In the oft-unfiltered Twitterverse, I suspect that Miguel Batista's name reached a trending crescendo at points far later during the Opening Day game than most had anticipated. During the game, in a fit of frustration that was boiling into anger, I labeled him the villain of 2011 as he flirted with the idea of throwing strikes. In retrospect, that was unfair. Batista did not call down to the bullpen and order himself to get loose. Batista did not summon himself from the 'pen to throw during a tied eighth inning. And Batista did not leave himself in the game too long after proving ineffective. Yes, in retrospect, Miguel Batista is more Oddjob to Tony La Russa's Goldfinger, a henchman who dutifully carries out the mastermind's bidding.*
*It is not lost on me that Oddjob is far more accurate with a hat than Batista is with a baseball.
There is no question that bullpen usage is a constantly fluctuating work in progress until La Russa settles on his roles. Even then, the necessities of a 162-game season throw the manager and his bullpen into situations that buck usage trends. Nonetheless, in the nail-biter Opening Day loss and the blowout loss on Saturday, La Russa's usage of his relievers was suprising.
Using Leverage Index and WPA, I wanted to demonstrate why I found Thursday and Saturday surprising. These two statistics can be used to gauge relievers and, to be honest, I found each of them rather daunting when first introduced to them. For those of you vaguely familiar or just getting introduced, here is a quick explanation for each.
The Fangraphs Saber Library gives us this description of Win Probability Added (WPA):
Most sabermetric statistics are context neutral, meaning they disregard the fact that during the course of a game some plays are more crucial than others. While wOBA rates all home runs as equal, we know intuitively that a home run in the third inning of a blowout is less important than a home run in the bottom of the ninth inning of a close game. Win Probability Added (WPA) captures this difference by measuring how individual players affect their team’s win expectancy on a play-by-play basis.
****
Cumulative, season-long WPA is not predictive and should not be used to make projections about a player’s talent, but it is a good describer of what has happened.
And here is the Fangraphs Saber Library description of Leverage Index (LI):
During the course of a game, some situations are more tense and suspenseful than others. For instance, we know that a one-run lead in the bottom of the ninth inning is more suspenseful than a one-run lead in the top of the third inning. Batting with two runners on and two outs in the eighth inning is filled with more pressure than batting in the same situation in the second inning. Leverage Index (LI) is merely an attempt to quantify this suspense so we can determine if a player has been used primarily in high-leverage or low-leverage situations.
****
An average (or neutral) LI is one, high leverage is 1.5 and above, and low leverage is below one.
Miguel Batista's WPA and LI tell a story over the last few seasons of a player who probably should not have a very high LI and is likely to have a relatively low WPA.
Heading into the 2007 season, Batista was a free agent in whom the Cardinals reportedly had some interest. The Mariners won the Batista sweepstakes and used Batista in their bullpen much the same way that Kyle McClellan was used in the St. Louis bullpen. [AUTHOR'S NOTE: I foolishly went by my memory.] Batista was used as a starter in by the Mariners in 2007 and was shifted out of the rotation and into the bullpen in 2008. In 2007, Batista's average LI (aLI on Baseball-Reference and pLI on Fangraphs) was 1.01 and a WPA of 1.14. In 2008, Batista's pLI rose a bit, to 1.10. It was 1.29 on average when he entered a game (gmLI on Fangraphs), so Batista was a fireman, of sorts, coming into high-leverage situations. That year, however, Batista was not so much putting out fires as he was helping to stoke them. Batista put up a WPA of -4.33. In 2009, Batista's aLI was 1.06 and his WPA was -1.27.
By way of comparison, Kyle McClellan had an pLI of 1.58 in his rookie season of 2008 and posted a -0.08 WPA. In 2009, K-Mac's pLI was 1.43; his WPA, 0.16. Last season, McClellan had a pLI of 1.16 and a WPA of 0.94. Comparing McClellan's time as a Cardinal reliever to Batista's time as a Mariner reliever, McClellan was used in higher leverage situations than was Batista. After 2008, when he faded down the stretch of his inaugural campaign, McClellan helped the Cardinals' chances of winning whereas Batista hurt the Mariners' chances of winning.
Given Batista's time in the Seattle bullpen, it is not much of a surprise that last season Jim Riggleman deployed the agingly ineffective righty in a meritorious manner--as Wonderpitcher. The Nationals manager primarily looked to Batista when the game was not close. Pitching for Washington, this means that Batista made 57 appearances yet his 2010 average Leverage Index was 0.651. His average LI (gmLI) when entering a Nationals game was 0.77. The highest pLI Batista had in an appearance for the Nationals during the 2010 season was 2.77, on season's final day.
Enter Batista on Opening Day, at La Russa's beck. The LI was 1.83. (A tie-game in the eighth inning with the bases empty gives the situation a pretty high LI.) This is the type of situation, La Russa typically would look to McClellan or Jason Motte. With McClellan in the rotation and unavailable, La Russa opted for Batista over Motte. By the time Batista had finished mixing in two outs with two hits and a walk, the LI had ballooned to 5.68. The pLI for Batista on Opening Day was 2.18. Luckily, Trever Miller came in and doused the fire, recording his first hold of the 2011 campaign.
On Saturday, Jason Motte came in to attempt to stop the bleeding of Jake Westbrook's walk extravaganza. The gmLI for Motte's appearance was 0.97, and it was the point in the game where the Cardinals would remain in striking distance or the Padres would put the contest out of reach. Needless to say, the Padres delivered what was effectively the death blow with Motte on the hill. Then Mitchell Boggs came in to a gmLI of 0.05 and finished out the 3 innings left in the game, seemingly performing the job most imagined to be Batista's.
The season is a series old, far too early to draw any conclusions about how La Russa will deploy his relief corps for its duration. Nonetheless, the early usage is troublesome as he appears to be leaning toward replacing Kyle McClellan with a pitcher who has not been particularly effective in 4 years. Over the coming weeks, it will be interesting to see whether the 40 year-old continues to be used in such pivotal situations.