July 24, 2007
Prospectus Hit and Run
Welcome to Prospectus Hit & Run, a new weekly feature here at BP. Consider this the official Hit List spinoff show, starring Seymour Skinner as Skinny Boy and the late Abe Simpson as the Love-Matic Grampa. In the two-and-a-half seasons since the Hit List's inception, the column has expanded to unwieldy proportions in a manner that has sometimes seemed even beyond my control. The powers that be have accused the Hit List of being on steroids, [Ed. Note: We really just thought he'd dialed it up to 11 a bit much], so we've decided to rechannel some of the analysis and energy that's gone into the weekly Hit Lists into this new companion feature, which will mine our copious stat reserves and explore some of the trends within the game, while keeping the Hit List a more manageable entity for all of us.
Thank You For Your Support Neutral
Most of our readers are likely familiar with the sortable stat reports, particularly the individual leaderboards for VORP, SNLVAR, and WXRL. But I suspect that relatively few subscribers take advantage of the ability to customize the reports, particularly on a team level.
Particularly for this year, I've relied on two such pitching reports. The first one is a starter report containing teams' various Support Neutral statistics. Though it's probably the worst acronym in the history of sabermetrics, SNLVAR (Support Neutral Lineup-adjusted Value Above Replacement) is a fine stat which tracks a starting pitcher's cumulative win expectancy based on the runs allowed, innings pitched, and base-out situation when he leaves the game. It looks at a starter's performance independent of the run support he receives from his offense and the job done by the relievers who follow him. For a good brush-up on the Support Neutral suite, see Derek Jacques' recent Prospectus Toolbox article on the subject.
Another cool tool from the Support Neutral toolbelt is Fair Run Average (FRA), which divvies up the responsibility for bequeathed baserunners that have scored between starters and relievers depending upon the base-out situation when the former departs. Note that within our customizable reports you can obtain Fair Run Average splits for starters and relievers. The five most effective rotations in each league:
AL FRA SNLVAR Athletics 3.97 16.6 Orioles 4.29 13.7 Angels 4.53 13.1 Red Sox 4.47 13.1 White Sox 4.58 12.7 NL FRA SNLVAR Padres 3.87 16.4 Mets 4.39 14.7 Braves 4.55 13.3 Cubs 4.21 13.3 Dodgers 4.55 12.0
Already, we've got a world of stories to tell. Three of the top five AL teams here are actually under .500, and a big reason for that is the squandering of effective starting pitching. The A's rotation not only features SNLVAR league leader Dan Haren, but also seventh-ranked Joe Blanton and 10th-ranked Chad Gaudin; they're clearly getting a collective raw deal from the rest of the club. Per FRA, they're nearly a run per game stingier than the average starter (4.88 FRA). To a lesser degree, the story is the same for the Orioles and the White Sox, two teams that have been playing out the string for quite a while. The O's rotation has both Erik Bedard and Jeremy Guthrie in the SNLVAR top 10, putting a couple of feathers in Leo Mazzone's cap. The ChiSox quintet, as I've discussed more than once recently, hasn't collectively succumbed to the doom foretold by PECOTA; Mark Buehrle is fourth in SNLVAR, while Javier Vazquez is ninth, and Jon Garland 18th. Rounding out the AL top five, we see the Red Sox and Angels, two teams with good rotations that have recently leveled off a bit. Boston's most effective starter, Josh Beckett, ranks just 15th in the league at 3.3 SNLVAR, and
The NL leaderboard is a bit more straightforward. The Padres, even with pitcher-friendly Petco as their home park, have a rotation that's been the class of the league. Jake Peavy and Chris Young are second and third individually, and Greg Maddux, David Wells, and the surprising Justin Germano are all within the top 35. Not too shabby in a league with 80 theoretical rotation slots. The Mets rotation has been impressive as well, especially considering their putative ace is rehabbing in the Dominican Republic. For what its worth, John Maine (fifth),
Descending to the depths within these same rankings:
AL FRA SNLVAR Royals 5.45 7.9 Mariners 5.45 7.9 D-Rays 6.01 4.8 Rangers 6.76 1.7 NL FRA SNLVAR Phillies 5.13 8.1 Nationals 5.47 8.1 Cardinals 5.79 5.2 Marlins 5.64 4.9
At an MLB-leading 7.5 wins above their third-order projection, the Mariners are overachieving this year. What's particularly surprising is that they're doing it with such a lousy rotation, one more befitting a sub-.500 club like the Royals, with whom they're tied here. Staff leaders Jarrod Washburn and Felix Hernandez are no higher than 25th and 28th in the league at 2.5 and 2.3 SNLVAR, respectively, and while Miguel Batista (2.1) has been decent, Cha Seung Baek, Jeff Weaver, Horacio Ramirez, and Ryan Feierabend have combined to give the Mariners 40 starts worth a total of 1.0 SNLVAR. Ouch.
