As regular readers know, I was very excited that the Phils would set a very under-rated record this year - the total team plate appearance record. Simply, a team that comes to the plate more has more opportunities to score because it's getting more opportunities per out.
But, alas, as reader RememberthePhitans pointed out this morning, the Phillies fell short. Scoring only 6, 5, 6, 6, 2, and 6 runs in their last six games just wasn't enough to push them over the top. They finished with 6537 total plate appearances. The new top five all-time looks like this:
The top 5:
- 1999 Indians - 6555
- 1996 Red Sox - 6545
- 1976 Reds - 6538
- 2007 Phillies - 6537
- 2003 Red Sox - 6530
But, one thing interesting stands out here. You'd think an all-time total plate appearance record would go along with an all-time OBP record, or at least close. But the Phillies OBP wasn't even the tops in the majors this year. It was .354 compared to .366 for the Yankees and .362 for the Red Sox. On the all-time list, the 1999 Indians had a .374 OBP, twenty points higher than the 2007 Phils.
How did the 2007 Phils get so close to setting the all-time record with a much lower OBP? I had an email conversation with Rany Jazayerli from Baseball Prospectus about this very topic late last week. The back-and-forth is below the fold.
Rany: You would think that total team PAs would correlate strongly with team OBP, as the fewer outs you make, the more chances you get to bat. But (as far as I know) they're not anywhere near the all-time record for OBP, or even the highest OBPs of the 162-game era. Strange.
David: That's what's strange about the stats. The 1999 Indians, the current record holder for plate appearances, had a .374 OBP, whereas the 2007 Phillies are at .355. That's a significant difference, yet the Phillies are on pace to have more plate appearances.
I'd love to see innings at-bat as a stat for this research. Have the Phillies had that many more innings, through extra inning games and/or rain-shortened games for the Indians? What else would explain it? (It's true for this year too - the Phillies have an OBP 9 points lower than the Yankees and 6 points lower than the Red Sox, but they have 100 and 134 more plate appearances, respectively, than those two teams.)
Or, have they gotten on base that many more times by error?
Are there other explanations?
Rany: Wow, you've got me there. I actually thought you might have made a mistake somewhere, because it's hard for me to figure out how two teams with OBPs 17 points apart (bb-ref.com has the Indians at .372) can have the same number of plate appearances. But you're right; the Phillies are on pace to break the record.
I thought it might have something to do with sacrifice hits, which don't count against OBP, but they're pretty similar in that regard; the Phillies have 63, the Indians had 54.
The biggest factor I can find is that the Phillies are ridiculously - as you probably know, historically - successful at stealing bases, with 134 SB vs. 18 CS. The Indians were 147-50, which means the Phillies have made 32 fewer "avoidable" outs. That's not trivial; that's another 50 or so PA right there.
The 1999 Indians threw 1450.1 IP; the Phillies are on pace for 1458.1. That's another 8 IP, or about 35 PA.
The 1999 Indians were 47-34 at home; the 2007 Phillies are 43-32 with 6 to play. A team that wins most of its home games will bat fewer times in the bottom of the 9th, but there's not much of a discrepancy there.
Using a simplified estimate of PA=(Outs)/(1-OBP), and an estimate of 4150 outs over the course of a season, then the Indians should have 6608 PA, the Phillies should have 6434. That's 174 PA to make up. They get 50 back on the bases, and 35 back from playing more innings, but that's still 90 PA I can't account for.
Well, they'll probably finish with about 10 fewer GIDP than the Indians. So that's 80.
I guess if we parsed the data, we'd find that the Phillies have more than their fair share of ROEs, while doing as good a job of avoiding outs on the bases when taking the extra base as they have when stealing one. But I really have no idea.
David: So, all in all, we have the following stats that contribute to a misaligned OBP and TPA:
- sacrifice hits
- caught stealings
- home wins prior to 9th inning
- extra innings
- double/tripe plays
- outs on the basepaths
- reached on errors
- anything we're missing?
With the Phillies, they're a NL team, so naturally they have more sacrifices. Their stolen base percentage is historically the best ever, which would indicate they're probably also better at avoiding double plays and outs on the basepaths (although not necessarily so). They're in the top third this year in innings pitched (8 more than the Yankees and 20 more than the Red Sox, the two teams with higher OBPs this year than them). I believe they have the most come-from-behind wins in the majors this year, which would indicate fewer lost 9th innings at home (although probably only a few such games making a difference). As for reaching on errors, intuitively, speed should correlate with that, but I seem to recall a BP article from years ago saying it doesn't. The BP stat page has the info though, and the Phils are 3rd in the majors with 68 ROE, compared to 64 for the Yankees and 53 for the Red Sox.
So, basically, all of these factors that give a team a higher TPA than their OBP would otherwise indicate are in the Phillies favor this year.