Lightning in the nine-hole

One of the authors was playing a computer baseball simulation*, as they both tend to do, and found himself innocently perusing the batting statistics for the pitching staff of the 1939 New York Yankees, a short-list candidate for the greatest baseball team ever. A name that naturally caught his eye was Red Ruffing, known for being one the best-hitting pitchers the game produced.

His performance in 1939 wasn’t overwhelming, but at .307/.347/.342, it was quite stout for his position. He clearly outperformed starting shortstop Frank Crosetti (.233/.315/.332), and ran a photo-finish with the man who took Lou Gehrig‘s place at first, Babe Dahlgren (.234/.312/.377). If you were the manager of this team, Tony LaRussa‘s theories aside, you would strongly consider batting Ruffing eighth in the order, possibly even seventh.

* We won’t name it. We would tell you what it sounds like, but that would imply that it’s a Rhymin’ Kind of game.

And there was more. Monte Pearson produced a gaudy .321/.379/.415 that season. He would definitely merit batting seventh, sixth if backup catcher Buddy Rosar (.276/.356/.343) was playing for Bill Dickey that day. Other pitchers were connecting, too. Steve Sundra posted a .265/.308/.367; Atley Donald finished at .250/.297/.300; Marius Russo came in at .244/.279/.293. And all these pitchers were at least half-time starters, not getting lucky in a minuscule sample size. (Manager Joe McCarthy spread the work out that season, having eight pitchers who started at least 11 games.)

Our co-writer began thinking he had stumbled upon the ’39 Yankees’ secret weapon. Sure, their batters were crushing it (well, most of them), but the pitchers were punching way above their position. Could this neglected aspect of the game have raised the Yankees from mere greatness to historical dominance? Could the possibly greatest team ever have had the greatest-batting pitching staff ever?

That’s where this article started—and it won’t give too much away to say that the answer is an unequivocal “no.” The 1939 Yankees had some fine-batting pitchers, but they had their share of anchor-men like Lefty Gomez (.151/.205/.178) and Bump Hadley (.177/.227/.210). It was also a high-offense era—the league triple-slashed .279/.352/.407, and even the AL’s pitchers managed an overall .197/.245/.240—so the numbers looked better than they really were. They’re very good, but not historically so.

And so what? We’ve got a bigger question to chase down. Which teams’ pitchers contributed the most to victory with their bats rather than their arms? Which team really did have the secret weapon?

Whaddya mean, pitchers?

The definition of “pitcher” seems obvious, but at its boundaries it can warp the numbers into incoherence. For example, when Jose Canseco threw an inning of relief in a runaway game (at a cost to his arm), did that make him a pitcher? Would his batting stats count with the staff’s—and since his 1993 Rangers were a DH team, would that make him effectively the whole pitching staff for our purposes?

No. That’s ridiculous. This example shows where we must start making adjustments, though where to stop is trickier. A player could be a part-time pitcher and part-time position player: Ossie Orwoll of the 1929 Philadelphia Athletics had 12 games on the mound and nine in the outfield. Also, a pitcher could bat well enough that he gets substantial pinch-hitting plate appearances: Don Newcombe, in his .359/.383/.632 campaign of 1955, pitched in 34 games and pinch-hit in 23 more. We needed to decide in these gray-area cases how much of a player’s batting should count toward the pitching staff numbers and how much should not.

We set a few basic rules. Everyone who pitched was counted as a pitcher, but all their plate appearances did not necessarily count. We deemed that, as a ceiling, a pitcher could expect to come to the plate once every two innings. A staff could thus claim 4.5 PA per nine innings pitched by each pitcher, the remainder discarded into the position-player bucket. If the player played no other defensive position (including DH for our purposes, rare and bizarre as it might be), he got credit for all his PAs.

This means Newcombe in 1955 gets credit for all his times up, including pinch-hitting. Orwoll, pitching 30 innings total, gets 15 of his 53 plate appearances counted for his 1929 staff. Canseco would have a full 0.5 PA pro-rated and counted for the 1993 Rangers’ pitching staff—but due to an insufficient sample size of 1.5 PAs total, we don’t include them in our survey. Just one more reason why he shouldn’t have been pitching.

