Scott Behson predicted another second half swoon for the Pirates. It takes a good man to admit when he’s wrong.
Happy to be So Wrong About the 2013 Pittsburgh Pirates
The thing that bugs me about most sportswriters is that they constantly make incorrect predictions but they never own up to their mistakes.
Well, a few months ago, I wrote a piece here at GMP Sports in which I explained the reasons why I expected the Pirates to suffer yet another second-half collapse and miss the playoffs. I was spectacularly wrong. But I’m not running away from my mistake. In this article, I’d like to review my prediction, explain why I was wrong, and why I am happy to be so wrong (yay, Pittsburgh!).
In making my case for yet another late-season swoon for the Bucs, I argued that the Pirates were too reliant on unproven starters and were overusing their bullpen. I should have known better. In retrospect, the very sabermetric tools I used to craft the article proved my undoing.
Looking back, one of the reasons that the Pirates were able to avoid a second-half slump and make their first playoff appearance in a generation was because of how they embraced sabermetrics and creatively integrated them into their game-day management. Here’s how they did it:
The 2012 Pirates were a well-below average in team fielding (they were dead last in 2010), but the 2013 version boasted the 4th best defense in baseball, utilizing most of the same players. What changed was how the Pirates used data. Specifically, they made extensive use of spray charts and opponents’ hitting data, and employed extreme defensive shifts as often as any other National League team—five times as often as they had the year before. As a result, it is estimated that they saved 68 runs more than they would have, given an average defense. That’s almost like starting every game with a half-run advantage before taking hitting and pitching into account.
This defensive strategy would not have worked if not for the starting pitching staff I so rudely maligned. The Pirates took some chances on reclamation projects like AJ Burnett, Francisco Liriano, Mark Melancon and Jason Grilli, who shared one important tendency. These were pitchers that induced a lot of ground balls. The Pirates also coached their up-and-coming pitchers in ground-ball strategies, such as the art of the two-seam fastball. Ground balls are converted to outs at a way higher rate than line drives, and, unlike fly balls, carry no risk of becoming homeruns. Ground balls are a pitcher’s second best friend (after strikeouts).
As a result of this two pronged strategy, the Pirated forced their opponents into hitting ground balls right into the teeth of their defensive shifts, leading to lots and lots of harmless outs. In fact, their ground ball rate of 52.5% is the highest since this stat was first calculated in 2002. In this case, the pitching and defense worked together beautifully, a true melding of advanced statistics and good old-fashioned management.
Also helping their pitching staff was the acquisition of defensive wiz, catcher Russell Martin. His 2 year/$17 million dollar contract may have looked like an overpay based on his hitting, but Martin is a master at getting borderline pitches to be called strikes. In fact, it is estimated that his framing of caught pitches led to about 30 runs saved over the course of a season. Combine this with the defensive shift and we’re at about 100 runs saved over an average team.
The bullpen I was so worried about? It turns out there was a plan here as well. Recent sabermetric research shows that starting pitchers tend to have the advantage the first two times through the batting order, but that the advantage shifts to the hitters (due to pitcher fatigue and the ability of hitters to make adjustments) the third and especially the fourth time through the order.
The Pirates are the first team to take this lesson so close to heart. They intentionally pulled starters even when they were pitching well into the 6th or 7th innings. In fact, their starters averaged only 5.7 innings per start, the second lowest in the NL (this is normally an indicator of poor pitching staffs, but in this case, the Pirates starters were excellent—top 5 in the NL in ERA, WHIP, ERA+, strikeouts and fewest homeruns allowed.
Pulling even effective starters when they come to bat in the 6th inning carries certain advantages. They don’t have to pace themselves as much, knowing they will not be asked to pitch a complete game. Further, you have the benefit of a pinch-hitter taking their at-bats (over the course of a season, reducing pitcher at-bats by 1/3 and giving them to actual hitters should add up to some extra runs scored).
The danger in giving starters an early hook is that you need a lot of depth in the bullpen. Here’s where the Pirates shined brightest. They had an excellent and deep 6-man bullpen, none of whom were overused (none of their six top relievers pitched over 80 innings). Together, they combined for a 2.89 ERA over 545 innings, and Manager of the Year (and deservedly so) Clint Hurdle did a great job keeping everyone fresh and effective.
So, the two flaws I saw in the Pirates turned out to be strengths, all because the Pirates understood and applied sabremetric lessons more extensively than I ever imagined.
I was wrong.
…But I couldn’t be happier about it. Long-suffering Pirates fans finally had a winner to cheer for. The Pirates won the NL Wild-Card game in what was the best atmosphere for a baseball game this season (perhaps only rivaled by Boston in the World Series clincher), and then gave the Cardinals all they could handle before losing in the NLCS. Finally, playoff baseball for Pittsburgh and its amazing fans. Congrats, and no more predictions of doom and gloom out of me. Lesson learned.