2007’s Most Historic and Impressive Team Splitsby Chris Jaffe
March 03, 2008
Over the last several months, I've raided the team splits information over at Baseball-Reference. For whatever sick and sorry reasons of my own, I decided to find team splits that interested me and download them into excel for every single team for which the site provides information.
Turns out, it's quite a lot of info. He has splits up for every team from 1957-2007—1,270 squads in all. And in this bizarre quest, I was like a kid in a candy store, taking an absurd variety of splits out—147 in all. (See references and resources at the end of this piece for a list of them.)
One nice result of this uber-nerdy activity: I can look up 2007 team splits, both pitching and hitting, and find out which were the most historically impressive. Anyone can say "Man, Baltimore sure didn't pitch well in September," but wouldn't it be nice to know exactly how they compared with every other squad from the last century?
Well, now you can find out the answer to that and other mysteries. And here the dozen team splits from 2007 that jumped out at me as the most historically impressive:
1) Maybe Baltimore should hire Leo Mazzo—oh, wait. Last year, from September 1 onward, the Baltimore Orioles posted a team ERA of 6.89, which is the single worst September/October ERA baseball in the last half century. It's almost certainly the worst since the 1930s.
And they didn't just sneak over the border with this one—the second worst score was 6.55. Among all months in which a team played at least 20 games, it's the seventh-highest ERA. (Worst of them all? April 1994, by the Minnesota Twins, with an ERA of 7.34). Most months topping Baltimore are mid-summer ones, so factor in the game's annual temperature curve, and it's even worse for them.
They allowed the most earned runs in September, the sixth-most overall runs, had the second-worst opponent on-base percentage, and allowed the eighth-most walks. Their ERA wasn't a fluke.
Actually, this understates the problem, because in late August they lost to the Rangers in historic fashion, 30-3 (helping them get the second-worst August ERA at 6.65). From that point onward, they played 39 games with a mortifying ERA of 7.61. I can find no precedent for a team pitching that poorly for that long a stretch.
What went wrong? Simple—Erik Bedard got injured and the rest of the team wasn't that good. Now the Orioles have traded him for prospects. Enjoy 2008, Baltimore.
2) Even still, maybe Tampa should call Mazzone. From 1957 to 2006, no bullpen posted an ERA over 6.00.
Enter the Devil Rays. Last year, they blasted the 6.00 barrier with an ungainly ERA of 6.16. They were all-around bad, allowing opposing hitters to tee off on them with a .303/.383/.493 line, for an .876 OPS, only one point behind the Coors-inflated 1999 Rockies for worst opponent OPS ever yielded by a bullpen.
I suppose if you look back in time, you can find a bullpen from the 1930s with worse numbers, but then again relievers threw fewer innings. This might be the most damaging bullpen ever to a team. I know there's an era adjustment that should be made with the current Silly Ball Era and before, but even in 1990 relievers didn't throw as much as they now do.
Fun fact: prior to this year, the 1996 Tigers had the worst bullpen ERA at 5.97. Though they lost that marker, they still have the worst starter ERA, at 6.64. Man, that team could not pitch!
3) Now that you mention it, I never have seen Pete Gray and Josh Bard at the same time together... This one really isn't a split (well, technically it is), but the info is so stunning, I have to include it.
The split in question is how pitchers do with base runners on. The stats under examination are opponent stolen bases and caught stealings. Obviously, they can't steal bases unless someone's on, so it's the full sample size, but ah well...
Looking at ability to throw out runners [SB/(SB+CS)], last year, the San Diego Padres were historically dreadful.
Actually, that's an understatement. They made historically dreadful teams look good. Runners stole 189 bases while only getting caught 20 times for a 90.4% success rate. That's...something.
For perspective, the second-worst success rate is 85.7%. Let's stop here a second—when you have 1,270 teams in a sample size, the difference between first and second shouldn't be 5%. The gap between them is greater than the gap between No. 2 and No. 25. Folks, I wasn't kidding when I said they made historically dreadful teams look good. Bard wasn't the only catcher, but he was their main one, and thieves succeeded 92% of the time against him.
Oh, and that second-worst team in the last half century? Would you believe it was the 2006 Padres? (Shakes head.) The Padres acquired Bard that year, and he played 93 games for them. Last year, it was 118.
Let me put it to you this way—those 189 stolen bases are the 18th-most allowed by any team. The 20 caught steals are the 1,268th-most. That's a bad combo. (The 1963 Yanks had 18 CS and that year's Red Sox had 17).
Actually, there was something going around last year. The third-worst total ever came from the 2007 Blue Jays, and the fourth-worst from the 2007 Cubs (who traded catchers with the Padres midseason). Teams are really picking their spots when to run. But if I'm a National League manager, I'm running every time Josh Bard is catching.
4) I guess we know what isn't the new market inefficiency. Billy Beane's Oakland A's had two notable splits in 2007.
They tied a record, as their right-handed pitchers picked off exactly one base runner all year. Not the most important factoid you'll ever come across, but rather impressive nonetheless. You'd think runners would at least fall over twice for them.