As this weekend showed beyond a doubt, this is home sweet home to the Devil Rays; their starters were rung up for 17 runs in their final two games against the Yankees this weekend. The 10-run pounding James Shields endured (with a little help from reliever Casey Fossum) knocked him from an impressive 12th in the AL down to 17th. But even with that bloodbath, the Rays are still comfortably ahead of the Rangers. In the 48 complete seasons in our database going back to 1959, only the 1996 Tigers starters have posted a higher Fair Run Average (7.20) than these Rangers. Not a single Texas starter is even a full win above replacement, and on a per start basis, Jamey Wright--Jamey Wright, people--has been the team's most effective pitcher thanks to a four-start string in which he yielded just six runs in 23 2/3 innings. To whichever team takes the bait at the deadline, good luck getting that out of him.
As with the AL, I had originally planned to just list the bottom three totals here, but again we've got a virtual tie between a contender and a pretender. Everyone expected the Nats to wind up down here given the cattle call for their spring rotation (where have you gone, Joel Hanrahan?). But remember when the Phillies had a surplus of experienced starters? With Freddy Garcia and Jon Lieber done for the year, and Brett Myers either in the bullpen or on the DL, they're serving up Cream of Adam Eaton every fifth day (5.84 ERA for the low, low price of $7.2 million) and hoping that Randy Lerch will return their phone calls before J.D. Durbin's chariot turns back into a pumpkin.
As bad as those two teams are, more than three wins separate them and the next two teams. If Branch Rickey, borrowing from Milton, famously said that luck is the residue of design, Walt Jocketty has conclusively proven that absent both luck (say, a healthy Chris Carpenter) and design (like signing a couple of bona fide starters to replace the departed Jeff Weaver and Jeff Suppan), you're merely left with residue. Converted closer Adam Wainwright, the team's top starter, ranks just 22nd in the league, and his 2.9 SNLVAR account for more than half of the Cards' total. The same can almost be said for 27th-ranked Sergio Mitre (2.4 SNLVAR) and the not-so-fresh Fish accompanying him in the rotation; only
You're Listening to WXRL
Did you know there's actually a radio station sharing the call letters of our newfangled relief stat (which stands for Win Expectation above Replacement, Lineup-adjusted)? It's true: WXRL 1300 AM out of Lancaster, New York is a country music station, but looking over their playlist full of hat acts, it's one that might cause Hank Williams and Johnny Cash to spin in their graves. Having said that, I gotta believe Tug McGraw, father of country star Tim McGraw and the National League's WXRL leader while pitching for World Champion Mets and Phillies clubs, might see things a little differently.
