May 11, 2012
Prospectus Hit and Run
Like many a Dodger fan, I found myself pulling out clumps of hair on Tuesday night. The Dodgers—a first-place team at 19-10 to that point, surprisingly—were facing the Giants (14-15) in L.A. Despite having Clayton Kershaw on the hill, they were on the short end of a 2-1 score, because with a man on base in the second inning, their ace left a high fastball to Brett Pill a bit too far out over the plate, and Pill drove it 384 feet into the left-field bleachers. The Dodgers had plated a run against Ryan Vogelsong in the bottom of the second thanks to a pair of doubles, but they could get no more, and as the innings passed, the situation grew more desperate.
Vogelsong had retired 10 hitters in a row going into the seventh inning and had thrown just 84 pitches, but the Dodgers put the first two batters of the inning aboard, and it looked as though they might finally break through. Juan Rivera, batting fifth in the lineup, smoked a single past shortstop Brandon Crawford, and James Loney smacked a single to right field. Up came Juan Uribe, who has done nothing but disappoint during his Dodger tenure; he entered the game hitting all of .246/.300/.292. Over a shot of the infield at double-play depth, Vin Scully noted that the Giants were "certainly not looking for Uribe to bunt." Indeed, a look at the portly infielder's track record showed that he had laid down exactly one sacrifice over the past three seasons, that on April 15 of this season, presumably because he misunderstood instructions from his accountant.
Track record to the contrary, Uribe did bunt Vogelsong's first pitch, and it didn't go well. The bunt hit the ground about a foot in front of the plate, and a nimble Buster Posey pounced on it in time to throw to Joaquin Arias at third base for an easy force, leaving enough time for Arias to throw to first base to complete the double play. It was a disastrous outcome, but even had it succeeded, the Dodgers would have had number-eight hitter, A.J. Ellis, at the plate, one spot ahead of the pitcher. Ellis has actually been the Dodgers' best hitter this side of Matt Kemp thus far (.296/.444/.451 coming in), and there would have been absolutely no reason to face him; an intentional walk would have loaded the bases, either bringing Kershaw (.130/.161/.130 for his career) to the plate in a non-bunt situation or forcing Dodger manager Don Mattingly to go to his bench for a pinch-hitter. As it was, Ellis batted with two outs and a man on second, popping out to end the threat.
After Kershaw set down the Giants, Vogelsong came back out for the eighth inning for the first time all season, his pitch count at 95. Bobby Abreu batted for Kershaw and worked a seven-pitch walk, and then speedy Dee Gordon bunted down the first base line. Vogelsong fielded the ball but mishandled it to the point that he couldn't get a throw off. Again the Dodgers had runners on first and second and no outs, number-two hitter Mark Ellis (.267/.378/.317 coming into the game) coming up, with Kemp and Andre Ethier to follow, and the opposing pitcher nearing the end of his night.
Even with the lesson of the previous inning in mind and the strong likelihood that opening up first base would give the Giants an excuse to intentionally walk Kemp, Mattingly ordered Ellis to bunt. He knocked one back to Vogelsong for a successful sacrifice to advance both runners. As sure as God made little green apples, Giants manager Brucy Bochy countered by ordering the intentional walk to Kemp, then brought in lefty Javier Lopez to face Ethier, who struggles mightily against southpaws (.215/.279/.329 in 516 PA from 2009-2011). As if on cue, Ethier grounded into a 6-4-3 double play, ending the inning.
Twice the Dodgers had runners on first and second and no outs. Using last year's values for run expectancy because the sample for this year's is too small, the Dodgers had a pair of situations where they could expect an average of 1.43 runs to score, with an 80.7 percent chance of scoring at least one run. A successful sacrifice that moved both runners up 90 feet would have lowered that expectancy to 1.29 runs—a 10 percent reduction—while raising the chance of scoring at least one run to 81.5 percent, and of scoring exactly one run from 42.1 percent to 46.8 percent, both very small gains. Furthermore, any gain in the likelihood of scoring had to be balanced against the possibility that something could go wrong, not only via a double play, but also a failure to advance one or both of the runners via another outcome such as a popup or a strikeout.
We've been over this before. James Click did the math years ago, searching for the break-even threshold for various common bunting situations and accounting for the chances that a sacrifice attempt may also fail or turn into a base hit or other "overachievement." Noting the possibility of the intentional walk following a successful sacrifice that opened up first base, he concluded that the threshold for a break-even play was given a hitter with a .206/.235/.263 line—worse than Uribe by nearly 100 points of OPS based upon this year's miserable numbers. Any hitter better than that was better off swinging away. He wrote:
When runners are on first and second, sacrificing is, again, not a good idea, a finding that is due almost entirely to the opposing manager's propensity to intentionally walk the next batter to keep the double play in order. This 10% decrease (approximately) in the scoring probability of the situation is enough to reduce the threshold across a great deal of current hitters.