We did some occasional data-fiddling beyond these few principles. Some pitchers with single-digit numbers of position-player games were allowed past their 4.5PA/9IP limits, things loosening as they approached zero games in the field. Four players who pitched very little and played no other position but pinch-hit often got manually limited, and three more got adjusted because their pitching/position game ratios were nowhere near the PA ratios. This changed the overall numbers very little, but satisfied some mild obsessive urges.

Whaddya mean, greatest?

There are several ways you can measure the batting greatness of a pitching staff, and we ran through most of them. We set a floor of 100 adjusted pitcher plate appearances for a team to qualify for the overall rankings, giving us 2,094 qualifying teams from 1876-2011. This excludes basically every DH team, plus a smattering of 1884 Union Association teams that played a few games, then expired. So if you’re looking for the 2011 Orioles pitchers and their epic .409/.409/.636 performance to top our lists, you’re in for disappointment.

The most basic method we used was the OPS+ calculation popularized by Baseball-Reference.com, the pitching staffs measured against all the batters in their respective leagues. As the following table of the best staffs under that method shows, though, there is an obvious problem.

 Rank  Year  Lg   Team          AVG    OBP    SLG    OPS+
  1    1884  UA  St. Louis     .287   .324   .394    142
  2    1882  NL  Boston        .281   .330   .399    133
  3    1886  AA  St. Louis     .279   .339   .394    125
  4    1876  NL  Chicago       .316   .329   .377    123
  5    1884  AA  Louisville    .263   .288   .367    117
  6t   1878  NL  Chicago       .276   .325   .345    115
  6t   1880  NL  Troy          .271   .289   .364    115
  8    1882  NL  Worcester     .262   .286   .395    114
  9    1876  NL  Louisville    .313   .315   .366    111
  10   1915  AL  Boston        .248   .324   .365    109

The top nine slots are taken by pitching cohorts from the primordial days of major league baseball, all 1886 or earlier. Going further, only two of the top 25 staffs played in 1890 or later. That’s because the development of pitchers as specialist players, not expected to contribute with their bats, was still in its infancy. The decline of pitcher offense has been almost continuous throughout major league history, though it may possibly be reaching a floor in our era.

image

While the first table did give us a valid answer, as pitchers batted a lot better then, it’s not really what we’re looking for. We want to see the pitchers who outperformed their fellow moundsmen the most, defying the presumption of hapless hacking and giving their teams the greatest offensive bonus at their position. So we devised PitOPS+, which is pretty much what it says: pitchers’ OPS+ as measured against the pitchers of the league. And we get rather a different set of teams using this method.

 Rank  Year  Lg   Team          AVG    OBP    SLG  PitOPS+
  1    2002  NL  Colorado      .254   .269   .338    203
  2    2010  NL  Milwaukee     .207   .249   .281    202
  3    1965  NL  Los Angeles   .187   .238   .257    196
  4    1915  AL  Boston        .248   .324   .365    195
  5    1988  NL  New York      .196   .217   .256    192
  6    2003  NL  St. Louis     .208   .251   .272    190
  7    1982  NL  Pittsburgh    .223   .251   .302    189
  8    1971  AL  Boston        .210   .249   .302    187
  9    1958  NL  Milwaukee     .223   .268   .302    186
 10t   1918  AL  Boston        .246   .331   .333    183
 10t   2003  NL  Chicago       .201   .220   .302    183

Before you ask, yes, we remembered to adjust for ballpark. The 2002 Colorado Rockies’ pitchers really were that good.

But a new problem has cropped up, not nearly as severe as for the pure OPS+, but still fairly serious. The PitOPS+ list is heavy with recent teams, four of the top 11 having played in the last decade—and it’s not too hard to see why. Those staffs were competing against a very low baseline, pitcher batting at its historical nadir. With the denominator so small, it takes only moderate changes in the numerator to make large changes in the ratio. In baseball terms, it doesn’t take very many hits, walks, or homers above the average to lift the PitOPS+ a lot.

So again, we have a valid answer, but not a fully satisfying one. Rather than tracking on-base and slugging, perhaps we need to look directly at what really counts in baseball: runs. Which staff created the most batting runs above the average for its league, and thus produced the greatest cumulative added value toward winning games?

