Mailbag: answering your commentsby Shane Tourtellotte
June 26, 2013
The Hardball Times, like any good website, encourages feedback from its readers. The Comments sections for our articles will sometimes hum with dozens of responses, a vigorous exchange that frequently brings the writers back to respond and keep the ball rolling.
Our readers will come up with some very good observations and questions, some of which can be handled in the give-and-take of the Comments section. Some of them require an extra effort.
I have gotten a few of the latter kind in the last few weeks, observations that made me want to chase down arcane facts. They might not be big and important enough to merit a full-sized post individually, but put them together and they'll fill out a web page quite nicely. I've done that with three of my recent comments, and this is the result.
You can consider this a Mixed Nuts version of my writings. If one morsel doesn't satisfy you, there's another one coming that may. (Any attempt to make a personal connection between me and "Mixed Nuts" will be either ignored or admitted to freely. Depends how I feel.)
Spahn and Shane and—wait, let me start over ...
My June 12 article about pitcher batting brought out an anecdotal claim I had never heard before, but which immediately fired up my tracer-chasing instincts. From AndrewJ:
Many years ago (pre-sabermetrics) someone theorized that Warren Spahn owed about 50 of his 363 career victories to the fact that he wasn’t lifted for a pinch-hitter late in games, thus getting the “W” in come-from-behind games which might have otherwise gone to a reliever. Not sure how that holds up.
This is a tricky one. It feels straightforward at first, but it requires a big dose of subjective judgment—or many small doses, covering a lot of games.
Before I get started, allow me to sweeten things with an anecdote of my own. Spahn's 363 career victories, most ever by a left-hander, could have been one win higher except for a quirk in the scoring rules. That phantom win would have been his very first, in his rookie season of 1942. It was a game his Boston Braves won, but he didn't, due to a recycling riot.
Yes, I'll explain that.
1942 was smack in the middle of the Second World War. Not only were everyday items being tightly rationed, but the armed forces were crying out for scrap metal to be melted down and recast into planes and ships and, hopefully, the bullet someone would put between Hitler's eyes.
The New York Giants, doing their patriotic bit, held a scrap-metal drive promotion for their double-header on Sept. 26. Any child bringing in enough scrap for the war effort would be admitted free.
The kids came through, big. Over 8,000* boys and girls brought in an estimated 56 tons of scrap metal—14 pounds apiece on average: no shirking there—and got their free admissions. The fortunate kids got to see their Giants beat the Braves, 6-4, in the opener and then take a 5-2 lead against some young pitcher named Spahn in the nightcap.
* Paid attendance was given as 2,916 that day, but The New York Times reported a crowd of 11,205. Presumably the difference comes from non-paid admissions, those being the scrap-metal kids.
It was in the middle of the eighth inning that thousands of restless, apparently bored kids with no parental supervision did what you'd expect: they made trouble. They swarmed the field, engulfing the Braves going out on defense, driving the players to the dugouts.
Umpire Ziggy Sears ordered an announcement be made for the field to be cleared, on pain of forfeit, but it couldn't be heard over the din of the kids. A mixed force of police, hired guards, ushers, and grounds crew tried and failed to sweep the diamond clear.
Sears did what he had to: he forfeited the game to Boston. Spahn avoided the loss, but the rules of baseball then and now state that a pitcher cannot be credited with a win in a forfeit if his team was not ahead when the game was forfeited. This is actually pretty fair, but it does mean that Spahn had to wait all the way until 1946, once he got done serving his country, for his first win in the big leagues.
For the games he did win, how many can plausibly be credited to Spahn being left in for his bat? The question is muddled due to the other good reason for leaving him in: his pitching. For a considerable stretch, he was the best pitcher on his team, starter or reliever. Retaining the benefits of that arm might have meant more than balancing his bat against some bench player's.
Of course, for this theory to make any sense, Spahn would have to be established as a good batter. Over his career he was, though not dominant by any means: a 43 OPS+ overall. However, it took some time for his bat to meet the standards of even an average pitcher. This table shows how long:
Year lgpOPS+ SpahnOPS+ 1942 N/A -1 1946 17 9 1947 19 12 1948 26 23 1949 25 11 1950 24 39
We don't have split records for pitchers as far back as 1942 yet, but there's little doubt that, for the four games he played that year, Spahn wasn't raising any eyebrows with his batting. Once he got his swing established, however, he stayed in the groove. In only one of his 15 seasons after the 1950 breakthrough did he have a below-average OPS+ for a pitcher, in 1957. He made up for that with his outstanding 131 OPS+ in 1958.