The Dodgers lost the game, but Mattingly, despite having shot his team in both feet at point-blank range, had no regrets. "Either one of those aren't decisions that I'd look back and change," he said afterwards. "I'm going to bunt the guys over, and put my best two RBI guys up there and make them make a choice," he said of the eighth-inning situation, either oblivious to, or in denial of, the fact that there was no choice to be made once first base opened for Kemp, given Ethier's platoon differential and the presence of a competent lefty reliever in the bullpen. The manager fell into a trap without even realizing it.
Mattingly isn't a bad manager. In his first year at the helm of the Dodgers, he took a poorly-designed team that was 11 games out of first place and 10 games under .500 at the All-Star break and brought them home with an 82-79 record, a 2.5-game improvement over the previous year. He did that amid a slew of injuries whose gaps were plugged largely by middling prospects fresh from Double-A, struggling with his lineups—too much of Tony Gwynn Jr., Justin Sellers, and Aaron Miles in the top two spots—but showing a deft touch with a bullpen that had lost Jonathan Broxton and Hong-Chih Kuo. Furthermore, he did it despite the ongoing distractions brought by Frank McCourt's quarrels with Jamie McCourt, Bud Selig, and the people of the Los Angeles area, who were staying away in droves. Mattingly did far more good than bad, all things considered, and this year, he's piloted a team that is tied for the best record in the National League.
Mattingly does have a weakness for the bunt, though. From the beginning of the 2011 season through Tuesday's game, the Dodgers dropped down 187 bunts, more than any other team save for the Marlins (194). Their 92 sacrifice bunts during that timespan ranked fifth, just six off the Marlins' pace. A little more than a third of those bunts and nearly half of those sacrifices came from pitchers, though, so in order to place Donnie Buntball's fetish in its proper context, it makes sense to remove those from the equation. When we do, here's how they rank:
Under Mattingly, the Dodgers have more non-pitcher bunt attempts than any NL team, though Mike Scioscia's Angels and Ozzie Guillen's White Sox (since taken over by the less bunt-happy Robin Ventura) exceeded even that total by wide margins. The Dodger non-pitchers rank fourth in bunt hits just behind the Twins and Marlins, but a mile behind the Angels; they owe that ranking largely to the part-time work of Gwynn (12) and Gordon (11). The Halos' Erick Aybar and Peter Bourjos both have 18 bunt hits in that span, trailing only the Marlins' Emilio Bonifacio (25) and the White Sox/Phillies' Juan Pierre (26).
Even so, it's not like the Dodger hitters have been particularly productive when they bunt. If we subtract their 46 sacrifices—a total that ranks "only" eighth in the majors, behind the White Sox, Royals, Tigers, Cardinals, Brewers, Angels, and Marlins—since as sacrifices, they don't count as times at bat, and use that as the denominator and bunt hits as the numerator, we find the team's "bunting average" to be near the bottom of the pile:
The major-league "bunting average" during that span is .430. In other words, the Dodgers aren't succeeding as often as they should be when they bunt; instead, they're just finding more routes via which they can give away outs.
Unlike his preference for the hit and run, which he appears to have inherited from Joe Torre (who inherited it from Don Zimmer, who had it implanted in his skull after a severe beaning), Mattingly's preference for the bunt isn't something that was handed down from his mentor. During Torre's 2008-2010 Dodger tenure, his team's non-pitchers bunted 217 times, the majors' 16th-highest total and eighth in the NL. Their bunt hits ranked 15th, their 86 sacrifices 19th. Their .382 bunting average was 47 points below the MLB average in that span.
It's tempting to say that Torre's Dodgers had more potent offenses than did Mattingly's, and thus had less need to bunt, but it's not exactly true. The 2008 squad ranked 13th in the NL in scoring at 4.32 runs per game, the 2009 squad fourth at 4.81, the 2010 one 11th at 4.12. Adjusting for ballpark and league, those teams' True Averages were .258, .267, and .260 respectively, while last year's model was at .259, this year's at .266; basically, they've been a middling offensive club during both eras, and any argument to be made that their low-scoring environment made one-run tactics more worthwhile during the latter tenure falls somewhat flat.
For as strong a start as they've had, the Dodger offense—which ranks third in the league in TAv—still resembles Kemp and the seven dwarves; the MVP candidate is hitting .385/.469/.780, the rest of the team .244/.320/.352. Thanks to the willingness of the Ellis "brothers" to take a walk, the split isn't that dire, but four regulars—Gordon, Uribe, Rivera, and Loney—are carrying True Averages under .230, and that's not even counting the pitcher's spot. The Dodgers don't need to be giving away any more outs, so Mattingly would be well-served to curb his bunt fetish.