Second, their pinch hitters are really horrible—have been for two years. In the team splits, Baseball-Reference has a stat called sOPS+. It's like normal OPS+, only instead of comparing to the league-wide OPS, it's compared to just the league-wide OPS+ for the numbers generated at that specific split. So, for example, a team's OPS for its pinch hitters would be compared to league-wide pinch-hitting stats (adjusted for park).
The 2007 A's had the 12th-worst pinch hitter sOPS+ since 1957. Their hitters posted an anemic .170/.250/.170 line. In 60 plate appearances, they didn't get a single extra-base hit. Mind you, that's progress for them. In 2006, they had the eighth-worst pinch hitter sOPS+ and a line of .114/.227/.171.
In the last two years, their pinch hitters have driven in only 14 runs. In contrast, the 2007 Phillies tied for the fourth-most RBIs by pinch hitters in the last half century with 56.
5) Winning with one hand tied behind their backs. The 2007 Yanks and Cubs both made the playoffs despite both having historically one-sided offenses.
The Yanks set several records for left-handed batting in the years under examination: most plate appearances by southpaw hitters (4059), most hits (1017—the first team to crack the Dee Fondy barrier), most runs (599), most RBIs (573), and most total bases (1622). They also had the second-most doubles (208, one behind the 2001 Twins), and 13th-most walks. Looking at their offense, it certainly makes sense.
The 2007 Cubs issue wasn't so much quantity; it was quality. They had an historically large split between their righty and lefty bats. The former posted an impressive sOPS+ of 112, while the latter floundered at a replacement-level mark of 74.
That 38-point edge in sOPS+ for right handers is the seventh-greatest in the last half century. In two-thirds of the team's at-bats, righties went .283/.345/.453 while their counterparts poked an ineffectual .248/.309/.358. Righties had an OPS similar to Miguel Tejeda's 2007 while the southpaws hit as badly as Craig Biggio did in his swan song.
Could be worse though—the 2005 Cubs had the fifth-biggest gap favoring righties.
6) Unlikeliest road warriors ever Last year, baseball witnessed its highest-ever BABIP in road games. It came in the AL. You're probably thinking it's a team from the East, right? After all, that's where the powerful Yankee and BoSox squads are.
Well, it's in that division, but not those teams. With a BABIP mark of .334, history was made by the Tampa Bay Devil Rays.
I have no explanation. They just did a great job getting hits off of balls in play. Who knew?
Just imagine how good their BABIP would've been if they could've faced their own defense!
7) These guys can't ever pace themselves. Over the last four years, the Tigers have done the worst jobs pacing themselves over the full season. If you compare first-half winning percentage to second half, they are the worst at maintaining their early performance for the full season.
So I suppose it's all too appropriate that they are historically dreadful at pacing themselves during ballgames. Baseball-Reference gives out offensive splits for scoring in the early innings (frames 1-3), middle innings (4-6), late (7-9) and extra innings.
Last year, the Tigers scored only 22.0% of their runs in the late innings, the second-worst mark of the 1,270 teams. Only the 1989 Cubs did worse. On the bright side, they did score the eighth-most runs ever in the first inning.
8) Doing what you're supposed to do—boy, and how! Last year, the Dodgers pitching staff did a nearly unprecedented performance beating down opposing pitchers.
Opposing hurlers hit an anemic .104/.133/.108 against them for an OPS of 241, the fourth-lowest ever. (And given how pitcher hitting has steadily declined over the decades, it probably really is the fourth-lowest in all history, not just since 1957.)
The best part? Did you notice that .104 batting average and .108 slugging percentage? Why yes, that isolated average of .004 is the lowest ever. In 352 plate appearances, they allowed one stinking double. That's it.
The slugging percentage is the second-lowest ever. The 1972 Orioles held other pitchers to a .106. It's also the 14th-lowest OBP. Sure, you're supposed to get the pitcher out, but one stinking double in 162 games? Damn!
9) No country for fly ball pitchers. The 2007 NL was not a pleasant place for hurlers who let the ball enter the air, and the AL was scarcely more forgiving.
The 2007 Brewers became the first team to ever hit over 100 homers against fly ball pitchers, with 108.
The second-most homers ever hit against airborne specialists is 97, from the 2007 Phillies.
In fifth place are the 2007 Marlins. Right behind them are the Reds, also from last year. And rounding out the top 10 are the Mets. That's one-third of the NL from one year making up the top 10 among 1,270 teams. Meanwhile, last year's Yanks and Tigers ranked eighth and 11th respectively.
Prior to last year, no league had ever seen 1,000 homers hit off of fly guys. Last year, 1,221 occurred in the NL.
Explanation? I'm sure the new round of parks helps, but they didn't all open last year, so that doesn't explain 2007's 20% increase. Maybe there were just more fly ball pitchers in general, leading to higher counting stats.
Well, that's easy to check. Add up all plate appearances against fly guys, and figure out HRA%. By this metric, the most homer-happy league was the 2007 NL, with the 2007 AL coming in second. The Brewers are "only" the second-best percentage (behind the 2006 White Sox), while five of the top 19 are 2007 teams.