Note: we used a fairly basic version of the Runs Created formula, given how far back our study went. Our version was (H+BB+HBP) * TB / AB+BB+HBP.

But if we look at it purely through counting Runs Created, we run into another consideration. Pitchers batted much less often in recent seasons than in ones long past, even ignoring the designated hitter. This is due partly to declining pitcher batting, partly to evolving in-game tactics and gentler use of pitchers’ arms. In the 2010-2011 National League, pitchers had 2.182 plate appearances per game, as compared to 2.903 PA/G in the 1960-1961 NL. (We haven’t reliable splits for earlier decades, but the numbers were surely higher still.) This leaves modern pitching staffs with fewer opportunities to pile up Runs Created and separate themselves from the league average.

Is it fair to debit later staffs, due at least partly to a universal culture of managerial decisions over which they have minimal control? We eventually decided it was not, so we devised a Runs Created rate statistic. We measured the staffs’ performances per 100 plate appearances, modified for park effects and neutralized to a run environment of 750 runs per 162 games. Here we reached what we considered a satisfactory end to our quest for a fair measure, primarily from conviction though perhaps a little from exhaustion.

 Rank  Year  Lg   Team          AVG    OBP    SLG    RCAA/100
  1    1915  AL  Boston        .248   .324   .365        6.3
  2    1918  AL  Boston        .246   .331   .333        6.0
  3t   2002  NL  Colorado      .254   .269   .338        4.4
  3t   1884  UA  St. Louis     .287   .324   .394        4.4
  5    1884  AA  Louisville    .263   .288   .367        4.3
  6    1958  NL  Milwaukee     .223   .268   .302        4.2
  7    1888  NL  Detroit       .237   .293   .317        4.1
  8t   1982  NL  Pittsburgh    .223   .251   .302        4.0
  8t   1886  AA  St. Louis     .279   .339   .394        4.0
 10t   2010  NL  Milwaukee     .207   .249   .281        3.9
 10t   1974  NL  Pittsburgh    .219   .280   .283        3.9
 12t   1965  NL  Los Angeles   .187   .238   .257        3.8
 12t   1882  NL  Boston        .281   .330   .399        3.8
 14t   1971  AL  Boston        .210   .249   .302        3.7
 14t   1878  NL  Chicago       .276   .325   .345        3.7
  16   1927  NL  Cincinnati    .256   .291   .321        3.6
  17   1926  NL  Cincinnati    .262   .311   .333        3.5

We find something like an even spread of teams from the long history of major league baseball, though with a tilt toward the 19th century. To satisfy those who would rather stick to modern play, we extended our usual top 10 list to 17 teams, giving us 11 squads from the post-1900 era. Now we can take a little closer look at the pitching staffs in question, to see what made them so extraordinary.

We’ll cover the 19th century teams first, so you can skip down a bit if you’re a modernist. Note that many, even most of the pre-modern slugging pitchers we’ll mention also did some position-play duty.

The St. Louis Maroons of the Union Association in 1884 tied for third on the list, but the UA is such a discredited organization, thoroughly dominated by the Maroons, that we will just note their presence and continue on.

The 1884 Louisville Eclipse of the American Association boasted Guy Hecker mashing a 149 OPS+ while pitching more than two-thirds of their innings. Hecker’s batting performance was second on the team to the Louisville Slugger himself, Pete Browning (and Hecker tied Browning for the team lead with four home runs). Two years later, Hecker would lead the whole league in batting average, after which he began playing the field more often than pitching in a precursor to Babe Ruth‘s career path.

The 1888 Detroit Wolverines had Pretzels Getzein (91 OPS+, and isn’t his name alone worth reading the article?) and Pete Conway (121 OPS+) pitch almost 800 innings and make 358 PAs. Utility fielder Ed Beatin chipped in with 107 IP and a 143 OPS+. Their exertions did not save the Wolverines from a calamitous 16-game losing streak that swept them out of contention.