It makes no sense that Spahn would be earning any added innings on account of his good hitting before he actually had good hitting. Assuming a dead minimum of one year's proficiency to gain that benefit of the doubt—probably too short, but I'm stretching the point in favor of the hypothesis—we needn't check any games before 1951. That means Spahn's first 86 career wins are off the table, but that's okay. There are plenty more where they came from.
I composed a generous first screen for games that might fit the profile. If Spahn pitched at least five innings (assuming he started), and came to bat after that with his team not yet ahead in a game it would eventually win, I counted it. Of his 277 eligible wins, 71 met these loose criteria.
The question looms: how much to tighten the criteria? There's a wide gulf between Spahn batting for himself in the fifth inning of a scoreless game and doing so in the 10th frame tied, 2-2. (Both of which he did in 1952.)
The matter is complicated by the difference in eras. Pitchers were used differently during Spahn's career than they are today. What we'd consider obvious pinch-hitting moments were iffy then, and situations we'd think debatable would leave 1950s bench-warmers firmly on the pine as Spahn went to the plate.
I had to use my judgment, fuzzy as it might be. I went over each of the 71 instances, and said "yes," "maybe," or "no" as to whether I'd send in a pinch-hitter, assuming a generic pitcher batting rather than Spahn but also accounting for Spahn's pitching skills if he stayed in the game. The "yes" tally ended up at 20, with "maybe" providing an added 19. Were I to run through the list again, my totals might vary a couple either way.
Regardless, that's not 50. That's not even "about" 50. Someone with flexible enough standards might be able to pick out 50 from the list, but most of us would poke Swiss cheese holes in their set. That said, there is something to the original claim. Call it 20, or 30 if you count half my "maybes," but either way, Spahn was getting chances to bat for himself when games hung in the balance that many other pitchers would not have, and that earned him extra wins.
The best illustration of this is that there are six games on the list in which he batted for himself in extra innings with the score tied. This didn't indicate total dominance from the mound. None of the games was a shutout when he batted in extras, and there were as many games in which he had given up one run (one) as where he had given up six. Yes, he'd coughed up six runs and was still batting for himself in the 11th.
The jewel of this collection was the nightcap on Sept. 13, 1956 against the Phillies. Spahn came to bat with one out in the top of the 10th, the score tied 2-2, and singled. In the 12th, the game now knotted at three, he batted for himself again with one down. He walked and would score the game-winning run. Odds are good we will never again see the starting pitcher score the winning run in the 12th.
Did his batting performance justify this confidence? Lifetime, Spahn hit .194/.234/.287. In the 71 games on this list, he had 96 plate appearances meeting my cutoff criteria. In those trips to the dish, he batted .224/.280/.447 with five home runs.
Nowadays, we've learned to expect an uptick in batting from someone facing an opposing pitcher for the second, third, or fourth time. This goes well beyond that. It may be a sample size artifact—and I'm not even breathing the word "clutch"—but when managers took a chance that he'd deliver, he tended to come through.
Keep in mind, these numbers form a specific subset of his career. There were times when Spahn was left in a game late and lost. There were times he was hooked early or at what we'd consider a normal stage. There were plenty of times when his team got an early lead and made the manager's decisions about Spahn a lot less stressful. This is a view through a keyhole—but it is an intriguing view.
Pitcher batting, and Jocketty-ing for position
For my same article on pitcher batting, reader Metsox mentioned someone swimming against the tide of indifference to hard-hitting hurlers:
Travis Wood's 296/345/556 triple-slash is coughing quietly in the corner. Posted a 146 wRC+ [year-to-date].
I replied to Metsox that sample size and the also-ran status of Wood's Cubs would probably keep his hot streak a secret. I did add, "But he does help give the Cubs' pitching staff an OPS+ of 44, against a current NL pitchers' average of 0. I may start a bit of digging here ..." And unsurprisingly, especially since I hinted I'd do such a thing at the end of the article itself, I did that very thing.
I looked up team pitchers' OPS+ in the National League between 1998 and 2012, the era of 16 NL teams. I then weighed them against league pOPS+, to see how pitcher-batting prowess waxed and waned over the years, and whether any specific teams, or specific general managers, might have tried to get a quiet edge out of this neglected aspect of the game. I wasn't jackhammering down to the bedrock of the numbers, just taking an attentive first pass.
To tie immediately back to Metsox's point, it's no surprise to find the Chicago Cubs doing well with pitcher batting this season. Over the previous 15 years, they have averaged 10 points of pOPS+ better than league average, the best figure in the NL and, thus, in baseball.
I proposed to look for organizational causes for such numbers, and I could credit Jim Hendry, the Cubs' GM from July 2002 through July 2011, if there weren't a far more obvious cause for much of this excellent run.
The cause's name is Carlos Zambrano.
