productive hitters. And best of all, TGF fits neatly into one of the categories of analysis Gary excluded from his obituary: integration of scouting data with performance data. Because TGF is actually a statistical representation of that most hackneyed of scouting clichés: The Good Face.

For decades, many baseball scouts have claimed the mystical ability to identify certain characteristics desirable in a young player by merely studying their facial features. The sabermetrically inclined, especially those outside the industry, have been known to scoff at this claim as just another discredited belief to which "baseball men" cling in a vain attempt to protect their oracular role in an organization. Not only was it impossible to quantify The Good Face, scouts didn't even seem able to agree on a standard definition, throwing around generic terms like "strength," "virility," "maturity," "determination," and "square-jawed confidence." The most unintentionally revealing description comes, perhaps not surprisingly, from Al Campanis: "I never used to sign a boy unless I could look in his face and see what I wanted to see: drive, determination, maturity, whatever." Huh? See what you wanted to see? Any description that ends with the word "whatever" is by definition both subjective and nebulous. Yet baseball men insisted that they could provide an accurate, objective measurement - but unlike stopwatches or radar guns (famously disdained by the old-timey), the only instrument that could measure The Good Face was available to just a select few: the eye of a veteran baseball scout.

The idea that a prospect's future could be read on his face strikes many as arrogant, deterministic hogwash. And yet … most of us make similar judgments every day, often based on far less evidence or experience than baseball scouts have available to them. Who hasn't shied away from the shifty-eyed used car salesman, or sought out the doctor with a caring face? Or drafted a guy in a fantasy baseball league because he "looks like a ballplayer?" The first time I saw Ryan Braun step up to the plate, I noticed something about his eyes - they seemed preternaturally focused, hyper-alert. He'd step one foot out of the box between pitches and take a deep breath, widening his eyes as he glanced around, and when he stepped back in his entire presence radiated … okay, I'll say it. Determination. Square-jawed confidence. Whatever. When I first saw Ryan Braun, I saw The Good Face.

You can find The Good Face in the dictionary, too. It's hiding there under an assumed name: "Physiognomy, the art of discovering temperament and character from outward appearance." A subject of intense debate ever since the Greeks first started associating an individual's traits with those of the animal they most closely resembled, Merriam-Webster cleverly sidesteps the "science or pseudo-science" argument by calling physiognomy an art. Scouts would probably agree with that assessment - art requires an artist (i.e., the scout), whereas science requires a testable, repeatable process. Physiognomy had long fallen into disrepute, likely due to its association with shakier or more insidious fields of study such as phrenology and eugenics. But recently physiognomy has been a popular, promising area of research and is likely to receive some votes for Comeback Discipline of the Year. So why not try and apply the same facial analysis techniques used to predict creditworthiness and business success to the problem of predicting player performance?

Happily, one recent study links an oft-desired athletic trait (aggressiveness) to a specific, measurable facial feature. The researchers are from Canada (where The Good Face looks like this), so the study naturally involves hockey and seems to indicate that players with higher facial width-to-height ratios tend to spend more time in the penalty box - the study's proxy for aggressive behavior - than those with lower width-to-height ratios. So if we assume that "aggressiveness" is a desirable characteristic in a baseball player, akin to "strength," "virility," "determination," and "whatever," there would seem to be an easy way to provide an accurate, objective measurement of a player's likelihood to possess that desirable characteristic: obtain the player's facial measurements.

And that's exactly what the TGF metric in the chart above represents: the facial width-to-height ratio for the 10 "most aggressive" and 10 "least aggressive" hitters in MLB. Obviously my calculation method for this isn't optimal. Since I'm just some random guy and the Angels aren't about to let me walk up to Vlad Guerrero and stick a ruler in his face, I had to rely on photographs of each player which I then cropped and measured to come up with the appropriate ratio. Rigor, schmigor - it's good enough for a quick thumbnail sketch, sample size be damned.

The interesting thing about TGF in the chart above, beyond the fact that it actually seems as if it might correlate to pitches per plate appearance, is the fact that high TGF (which indicates greater aggression) coincides with seeing more pitches per at bat, not fewer. Players that put balls in play earlier in the count, and don't walk much, are traditionally described as "aggressive" hitters. But TGF seems to be saying the opposite -- aggressive hitters actually work deeper counts, relying on a more controlled aggression to wait for a pitch they can really attack. Perhaps this means players with lower TGF lack the "square-jawed confidence" to take a less hittable pitch for a strike because they're worried they won't see a more hittable pitch later. Perhaps they swing out of fear, not aggression. Or perhaps I'm just talking out of my … heinder, tuchas, whatever.

If the TGF metric does in fact correlate to better plate discipline, and thus greater production, maybe The Good Face actually does exist. Maybe TGF is one component of it, and maybe scouts have a finely-honed sense of proportion which registers just by looking at a kid. Or maybe not. But as for Ryan Braun, his TGF score is 0.696 - higher by far than anyone else in the chart. Sometimes you really should just believe what you see.

Ken Funck is an author of Baseball Prospectus. Follow @KenFunck
Click here to see Ken's other articles. You can contact Ken by clicking here