The 1886 St. Louis Browns won the American Association, thanks partly to a ferocious one-two punch from their top hurlers. Dave Foutz hit for an 111 OPS+ while pitching over 500 innings, while Bob Caruthers threw nearly 400 and posted a crushing 200 OPS+. Their success carried into the post-season exhibition with the Chicago White Stockings, making the Browns the only AA team ever to win the “World Series.”

Grasshopper Jim Whitney of the 1882 Boston Red Stockings hit at a 183 OPS+ clip, easily team best, while pitching more than half Boston’s innings (and playing some outfield and first base). Fellow hurler Bobby Mathews‘ 68 OPS+ did drag down the average: league pitchers overall produced an 82 OPS+.

Terry Larkin (121 OPS+) did over nine-tenths of the pitching, and pitcher batting, for the 1878 Chicago White Stockings. Cap Anson‘s squad could build only a .500 finish from his contribution, though.

Now for the more modern clubs, 1901 and on. We’ll work in reverse order, for greater drama and because we’re saving the obvious matter of Numbers One and Two for last.

The 1926-’27 Cincinnati Reds had two career 80+ OPS+ pitchers in Carl Mays and Red Lucas. Red outdid himself both years, though we do dock him in 1927 for 10 games in the field. Mays stumbled in ’26, but Pete Donohue (102 OPS+) and Dolf Luque (118 OPS+) had their best full seasons at the plate to compensate.

For the 1971 Red Sox, Gary Peters was a lifetime 70 OPS+ batter (in a depressed batting era for pitchers) having his second-best batting season (91). Sonny Siebert was always feast or famine at the plate (four seasons below 0): 1971 was a 121 OPS+ banquet for him. Most of the remaining staff was a batting wash, but Peters and Siebert were enough to lift Boston onto the list.

The 1965 Dodgers saw Don Drysdale, a career 45 OPS+ batting threat, soar to a deadly 140. Sandy Koufax chipped in with the lone positive OPS+ of his career (a 29). Nobody else did much with the lumber (save Nick Willhite in 13 PAs), but with league pitchers below zero in OPS+ that year, it—or rather he—was enough. Their P-batting margin over second-place San Francisco was around two wins, and the Dodgers won the pennant by two.

The 1974 Pirates boasted two great pitcher-batting seasons from Jim Rooker (109; lifetime 41) and Ken Brett (122; lifetime 94), though neither was a career peak. The other starters were above league-pitcher average or at least close, and the staff batting, producing close to two wins above league average, may have been enough to push the Bucs to a division title (1 1/2 games ahead of St. Louis and its league-average P-batting).

The 2010 Brewers are a surprise, though mainly because nobody really thinks about pitcher batting in the 21st century. They followed the emerging pattern, as two starters—Yovani Gallardo (122; lifetime 70) and Chris Narveson (94; lifetime 46)—demolished the league-average pitchers’ OPS+ of -5. Their other three starters were either average (Dave Bush; -5) or well above (Randy Wolf at 49; Manny Parra at 36), and a lousy-swinging bullpen didn’t drag them down much.

The 1982 Pirates staff continues the pattern. Of their five most frequent starters, two were outstanding batters—Rick Rhoden (100; lifetime 59) and Don Robinson (98; lifetime 62)—while John Candelaria (54) was good, Manny Sarmiento (21) was pretty good, and Larry McWilliams (4) was average. All five outperformed their career marks significantly.

The 1958 Milwaukee Braves did it with two pitchers, only more so. Warren Spahn was a pretty decent hitter (lifetime 42) who had easily his best batting season, producing a 131 OPS+. Lew Burdette hoisted his career 29 to a 73 (third-best lifetime: his two best seasons were oddly in the 1964-’65 twilight of his career). Those two alone accounted for almost half their team’s pitcher PAs, 230 of 483. The rest of the staff ended up average, but Spahn and Burdette ate up enough innings and at-bats to raise them into the P-batting stratosphere.

The 2002 Colorado Rockies effectively had a six-man rotation, and they all contributed as hitters. Mike Hampton (112; lifetime 67) won his fourth straight Silver Slugger (yes, they give those out for pitchers, too). Against a league OPS+ of -1, Rookie of the Year Jason Jennings posted a 79 (lifetime 29); Denny Neagle a 53 (7); Shawn Chacon a 35 (-9); Denny Stark a 16 (-17). John Thomson came in at -1 before his trade to the Mets, but average production from the back of your rotation is always a great floor to build on. Third all-time could not raise the team past 73-89, though.

