Friday, October 2, 2009
Unseen differences in performance: The 2007 Rockies and Diamondbacks
Jeff Passan notes the divergent paths to 2009 between the two NL West 2007 playoff teams, the Rockies and Diamondbacks, and how the Rockies have returned to the spotlight while the Diamondbacks have fallen off the map. Passan laments one of the best young teams in baseball in the Diamondbacks sputtering while the Rockies returned to glory after an off 2008. A closer look shows that, while indeed one team has sputtered while another thrived, the difference between the two teams was evident even during their dual postseason runs.
In 2007, the Diamondbacks finished with the 3rd fewest runs scored and only the 5th best run prevention in the NL. At 712 runs scored and 732 allowed, their pythagorean record indicated they ought to have finished somewhere around 79-83, well off the pack... not 90-72 and 1st in the NL West. Brandon Webb fronted an unspectacular rotation: Livan Hernandez (4.93 ERA), Doug Davis (4.25), Micah Owings (4.30) and a spate of mediocre others alongside an injury-prone Randy Johnson (3.81, but only 10 starts and 56.2 IP), whose season was done after June.
In August the D-Backs took advantage of a stretch against weak opponents (Pittsburgh, Washington, Florida) to help ring up a 16-12 record despite getting outscored 141 to 127. In June they eked out a 14-13 record despite getting outscored 127 to 108, and went 8-7 in Interleague play despite getting outscored 80 to 71. With the help of such fortuitous stretches, they went 43-29 after the All Star Break to propel them to the NL West title.
How did they get so lucky? Like many teams that defy their pythag projected record, the D-Backs had one of the best bullpens in baseball that year, which helped them in close games. Their top five relievers had ERAs no worse than 3.27 (Tony Peña Jr.), and three of those five, including closer Jose Valverde, had sub-3 ERAs. Go figure they ran a solid 32-20 record in one run games. Arizona also led the NL with 51 saves (as Valverde rang up a career high 47) while only blowing 15 saves. Once the D-Backs had a lead, a deep bullpen did a good job of making sure they didn't (usually) give it up.
Now granted, if you crunch the numbers and consider the history of those pitchers you'll find the bullpen itself wasn't necessarily spectacular. The defense was a key reason they ruled it so hard: Arizona's defensive runs saved above average per Baseball Reference was +27.4, good for 4th in the NL. The key contributors were outfielders Eric Byrnes (+20.3) and Carlos Quentin (+11.5), and while CF Chris Young was the biggest liability (-13.3), having two effective corner outfielders on each side helped neutralize any shortcomings, plus defensive replacements Scott Hairston (+4.5) and Jeff Salazar (+4.5). Aside from Stephen Drew at SS (-7.2), no other D-Back cost his team more than 4 runs in the field. The bullpen had an added edge not just with the good corner outfielders, but Young could be replaced late in the game by a decent backup, turning the outfield into death for flyballs.
As for the lopsided run differential, here's an interesting breakdown:
Wins by 4 + runs: 30
Wins by 1-3 runs: 60
Loss by 1-3 runs: 36
Loss by 4 + runs: 36
While there's a fairly even ratio of routed wins to routed losses, with a few more rout losses, there are far more closer 1-3 run wins than losses. This is where the bullpen and defensive replacements came in: Once the D-Backs got a slim lead, they were able to hold it far more often than not. But they did not frequently dominate, nor did they frequently get dominated. Had they had more even fortune in those close games, their record might have been closer to that projected 79-83 mark and I probably am not discussing them.
Meanwhile, the Rockies finished one game back at 89-73, deadlocked with the San Diego Padres and making the playoffs only after winning a playoff game with help from a dubious sac fly run from Matt Holliday. Their Pythag record walks the walk (860 scored, 758 allowed, 91-72 expected), as did the Padres record (741 in and 666 out for a 89-74 expected record that exactly matched their final record). The Rockies were 2nd in the league in runs scored with obvious help from Coors Field, scoring 478 runs at home but a more average 382 on the road, while going 51-31 at home and a rather mediocre 39-42 on the road.
Colorado's run prevention was actually a tad above average (4.65 runs allowed; the NL average was 4.78). Their .701 defensive efficiency was 2nd best in the NL. Their defense saved an NL best 69.5 runs. They too had an effective bullpen, but even deeper, with six regulars finishing with ERAs below 4, even though none of their starting pitchers had an ERA below 4. Their lineup tied for the NL lead with a team .354 OBP. Their slugging and OPS, like their runs scored per game (5.28), were 2nd to Phialdelphia, which plays in one of baseball's hitter friendliest parks in Citizens Bank Park. This team was clearly, in most ways, better than the Diamondbacks.