For his 11 years pitching for Chicago, 2001 to 2011, Zambrano posted a 62 OPS+, about as good as any pitcher in the current era. Discounting the stub of his 2001 rookie season, for a decade Zambrano took almost a fifth of the Cubs' pitcher plate appearances, 706 out of 3,635. Filter Zambrano out of the equation, and the Cubs have below-average pitcher batting for that decade.
Zambrano's arrival preceded Hendry, so there's no credit there. He was gone shortly after Hendry was fired, but that's not Theo Epstein's indifference to pitcher batting: Zambrano's volatility provided ample reason to wave good-bye. The Cubs' run of nine-hole excellence winds up being largely the serendipity of one player's presence.
Second in pitcher batting for this stretch is St. Louis, about seven points above average the last 15 years. Their best stretch was 2003-2009, with six seasons more than 10 points above average and as high as 33 points above (plus one slightly negative hiccup).
That run began under GM Walt Jocketty, but it persisted for two seasons after John Mozeliak took the reins. Also, Jocketty had been in office since the end of 1994: it would be a bit odd if he developed a strategy for collecting hot-hitting pitchers that deep into his tenure.
Then again, Jocketty landed with Cincinnati for the 2009 season, and after one middling year while he got up to speed, the Reds have had three seasons well above average since. Then again again, they're well under-performing this year. The Jocketty case is interesting, but skepticism is in order.
Of course, St. Louis also had a manager noted for innovative strategy involving pitchers in the batting order. Could Tony LaRussa have been nudged toward batting his pitchers eighth because there wasn't as big a gap between them and the weakest position player as there was on other teams? Might he even have been asking Jocketty for better-batting pitchers so he could run this experiment with a somewhat larger safety net? Maybe these secrets are in LaRussa's latest book, but I haven't read it, yet.
The worst pitcher-batting team in this survey is, not too surprisingly, the Pittsburgh Pirates, nine points below average per year. I doubt this is a conscious strategy. The Bucs have had 20 consecutive seasons under .500, and teams that are generally bad tend to be bad at any particular skill you happen to select.
(The string is carrying over into what's been a very good year so far: as I write on Tuesday morning, Pittsburgh's pitchers are batting .077/.097/.077 for a -50 OPS+. That has to regress upward soon, but yikes!)
Other stretches of bad batting either can be attributed to an overall poor team (such as the 2006-2011 Nationals) or overlap the tenures of several GMs and so appear to be nobody's particular idea (Milwaukee and Cincinnati). Good pitcher batting, the subject of my interest, falls into few obvious patterns. Aside from the Jocketty case, there is just one that might indicate deliberate planning, with the emphasis on "might."
Josh Byrnes took over as Arizona Diamondbacks general manager in Oct., 2005. Their pitcher batting had been below average for the last five years under Joe Garagiola Jr., and remained so in Byrnes' first year.
After that grace period for Byrnes to start doing things his way, the D-backs pitchers began batting well for the five following years. Byrnes would be around for just four of them, getting fired in July of 2010. After his successor, Kevin Towers, had a year's grace to start getting things done his way, D-backs pitcher-batting plunged back into negatives in 2012 and is pretty poor so far this year.
This could well be coincidence. Given all the years and teams being studied, a lucky confluence here and there is not unexpected. The proof could stand with the team where Byrnes fetched up, in San Diego. 2012 was the grace year there, and so far in 2013, their pitcher batting has been a little below average, much like the last two years.
At the moment, it's not looking too good for Byrnes being a secret P-batting fan. Indeed, it isn't looking good for there being any such GM in baseball today, suggestive as Jocketty's case is. The thin percentage to be gained from the nine-batter (or occasional eight-batter, if LaRussa's your manager) doesn't seem to be enough to have tempted any front offices in recent years. But I'll keep half an eye on Jocketty and Byrnes in case the patterns get stronger.
The night the running stopped
On June 14, I wrote a THT Live article about the third 18-plus-inning game played in the majors in under a week, and from there about other years with a lot of marathon games. 1967 holds the record at nine, three of them played by the Yankees, two of those against the Red Sox. Reader Tom Dockery remembered one of those games in connection with a little piece of television history:
One of those 1967 Yankee games was in the second game of a doubleheader on Aug. 29. I watched only the last eight innings of that 20 inning affair after having watched the last episode of The Fugitive.
For the benefit of some of our younger readers, I will briefly explain why this is extremely interesting. It has nothing to do with the feature film later based on the TV series, even though it starred one guy who would later play Branch Rickey, and another who would later play Ty Cobb. The movie itself exists due to the echoes that still resounded from what was the biggest television event of its time.
In very short, The Fugitive was about Dr. Richard Kimble, wrongfully condemned to death for the murder of his wife. He escapes captivity and wanders the country, at once eluding pursuit by Lt. Gerard and hunting for the one-armed man Kimble believes committed the crime. The series ran for four years, concluding with a two-part episode that wrapped up the story, something fairly unusual for its era.
