# When players delve into advanced baseball stats

For decades fans have grasped the need for advanced analysis of the numbers behind baseball in hopes of better understanding the game. Now, there are three young pitchers who have come out and said that they use saber-esque stats in analyzing baseball, and sometimes their own game: Brian Bannister, Ross Ohlendorf, and Max Scherzer. Their similar backgrounds are intriguing (all close-aged, right-handed starting pitchers with college educations), as is the way in which they go about their analysis. However, another interesting theme is how the media has gone about reporting and portraying these players.

Brian Bannister has to be the leader of the group thus far. He made friends around the sabermetric community when he said last week that, “If Bill James had a 90-mph fastball, he’d be me…I think people universally agree — in the sabermetric community and the fan community and in the media community — that sabermetrics are effective at identifying successful baseball players and ways to win this game.” That’s a pretty amazing, and downright gutsy statement from a baseball player considering the atmosphere of clubhouses. While advanced analysis has permeated into front offices, there doesn’t seem to be much talk of WAR or wOBA around the dugout water coolers. Bannister, the son of former major leaguer Floyd Bannister, attended the University of Southern California and also runs a photography studio in Phoenix. He claims he’s been successful in using advanced stats to improve his game, including getting more grounders; the numbers verify his claims, as his GB% has gone from 37.5% last year to 51% this year.

However, it’s been interesting to see how the media has portrayed his number-loving compared to the people who have really put the work behind the scenes (e.g. some of the staff here, including the great Colin Wyers and Dave Studeman). Scott Lauber of the Courier Post Online did a piece on Bannister’s use of xFIP, and the exaggeration involved in his following statement is too much:

Bannister, 7-8 with a 3.97 ERA in 21 starts for the Royals, conceded his favorite statistic is “really technical,” and the formula for computing xFIP is so complex that it seemingly requires Einstein-like knowledge of higher mathematics.

Einstein-like knowledge of higher mathematics? Really? Here’s the formula for xFIP: ((FB*.11)*13+(BB+HBP-IBB)*3-K*2)/IP. While it may be more complex than other traditional stats, one does not need any higher math degree to figure out how it works. But Lauber isn’t alone. Bob Dutton at KansasCity.com also reported on Bannister and again thought his understanding of xFIP made him a genius:

“I think the ultimate stat for a pitcher is xFIP,” Bannister said before pausing and offering a wry grin. “I know that’s getting really technical. It’s fielder independent pitching adjusted for your home-run rate back to the league average…” He’s right. It is technical…It could be that xFIP was what Matt Damon scratched out on the blackboard in “Good Will Hunting.”

While a hyperbole in written media isn’t exactly a rarity, it certainly is ironic considering the baseball media’s portrayal of the inventors of these numbers as geeks in their mom’s basements. This can be seen in an article Dutton wrote just a few days ago when he channelled Murray Chass by calling xFIP and BABIP, “new-age baseball jargon.” He wasn’t saying it in a demeaning fashion (I don’t think), but the subconscious framework was laid down; any non-traditional stat is weird.

Ross Ohlendorf has been getting similar treatment ever since Tim Kurkjian’s piece on him. Ohlendorf, who is certainly an extremely bright guy (Princeton grad who knows his math), put together a model of determining the value of draft picks for his thesis in college. And while I’m sure the paper is certainly interesting, Kurkjian goes a little overboard when he says that, “The 126-page thesis is brilliantly written and so complex, only a mathematician would be able to completely comprehend its meaning.” I certainly hope (and believe) that that’s not true. If only a mathematician could understand it, something has to be wrong. Ohlendorf’s study shouldn’t have involved too much calculus or string theory.

But moreover, your average fan is doing stuff like this all the time. In fact, Sky Andrecheck at Baseball Analysts did a very similar study a few months ago, as has The Hardball Times’ very own Victor Wang. While their work is fantastic, you don’t have to be Will Hunting to understand it, which is certainly a good thing; overly complex work isn’t good.

Finally, and briefly, is Max Scherzer. Scherzer, who conducted a great Q&A last year with Eric Seidman over at Baseball Prospectus, recently said that he “values the pitching statistics that take fielding out of the equation and recently has become particularly interested in a stat called tERA, which assigns values to every batted ball based on trajectory, velocity and location.” That, my friends, is simply awesome.

Although the tone of this article may have seemed a little critical of the media, I have to say that there is nothing better than hearing MLB players talk about the stats we use and love. While it may be funny to imagine Scherzer or Bannister logging into Fangraphs or The Hardball Times, the truth is that it is happening, and while the media may still be slow to embrace advanced analysis, players entering the picture will definitely speed up the process. So when we hear about Bannister pouring over his ground ball numbers, or Stanford grad/Cubs outfielder/former Stats Inc. intern Sam Fuld talking about economics, let’s give a hand to the guys and girls behind the scenes who are invaluable at crunching the numbers. They are making waves, one player at a time.

Print This Post

Nice job, Pat. But xFIP is a little more technical than you have it. The home run part is normalized to league average, and a league-specific constant is added to make the entire equation add up to the league’s ERA.

Awesome read, Pat.

Thanks, Dave and Dan. Oh, and thanks again Dave, as I changed the formula because I accidentally had FIP instead of xFIP.

Dave, isn’t xFIP also normalized to park HR/FB rate?

Anyway, nice piece Pat. This was my exact reaction when I read the ESPN piece on Olendorf.

Ohlendorf also wrote some sort of thesis report on baseball, I believe it might have been on the draft. I have tried to locate it but have not had any luck thus far. Anybody seen it before? Or know where to get a copy?

Sorry, my webpage only showed only the first part of the article so I didn’t know it discussed Ohlendorf’s thesis later. Still would like to read it myself.

Also, to give some perspective, not everyone thinks that algebra is easy. So even though these reporters undoubtedly attended college, for them college algebra is to them as quantum physics is to most of those who are into sabermetrics. We got into saber because we loved math, but not everyone gets math the way we do.

To get a copy of Ohlendorf’s thesis, you should contact the Princeton University Archives at the Seeley G. Mudd Manuscript Library, 65 Olden Street, Princeton, NJ 08540, phone 609-258-6345, fax 609-258-3385, or send email to

to locate a thesis of interest.

Here is the link to follow….

http://libweb5.princeton.edu/theses/index.htm

THESIS NO.: 19720

TITLE: Investing in Prospects: A Look at the Financial Successes of Major League Baseball Rule IV Drafts from 1989 to 1993 (140 pages).

AUTHOR: Ohlendorf, Curtis Ross (2006), Operations Research and Financial Engineering

ADVISOR: Carmona, Rene A.

LOCATED AT: MUDD

“So even though these reporters undoubtedly attended college, for them college algebra is to them as quantum physics is to most of those who are into sabermetrics. We got into saber because we loved math, but not everyone gets math the way we do.”

True, I’m sure, but we’re not even talking about college algebra. Solving for xFIP (once you have the constants—and those you can approximate even if you don’t) isn’t any more difficult than finding OBP or SLG.

Some mainstream columnists have their issues with OBP, but I don’t believe the inscrutability of its calculation is one of them.

Good article, Pat.

Any info on the Shcerzer referenced tERA? I did a google search here at work and had to close the window quickly – I got a ton of pictures of tERA Patrick – and I didn’t even know she could pitch

Bigtoe- tERA is just Statcorner’s tRA adjusted for earned runs instead of just runs.Rule of thumb is to just multiply tRA by .92 to get tERA, as about 92 percent of all runs are earned