Milwaukee also had the second-best isolated average against them, an especially impressive achievement given that the rest of the top nine all came in designated hitter leagues.
I dunno...hitters just went shitass crazy on them.
10) Manic Manny Acta. New Nationals manager Manny Acta won accolades by raising his team, with its all Quadruple-A starting rotation, up to the heights of craptitude. Instead of being historically dreadful, as some predicted, they just weren't very good.
One key point was his willingness to go to his bullpen. The 2007 Nationals tied a record set the year before by the Cubs by using pitchers on consecutive days 101 times. While the Cubs win a tie-breaker (they once used a pitcher on both ends of a double header, something the Nationals never did), Washington's experience may have been more impressive.
The Nationals used pitchers on no rest for 183 of their innings, shattering the previous mark of 168. I have no idea what it means for their future, but that's one thing that got them through 2007.
11) The Angels way. In June, Anaheim set an offensive record perfectly keeping with the contact-happy philosophy promoted by manager Mike Scioscia. They posted a team-wide batting average of .319 for that month, the best June bloom since...well, probably since the 1930s.
In months at least 20 games have been played, only two squads had ever done better since 1957. The 1998 Rangers went .322 in April, and the 1980 Kansas City Bretts hit an electrifying .332 in July.
Improbably, the Angels' mark was perfectly tied the next month, as the Yankees scored the exact same batting average of .31944444.
Whereas the Angels could only parlay their performance into the 23rd-best OBP and 70th-best runs per game in June-dom, the Yanks had the best OBP in all July (even better than the 1980 Royals), and third-best runs/game.
12) Abandon all hope, ye who enter. Last, but not least, the Giants threatened a mark for offensive futility last year. Not very surprising, is it?
Why the league put fly ball pitchers on the run, the Giants provided safe haven for ground ball pitchers. In 1,970 PA against the ground guys, the Giants only hit 19 homers. That ratio of one homer every 104 plate appearances is the fifth-worst since 1957. Imagine how bad it would be without Barry Bonds.
I know I said I'd only list a dozen, but there's just one more I want to briefly cover in this baker's dozen.
13) The Jim Palmer Seal of Approval. Last year, the Nationals pitchers did not allow a single grand slam. That's nice, but by itself is by no means historic. Many other staffs can make the same claim, including the 2007 Indians.
However, one key fact separates DC from the rest. Those other staffs all had no more than 170 occasions when they faced a hitter with the bases loaded. The Nationals did it 198 times. For the record, over 200 teams in the last 51 years have had more than 170 times that they pitched with the bases loaded. The Nationals, who are obviously nowhere near that 170/171 border, are the only ones to survive that experience unscathed.
There are a few others I could toss in. The Reds had an offensive righty/lefty split almost as severe as the Cubs, though favoring the lefties in this case. Milwaukee had the sixth-highest percentage of their team's losses allotted to their bullpen (32 of 79), which given the importance of relievers before 1957 almost certainly makes it the sixth-highest ever. But the law of diminishing returns starts to kick in at some point, and this column is long enough.
I'm sure I missed some doozies, but the above are the 2007 team splits I found most impressive.
References and Resources
First off, I could not have done this research without Baseball-Reference's Play Index. Sure, anyone can access the team splits, but if you subscribe to PI, you only have to go to the league splits page, click on a the red text for a given split, and then every team's stats for that split will appear. It saves at least 90% of the work.
I looked up 147 splits for all team—65 pitching and 82 hitting. Rather than last each individual split, I'll give you the categories I took them from and how many splits existed in each category. The 65 pitching splits were: platoon (8), home/away (2), halves (2), months (6), outcome (2), pitcher role (2), opposition defensive position (2), bases occupied (10), outs (3), clutch stats (7*), innings (13*), and days rest (14).
I * clutch stats and innings because there are actually eight and 14 stats there currently. When I started taking out stats, there were eight clutch stats, then one—close/late—disappeared. It came back, but I only had a partial sample, so I tossed it out. I only have 13 for innings because, for reasons unclear to me, Baseball-Reference lists "Extra Innings" and "Innings 10+" as two separate splits. Seems a little redundant to me.
The following 82 hitting splits were tapped: platoon (8), home/away (2), halves (2), months (6), defensive position (15), batting order (12), bases occupied (10), outs (3), clutch (7*), innings (13*), power/finesse (2*), ground ball/fly ball (2*).
I have the same problem with innings and clutch stats here. With power/finesse & GB/FB, I made a stupid error. Both have three categories—a neutral, and two ends. I accidently filed the two neutrals together, causing them to be hopelessly entangled. They're out of the sample.
In some cases, there are fewer than 147 splits for a team—strike years can lose a month, with DH-leagues I don't include how pitchers bat, nor do I use opposition defensive position. Whenever feasible and possible, I used all splits.
History instructor by day, statnerd by night, Chris Jaffe leads one of the most exciting double lives imaginable; with the exception of every other double life possible to imagine. Despite his lack of comic-book-hero-worthiness, Chris enjoys farting around with this stuff. His new book, Evaluating Baseball's Managers is available for order. Chris welcomes responses to his articles via e-mail. Oh, and now he's on twitter.
<< Return to Article