Instead of ending the series in May, as was and is common on network TV, ABC withheld the two-parter over summer reruns, airing the episodes in late August, 1967. They were hoping to build suspense and anticipation, and their success was monumental. The final episode scored a 45.9 Nielsen rating and a 72 share, representing the highest viewership numbers ever for a TV series. The record would stand for 13 years, until someone took a couple shots at J.R. Ewing.
This was a night when a huge number of people stayed home to watch a TV show, preferring it over such alternative forms of entertainment as, say, baseball. With all those people at home, the attendance numbers at games that evening must have suffered. Right?
That's the question I asked myself, anyway. Then I went to figure out the answer.
The notion seems pretty easy to check. August 29, 1967, fell on a Tuesday, a weekday. Fortunately, the days before and after were also weekdays: we don't get the naturally higher attendance of weekend games skewing the attendance numbers. If we check attendance figures for the night of The Fugitive finale and the adjacent days, we should see how much of a drop-off happens.
A fair comparison does require wholly comparable circumstances, which were pretty elusive for this three-day span. Three pairs of teams played double-headers, two of them the night of the finale. That artificially boosts attendance numbers, so I had to exclude them.
Series with day games likewise warped the figures, costing me the Dodgers-Giants series. The Astros and Cubs played all their games in the day at lights-less Wrigley Field, but since the finale aired several hours after the Tuesday game would have ended, it couldn't possibly have had an effect on attendance.
That left five sets of teams playing one game the night of The Fugitive's finale, as well as single night games the day before and the day after.
Attendance 8/28/67 8/29/67 8/30/67 Pit @ Atl 8,725 6,516 8,674 Phi @ Cin 9,524 8,667 6,453 Cle @ KCA 3,933 4,326 4,174 Bal @ Min 27,723 26,291 26,034 ChW @ WsA 5,875 7,982 9,492
The effect of, up to that time, the most popular television episode in American history is not immediately apparent. Attendance hit a low that night in Atlanta, but it also hit a high in Kansas City. In the other three cities, Tuesday's attendance ended up in the middle. I could walk away and say there was no effect, but I'm going one level deeper instead.
Adding up attendance for the five series doesn't really work, because the numbers are overwhelmed by the Orioles-Twins series*. Instead, I took averages for each three-game set, calculated how much each game was above or below the mean, and produced combined figures for each day, weighting the series equally.
* The huge disparity for the games in Minnesota is because the Twins were the only home team of the five to be in a tight pennant race. All the others were double-digit games out of the lead in late August.
Att. Factor 8/28/67 8/29/67 8/30/67 [Average = 1.00] Pit @ Atl 1.0945 0.8174 1.0881 Phi @ Cin 1.1594 1.0551 0.7855 Cle @ KCA 0.9490 1.0438 1.0072 Bal @ Min 1.0390 0.9853 0.9757 ChW @ WsA 0.7549 1.0256 1.2196 Total 0.99936 0.98544 1.01522
Take the Monday and Wednesday totals together, measure Tuesday against that, and the conclusion is that The Fugitive cost baseball games about 2.2 percent of their "proper" attendance. You could, instead, consider Monday alone the proper baseline, and say that Wednesday's higher figures came from people deferring their trip to the ballpark by one night. Working this way, the Fugitive Effect ends up around 1.5 percent.
Even if we consider a sample of 15 games conclusive, the conclusion is that The Fugitive barely dented ballpark attendance. Millions stayed home to watch ABC, but only a few hundred per city abandoned their hometown teams for this historic television event. I had not expected attendance to be that resilient, and it makes me a little proud for the sport. At least in 1967, baseball was still the national pastime, no matter what was on television.
I would have carried this question further, but TV history is against it. The resolution of the "Who Shot J.R.?" cliffhanger on Dallas—the episode that surpassed The Fugitive's record—aired in November of 1980, out of baseball season. The show that beat it, the series finale of M*A*S*H, aired in February of 1983.
Since then, the fracturing of the TV audience by scores and then hundreds of channels has made such immense audiences for TV series a thing of the past. You can only get ratings numbers like that nowadays for ... well, the Super Bowl.
If anybody wants to put a baseball game up against that, good luck to them. They'll need it.
References and Resources
Online archives of The New York Times
Baseball America for team executive rolls
Information regarding The Fugitive came from, among other sites, Wikipedia. I feel like such a slacker ...
Shane Tourtellotte is a long-time, occasionally-nominated science fiction writer, currently living in Asheville, North Carolina. He will tell you all about the baseball novel he’s shopping if you give him an inch.