And then there are No. 1 and No. 2, and the elephant in the room. It seems almost unfair to pit pitching staffs against each other at the dish, knowing one team in one era has the ringer of all ringers. Sure enough, the 1915 and 1918 Red Sox top our list, thanks to the irresistible force that was Babe Ruth.

Except that’s not entirely the case.

The 1915 Red Sox had six pitchers of 75 IP or more who hit better than league average for all batters: Smokey Joe Wood‘s 110 OPS+ was good for only sixth-best on the roster. Ernie Shore was the sole “rotation” pitcher with a poor batting year (and it was admittedly bad at 1 OPS+). Babe Ruth’s 188 was tremendous, well surpassing position leader Tris Speaker‘s 151, but it was backed by very strong, incredibly deep support.

As for 1918, the Babe was batting even better, compiling a 192 OPS+—but he did this as a part-time pitcher. When we prorate his playing time, only 22 percent of his plate appearances get credited to the pitching bucket, putting his adjusted PAs third on the staff. The other hurlers needed to contribute, and here a familiar pattern reasserts itself, as two pitchers led the way. Carl Mays (121) and Bullet Joe Bush (101) pitched half of Boston’s innings that year. (Mays’ excellence is no surprise: recall he was on the 1926-’27 Reds.) Their great hitting, plus the average-to-good work of Sad Sam Jones and Dutch Leonard filling out the starting cohort, lifted Boston back to the top of the list.

Both teams managed to do something no other team since 1886 has accomplished: produce a better OPS+ from its pitchers than from the rest of its players. The 1915 Boston pitchers produced a 109 against 105 for the other players; in 1918 it was 102 against 97. A dozen clubs managed it from 1876-1886, but these Red Sox were, and surely will always be, the last.

The Red Sox won the World Series both years, pacing the American League by 2 1/2 games each time. The added wins above average their pitchers added with their bats each season hovers around three. The second-place team in 1915, Detroit, had league-average P-batting; the 1918 runners-up, Cleveland, were perhaps half a win below average. One can plausibly argue that pitcher batting won the Red Sox those two championships, and Babe Ruth was only part of that story.

The flip side

Before we wrap up, we will take a brief glimpse at the opposite end of the spectrum: the worst-batting pitching staffs of all time.

Room for failure in this regard is narrower than the room for success, since average performance is much closer to zero than to perfection. There are six qualifying teams that do three runs or more worse than average per 100 plate appearances, while there are 39 that gain at least 3R/100PA.

Measured against league OPS+, the worst-batting staffs cluster in recent years, due naturally to the progressive deterioration of pitcher batting. The most notable trend-bucker is the 1965 Tigers, finishing third-to-last both against OPS+ and PitOPS+. Detroit’s staff produced a .080/.125/.090 line, versus .128/.178/.172 for the league’s pitchers, and against a standard OPS+ of zero, the Tigers’ hurlers combined for a grotesque -38. Mickey Lolich (.058/.080/.058) and Denny McLain (.054/.123/.054) both far undershot their usual merely poor batting ability, and nobody with any substantial PAs stepped up to fill the gap.

Measured by our Runs Created per Plate Appearance metric, the most hopeless batting pitchers of all time belonged to the 1890 Baltimore Orioles of the American Association. However, this club played barely a month and a half of the season, one where there were three major leagues spreading the talent pool thin, so we may be forgiven for setting their claim aside. In their place, we nominate as the worst batters of all time the pitching staff of the 1956 Kansas City Athletics.

Twenty-one players pitched for the A’s that season, and only two of them produced batting lines better than the pitchers’ league-average OPS+ of 25. The best of those was Jack Crimian, at 47 in 30 PAs. (Ironically, this was the worst season OPS+ of his short major-league career.) Staff workhorse Art Ditmar stumbled to a -6, and then it got gruesome. The next three highest pitchers in plate appearances—Wally Burnette, Tom Gorman, and Lou Kretlow—posted batting averages of .051, .051, and .061. (Those were also their slugging percentages.) In 130 combined PAs, they scratched together six hits and, somehow, nine walks. Bobby Shantz‘ usually decent bat (38) failed him to the tune of a -19, and Tommy Lasorda, in his last major league action, added a 1-for-13 to the mix.