My customized team WXRL report reveals which bullpens have most of their oars pulling in the right direction. Rather than dissect those rankings in parallel to what we did above, we'll look at things a bit differently. When Keith Woolner introduced it a few years ago, WXRL pushed another fine stat, Adjusted Runs Prevented, into the shadows. Whereas ARP accounts for the base-out situation in which a reliever inherits runners by treating them equally regardless of inning or relative score, WXRL incorporates leverage and the team's expected chances of winning into the mix (once again, Derek Jacques has your brush-up). Often the two stats are more or less in agreement, but sometimes they're not; a team may be doing a decent job of dealing with inherited runners as a whole, but a few high-leverage failures can throw their WXRL out of whack. Here are the teams' respective rankings (1-30) in both categories:
Team WXRL ARP Difference D'Backs 4 14 10 Tigers 17 25 8 Indians 11 18 7 Brewers 8 15 7 Braves 15 21 6 Pirates 16 22 6 Cardinals 13 19 6 Angels 12 17 5 Astros 23 26 3 White Sox 27 29 2 Phillies 25 27 2 Nationals 7 9 2 Red Sox 1 2 1 Dodgers 5 6 1 Mets 10 11 1 Mariners 3 4 1 Reds 28 28 0 Devil Rays 30 30 0 Twins 6 5 -1 Padres 2 1 -1 Royals 14 12 -2 Rockies 26 23 -3 Marlins 20 16 -4 Athletics 24 20 -4 Orioles 29 23 -6 Rangers 9 3 -6 Blue Jays 18 10 -8 Cubs 22 13 -9 Yankees 19 7 -12 Giants 21 8 -13
For just about half of the teams (14 out of 30), the difference between the two lists is trivial; they're no further than three spots apart. What the list is saying for the rest is that relative to their overall bullpen performance, the teams at the top have done a better job of rising to the occasion than the ones at the bottom; they've especially taken advantage of their high-leverage situations. Note the lead here is held by the Diamondbacks, who are an NL-best 7.4 wins above their third-order projection, and that nine of the top 10 teams-all but the Braves-are ahead of their own third-order projections. At the other end of the scale, we've heard plenty about the bullpen failures of the Cubs and Yankees at various times this year, but the Giants? Between the minute-to-minute updates on Barry Bonds, the zombie lineup around him, and the solid but ill-supported rotation (including the fascinating Tim Lincecum), the bullpen has been pretty low on the list of things to pay attention to out by the bay. Then again, there's a reason Armando Benitez was banished to Florida, and it ain't the dominance of his replacement, Brad Hennessey.
The Close Ones
What's especially interesting about the list above is that it appears to correlate pretty well with team records in close games, that is, those games decided by one or two runs runs. Here's that info; +/- is the number of games above or below .500 that a team is in this category:
Team W-L +/- Indians 29-20 9 Brewers 25-16 9 Mets 22-13 9 D'backs 29-21 8 Mariners 23-15 8 Dodgers 27-20 7 Cardinals 20-14 6 Angels 20-15 5 Rockies 26-21 5 Tigers 25-20 5 Red Sox 23-19 4 Athletics 25-21 4 D-Rays 24-21 3 Braves 22-20 2 White Sox 20-20 0 Marlins 22-22 0 Padres 25-25 0 Nationals 25-25 0 Cubs 22-25 -3 Twins 21-24 -3 Pirates 23-26 -3 Rangers 16-19 -3 Blue Jays 24-27 -3 Royals 19-25 -6 Phillies 15-21 -6 Astros 20-27 -7 Reds 18-28 -10 Orioles 18-31 -13 Yankees 13-26 -13 Giants 20-34 -14
In fact, that list showing the differential rankings of WXRL and ARP correlates especially well with the teams' records in close games (r = .65), better than either the WXRL (r = 0.5) or ARP rankings (r = 0.06) alone. Deadlines and space issues prevent me from launching a deeper inquiry to see how well this 90-something game trend holds up under larger sample sizes; we'll hit this topic again at a later date. Meanwhile, the moral of the story is never underestimate the importance of a good bullpen in helping a team win the close ones.
Who Wants to Go Streaking?
Quick, what's the longest hitting streak this year? You're probably thinking it was Ichiro Suzuki-he's always a good guess-or maybe that guy with the, uh, pants. It's actually Casey Blake, at 26 games, and while that's cool, bookmarking this link to the increasingly awesome Baseball-Reference.com Play Index so that you can check out the leaderboard instead of pretending to work is far, far cooler.
Mailing It In
Hello Jay, I always read your 'Hit List', eagerly anticipating the day my beloved Washington Nationals rise above 30th place. Okay, they spent a few weeks above, but now they seemed doomed to last forever. Since starting the season 9-25, they are 33-31, including 9-9 vs. the much vaunted American League. At some point, shouldn't actual games won factor into the rankings? Regards, TM
TM, actual wins do factor into the Hit List in the form of actual winning percentage. It's 1/4 of the recipe, along with first-, second-, and third-order winning percentages. And that's the real problem for the Nats-through the games leading up to last Friday's Hit List, they had the 28th-best actual winning percentage. They also had the 29th-best first- and third-order winning percentages, and they were dead last in second-order. All told, they missed leaving the basement by just .003, but you can cheer up. It's a good bet that the severe beatdowns the Devil Rays received at the hands of the Yankees will be enough to help your beloved club assert itself as the Hit List's 29th-best team next week. We'll see about that come Friday.