Logan's Run

It's Nook Logan Week. He's a headliner during the Nats Caravan, MLB.com heralds him as "a keeper," and the Post notes Nook has caught the eye of "Baseball Men":

I'll confess I'm curious if the "Baseball men" calculated this figure; my inclination would be that they didn't, seeing as doing so would likely require a computer, which every true baseball man eschews. But someone calculated it, and it's got my attention. Some questions:

• Is it true?
  
• Is it meaningful?
  
• Is it significant?

Sixty-four one-hundredths of a hit per nine innings: the author of the Post's article desires it to sound pretty enormous, and I'm happy to oblige. It is enormous. Last season, Washington pitchers hurled 1436.1 innings. That 0.64 hits per nine innings reduction would account for 102 fewer hits every 1436.1 innings. According to linear weights, a single is (generally speaking) worth 0.47 runs. That means, knowingly or unknowingly, the article indicates the quality of Logan's defense would save 48 runs over a 1436.1-inning season. At the rate of one win per 10 runs saved, that means Logan would account for almost five additional wins if he played center like he played center in September. Enormous, indeed.

Is the 0.64 reduction per nine innings true? I have no clue. But it's a clunky way of looking at the issue. It counts a home run as a hit, even though there's no shot any fielder (with the most limited of exceptions) could have prevented it; additionally, it counts a strikeout as an out, even though no outfielder could have contributed to it. How about something less clunky?

I've figured the batting average on balls in play ("BABIP") against the Nats for every game, reduced down to each game started by each centerfielder. This could not possibly have been done but for Baseball Musings' Play-by-Play Database, which is pretty amazing. And slow. But mainly amazing.

Just so we're clear, BABIP = (Hits - Home Runs)/(At-Bats - Strikeouts - Home Runs). And, just so anyone skeptical can repeat this same study if he or she so desires, I might as well list the games started by each of the Nats' centerfielders in 2006:

• Watson: 4/3-4/8, 4/11-4/12.
• Byrd: 4/9-4/10, 4/13, 4/16, 4/25, 4/29-5/5, 5/7-5/12, 5/17-5/18, 5/28-6/3, 6/9-6/14, 6/24-6/30, 7/4-7/5.
• Church: 4/14, 4/18-4/24, 4/26-4/28, 5/6, 5/13-5/16, 5/19, 7/23-8/8, 8/10, 8/13, 8/17, 8/21-8/22, 8/26-8/30, 9/2 (first game), 9/24, 9/30-10/1.
• Jackson: 4/15, 5/22-5/27, 6/4-6/8, 6/15, 6/21-6/23.
• Anderson: 7/1-7/3, 7/6, 8/31.
• Escobar 5/20-5/21, 7/17, 7/19-7/21, 8/9, 8/11-8/12, 8/14-8/16, 8/18-8/20, 8/23-8/25.
• Kearns: 7/14-7/16, 7/18.
• Matos: 7/22.
• Logan: 9/2 (second game)-9/23, 9/25-9/29.

Using the Day-by-Day Database, I calculated the BABIP for all of those dates. (Speaking of which, I would have greatly appreciated had Frank Robinson not switched centerfielders from game to game on so many occasions.) Results for each player follow, highest BABIP to lowest:

Is it true the Nats allowed fewer hits (on balls in play) with Logan in center? Yes. Marlon Byrd played the most games there, and the opposition's BABIP with Logan out there was about 35 points lower. Ryan Church played the second-most, and the BABIP difference was about 15 points.

Is it true the Nats had some sort of hit prevention superiority with Logan in center? Not really. Logan played the third-most games at the position (25). The Nats had virtually the same opposition BABIP with the fourth-most-used centerfielder, Alex Escobar (23 games), out there. This isn't to say Logan's not a very good centerfielder; for all I know, he might be the best ever. It's just the difference between him and the other centerfielders trotted out there---as measured by opposition BABIP---was, generally speaking, not dramatically lower. This is all a roundabout way of saying Marlon Byrd deserves as much discredit as Logan deserves credit.

* * * *

Of course, this is all premised on the basis that a centerfielder has a meaningful role in hit-prevention or BABIP-lowering. More specifically, we return to the seeming assertion that Logan particularly accounted for a reduction in hits allowed by Washington's pitching staff. Baseball men might believe that, and they might be right; I'm not at all saying they're not, as I respect the opinions of baseball men (and even Baseball Men). They know more than me about baseball, quite naturally. But this tossing around of hits allowed stats is about as tidy as a drunk tossing about his dinner.

Earlier, Capitol Punishment inserted some possible context. Lest one believe a centerfielder alone---or primarily---is responsible for hit prevention, I'd offer it's not quite that clean. It's rather more noisy, and when the defensive personnel changes, how do you account for that? Moreover, when the composition of the pitching staff changes, what happens? I'm thinking specifically of Livan Hernandez, who was dealt before Logan's arrival. The guy surrendered 176 hits in 147 innings for the Nats, and then the ratio was about even upon his trade to Arizona. Was the D-Backs' defense possibly that much better? If so, was the difference so intertwined with the quality of play in center?

* * * *

This leads me to the final consideration: Is any kind of hit prevention stat for the Nats with Logan in the lineup (hits per nine innings or BABIP against) in any way significant? He played little more than three weeks in center. I dare you to take a player listed above, plug one week he played into the Day-by-Day Database, then plug in another week he played. Try this with Jackson, or Byrd, or Church, or whomever. You'll find quite a difference. At times, the differences seem---if you'll pardon the express---random.

When you've got so many players splitting time so tightly, even individual pitcher performance can make a tremendous difference. Pedro Astacio pitched an inexplicable two-hit shutout on August 15. Who was the centerfielder? Escobar. This was one of his 23 appearances in center. You think that helped the opposition BABIP in games he manned center?

What about Beltran Perez, who joined the staff around the same time Logan joined the club? Perez made three starts for the Nats. In the first two, Logan manned center; these games comprised nearly ten percent of Logan's appearances in center. Perez surrendered all of five hits in 12.1 innings (including a six-inning one-hitter). In his final start, with Church in center, Perez got bombed, allowing six runs on six hits in three innings. What if fate had switched things around and the Perez clunker occurred while Logan was in center? That neat little hits per nine innings stat would have gone up, right? Or did the particular centerfielder on those days mean so much?

* * * *

Please understand I'm not in the least demeaning Logan's ability to play center, and it's certainly possible he's excellent out there. There's an argument to be made that he should open as the regular centerfielder---maybe not a winning argument, but the argument does exist. I'd just rather the argument be well-constructed. Heck, I'd take a bald assertion over barren facts.