This sad batting record fit well with a team that finished 52-102, seven games clear in the cellar. (The pitchers’ bats cost them roughly a win and a third of that margin, a small un-contribution in the grand scheme.) There would be some release from this purgatory, though: Ditmar and Shantz rode the famed trade pipeline to the New York Yankees in the offseason. In New York, they enjoyed years of individual and team success—and not incidentally, hit much better.

In conclusion

Have we discovered the secret weapon for team success after all? Probably not, though in a few outlying cases pitcher-batting excellence may have tipped pennant races, and that deserves some appreciation. Batsmen like Don Drysdale and Ken Brett made real contributions with something other than their pitching arms, though in most cases it took more than one breakout pitcher-batting season to raise a team to the top of the rankings. And recall the 1939 Yankees, where this trip began. A few sub-par bats undercut the work of Ruffing, Pearson and others. An all-time staff could afford almost no weaknesses.

If there is one outstanding piece of advice one can take from our survey, it is this: If at all possible, sign Babe Ruth. (But you probably knew that already.) If there’s a second piece of advice one can take, though, it would be to find good performers to back up your star. That’s a little more practicable, and has applications far beyond just finding a pitcher who can mash.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: Fluke Watch: Johan Santana
Next: Trader’s corner: week four »

Comments

  1. Shane Tourtellotte said...

    On the nose, Brandon.  It was 2003.
    Pitchers (373 PA): .202/.221/.304
    Pinch-hitters (236 PA): .155/.244/.223
    Even Billy Beane at his on-base hungriest would take Chicago’s pitchers over the pinch-hitters.  The hurlers’ OPS was up almost 200 points from 2002, though, so it wasn’t all due to a collapsing bench.

  2. Brandon Isleib said...

    IIRC, there’s a Cubs team in the 2000s – 2003? – whose P OPS exceeded that of their PHs.  It was more a reflection on the latter than the former, but it’s still obviously abnormal.

  3. Paul G. said...

    I am told that there are some technical issues with multiple bylines, so I am putting mine here.  It’s my first article!

    A master of computers and computer-related accessories (to those with less technical backgrounds, it’s called “magic”), Paul Golba is probably either fixing a computer problem in or near his home in New Jersey or watching the Yankees induce the hatred of the rest of the world.  He also occasionally writes something.

  4. John D. said...

    Just a question in regards to your conclusion, where did that 1939 yankees team end up position wise (placement) in the grand scheme of things?  You mentioned why they weren’t at the top, I was just curious where they did end up.

  5. Shane Tourtellotte said...

    John D.:  Out of 2094 qualifying teams in our survey, the 1939 Yankees came in a tie for 104th (a 15-way tie), producing 2.1 Runs Created Above Average per 100 PA.  They led the AL that year, though Pittsburgh in the NL finished with a better mark.  A fine showing, but not truly great.

  6. Paul G. said...

    To elaborate a bit further on the 1939 Yankees, if you use one of the other measures that we mentioned, they still come across as good but not great.  The staff’s “normal” OPS+ is 49, tied for 322nd place.  That ties them with the 2002 Colorado Rockies, our #3 set of bats, but in a very different offensive context.  (Pitchers in the 1939 AL averaged an OPS+ of 24; in the 2002 NL it was -1.)  On the PitOPS+ front they combined for a 140, good for 136th all-time, but they are actually second best for the year being inched out by the Pirates (PitOPS+ 141).

    Now, among Yankee teams they do stand out.  They are the second best hitting pitching staff in the pre-DH era of the franchise (+2.1), just barely ahead of 1921 and 1950 (+2.0).  The best is the 1922 squad (+2.6) which featured the services of Bullet Joe Bush, Sad Sam Jones, and the ubiquitous Carl Mays.  If those sound familiar, along with the Babe they were the backbone of the 1918 Red Sox staff, which also along with the Babe were shipped off to the Yankees for a song resulting in the changing fortunes of the two franchises.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current ye@r *