September 21, 2008

Prospectus Today

No Fish Story in Tampa Bay

by Joe Sheehan

Printer-friendly

Contact Author

In their 11^th season of play, after a decade of never winning 71 games, of never finishing above .500, and never coming close to contention in any year, the Tampa Bay Rays clinched a playoff spot yesterday with a 7-2 win over the Twins. The win was their 92^nd of the season, 22 more than they'd ever won before. With a week left in the regular season, the Rays, though nominally battling the Red Sox for the AL East title, can work on setting their post-season rotation, rest their battered lineup, and prepare for the first Division Series in their history.

I don't get to brag about this. While Nate Silver's PECOTA projection system famously saw this coming, pegging the Rays to go 88-74, I distanced myself from its optimism and called for an improvement to 77-85. At the time, I felt I was being pretty generous to a team that had won 66 games in 2007, had made few major acquisitions, and was playing in a division with the Yankees, Red Sox, and Blue Jays. What primarily drove the prediction, however, was this idea: the Rays allowed 944 runs in 2007, and there seemed to be a limit as to how many fewer runs a team could allow from one year to the next. It didn't seem reasonable that the Rays could lop off enough runs in one season to go from one of the worst teams in baseball to one of the best.

In May, I examined the concept further.

Well, now it's May 1, and the Rays are on pace to allow 654 runs. Some of that is a league-wide downturn in offense, but more of it is defense. In 2007 the Rays had the worst Defensive Efficiency in the 49 years in our database, but they now have the third-best defense in baseball by that metric this time around. That points to a concerted effort to upgrade the defense, to make personnel decisions based on getting more outs on balls in play. That's paid off in allowing nearly two runs a game less in 2008.
...

Just to allow 744 runs, the Rays would have to lop 200 runs off of their 2007 mark. Jason Paré looked it up and found just 15 teams that have allowed 200 fewer runs from one season to the next since 1959:

Team     Years      RA    RA2   Diff.
DET    1996-97    1103    790    313
SDN    1997-98     891    635    256
SDN    1977-78     833    598    235
CHN    1962-63     806    578    228
CLE    1971-72     747    519    228
HOU    1970-71     763    536    227
CLE    1987-88     957    731    226
CHA    1970-71     822    597    225
BAL    1996-97     903    681    222
OAK    1979-80     860    642    218
CLE    2004-05     857    642    215
SDN    1970-71     788    582    206
CHN    2000-01     904    701    203
MIL    1987-88     817    616    201
HOU    1967-68     678    477    201

...

The point is that a team improves by 200 runs on defense about once every three or four years, which is why I remain skeptical about the Rays' ability to do so.

The Rays have allowed 628 runs with nine games to play. They won't set the record for one-year improvement, but they should manage to finish second to the 1997 Tigers. That's an amazing feat.

One reader, R.R., has consistently objected to this analysis of the Rays:

I don't see how it matters whether it is unusual for teams to lop 200-plus runs off their runs allowed from one year to the next. That fact might simply be because teams do not ordinarily turn over personnel that drastically, or that the focus tends to be on changing offense rather than defense, or perhaps some other factors.

It seems to me that the proper approach is to compare the run-prevention sources of the Rays in 2008 to those of 2007. Is Bartlett significantly better than the group that played shortstop last year? Is Iwamura an upgrade over last year's second basemen or Longoria an upgrade over Iwamura at third? Has Upton improved in center field and is Gross a better outfielder than Young?

And similarly, is the rotation and the bullpen significantly better? If the answers to most of those questions is yes, what difference does it make if there is little history for such a dramatic turnaround? It seems to me that is relying on a somewhat irrelevant factoid, almost like a superstition rather than analysis.

To his credit, R.R. made this point a number of times throughout the season. My argument is that the list above represents what we might call the range of reasonable outcomes. We don't predict teams to win 122 games, or pitchers to have 0.88 ERAs, or batters to hit 81 home runs. When predicting the future, we are largely bound by what has happened in the past, and by the range of reasonable outcomes. Perhaps it was reasonable to project the Rays to allow just 660 or so runs this year, and to have the second-biggest improvement since Eisenhower was in office, but that's not a prediction I felt comfortable making. It didn't seem reasonable.

What the Rays have done this year is historic, and full credit should go to their management team, their field staff, and their players. Their Defensive Efficiency is the best in MLB, a year after that database-worst mark. The improvement has helped save more than 300 hits as opposed to last season: the Rays allowed 1,450 singles, doubles, and triples last year; that figure is 1,120 this year with a week to go. Rays pitchers haven't hurt themselves quite as much either, lopping off about 70 walks and 40 homers this season. The missing hits, though, prevented by a defense that was better at nearly every position, is why the Rays will allow fewer than 670 runs, why they will win 95 or more games, and why we're about to find out if the baseball fans of central Florida will show up for post-season games. There is no better story in baseball this season.

Joe Sheehan is an author of Baseball Prospectus. 
Click here to see Joe's other articles. You can contact Joe by clicking here