September 23, 2009

Prospectus Hit and Run

The Underachieving Class of 2009

by Jay Jaffe

Printer-friendly

Last week, I checked in on the Angels' quest to make sabermetric history. Examining their third-order Pythagenpat projection-their Pythagorean record based upon their hits, walks, total bases, stolen bases, and outs of all kinds, as well as those of their opponents, all adjusted for park, league, and quality of competition-I discovered that they were on their way to becoming the first team to finish at least 10 games above their third-order projection for a second year in a row. At the time, the Halos' D3, the difference between their third-order wins and their actual wins as published in our Adjusted Standings report, was 11.5 games, but I noted that suggesting they were "on pace" for an even higher mark was a misnomer, since teams that are outperforming their Pythagorean records by wide margins in either direction tend to regress to the mean. As if on cue, that D3 has shrunk to 10.1 games at this writing.

Meanwhile, there's also potential history being made at the other, less happy end of the Pythagorean spectrum. Since 1901, twenty-five teams have finished at least 10 games below their third-order Pythagenpat projection. Only twice have two teams done so within the same year, first time in 1912 (when both the Brooklyn Dodgers and Boston Braves achieved ignominy), and then again in 1993 (when the Mets and Padres did it). This year, no less than four teams are threatening to join those ranks, including two from the same division:

Rnk  Year Team         W-L    Pct    R    RA   AEQR  AEQRA    D3
 1   1993 Mets       59-103  .364   672   744   672   736   -15.1
 2   1935 Braves     38-115  .248   575   852   593   835   -14.6
 3   1986 Pirates    64-98   .395   663   700   666   697   -13.6
 4   2009 Nationals  51-99   .340   661   825   664   773   -13.2
 5   1946 A's        49-105  .318   529   680   529   662   -12.8
 6   1905 Browns     54-99   .353   512   608   521   601   -12.7
 7   1937 Reds       56-98   .364   612   706   620   700   -12.4
 8   1939 Browns     43-111  .279   733  1035   752  1003   -12.2
 9   1962 Mets       40-120  .250   617   948   631   924   -12.1
10   1917 Pirates    51-103  .331   464   595   468   579   -11.9
11t  1975 Astros     64-97   .398   664   711   668   711   -11.8
11t  1984 Pirates    75-87   .463   615   567   612   564   -11.8
13   2001 Rockies    73-89   .451   923   906   910   870   -11.5
14   1993 Padres     61-101  .377   679   772   681   764   -11.4
15   2009 Blue Jays  68-83   .450   727   719   745   714   -11.3
16t  1924 Cardinals  65-89   .422   740   750   745   752   -11.1
16t  1961 Phillies   47-107  .305   584   796   599   782   -11.1
18   1907 Reds       66-87   .431   526   519   527   522   -11.0
19   1967 Orioles    76-85   .472   654   592   657   602   -11.0
20   1936 Phillies   54-100  .351   726   874   739   869   -10.9
21   2006 Indians    78-84   .481   870   782   882   800   -10.7
22t  1912 Dodgers    58-95   .379   651   744   665   742   -10.4
22t  1952 Tigers     50-104  .325   557   738   563   716   -10.4
23   2009 D'backs    66-86   .434   678   735   693   690   -10.3
24   1919 Senators   56-84   .400   533   570   533   565   -10.2
25t  1912 Braves     52-101  .340   693   871   705   857   -10.1
25t  1928 Phillies   43-109  .283   660   957   682   936   -10.1
25t  1972 Giants     69-86   .445   662   649   662   648   -10.1
30t  2009 Rays       77-74   .510   748   691   774   662    -9.6

Recall that the overachievers list skews towards recent history, with the Wild Card era producing eight of the 21 teams who have finished at least 10 games above their expected records. This one, on the other hand, tilts heavily towards the pre-World War II era, producing 12 of the 25 who've finished at least 10 games below their expected records. Not counting this year's bountiful class, just two of the top underachievers are from the Wild Card era.

The main reason for that, I suspect, has to do with bullpen usage. As noted last year and again in last week's piece, a strong bullpen is a consistent means of such overachievement; the historical correlation between a team's cumulative WXRL and its D3 is .42, whereas it's just .20 for SNLVAR. It makes some amount of sense that the current era might produce more overachievers and fewer underachievers because of the fact that WXRL rates and Leverage scores have been on the rise historically, as bullpens have assumed a higher percentage of innings and increased specialization has tailored more specific roles than 20 or 30 years ago:

Note that Bruce Sutter's advent as the modern closer marks something of a turning point here. WXRL rates rose above 0.1 per nine innings only four times from 1954 through 1979. By that point, Cubs manager Herman Franks had begun his attempt to limit Sutter's deployment to close games in which the Cubs had a lead-save situations, in other words. The strategy began to take hold, and the only time WXRL rates have been below 0.1 per nine innings since was in the 1981 strike year. They're now about 40 percent higher than they were 30 years ago.

Anyway, from the overachiever group, 15 of the 16 teams hailing from the Retrosheet era (1954 onward) finished in the league's top three in WXRL. There's a similar consistency to the underachiever group:

Year   Team       WXRL    Rank
1961   Phillies    1.6    6/8
1962   Mets       -1.1   10/10
1967   Orioles     9.3    2/10
1972   Giants     -1.8   12/12
1975   Astros      0.8   11/12
1984   Pirates     2.8    9/12
1986   Pirates     2.1   11/12
1993   Mets       -3.1   14/14
1993   Padres      5.5   10/14
2001   Rockies     0.1   14/14
2006   Indians    -1.5   13/14

Of the 11 underachievers from the Retrosheet era, eight finished in the bottom three of their league, two others finished with the fourth- and fifth-worst rankings, and then there are the 1967 Orioles, the list's odd ducks. The defending World Champions at the time, they outscored their opponents by a handy 62-run margin via the league's second-best offense according to EqA. Their bullpen ranked second in WXRL, and their third-order projection called for 87 wins. Yet their rotation ranked ninth in SNLVAR in a 10-team league, they were an anemic 33-55 in games decided by one or two runs, and their differential was further distorted by the fact that they went 18-8 in games decided by six or more runs.

The four "Class of 2009" underachievers certainly fit the schema, as each of them ranks in the bottom four in their league in WXRL. Taking a closer look at what's gone wrong with their bullpens:

• The Rays rank 11^th in the AL with 4.4 WXRL, more than 10 wins off last year's league-leading performance, as J.P. Howell and Grant Balfour have regressed towards league average, and newcomers Brian Shouse, Joe Nelson, and Russ Springer have been subpar or worse-a reminder of just how rare it is for relievers to remain significantly above-average from year to year. Mirroring the tendency of overachieving teams to win the close ones but do less well in blowouts, the Rays are 17-23 in one-run games, 34-35 in games decided by one or two runs, and 21-10 in games decided by six or more runs. What's most interesting about them relative to this list is that they might be the first team to finish .500 or better while still falling 10 games shy of their third-order projection, and at the very least, they need just two more wins to surpass the 2006 Indians as the best team to make the list. So they've got that going for them.

• The Blue Jays rank 12^th in the AL with 4.2 WXRL. Their bullpen has certainly seen its share of upheaval this season, starting with B.J. Ryan losing his job as closer and then being released just prior to the All-Star break with about $15 million still remaining on his contract. Scott Downs was passable as closer before getting hurt in early August, while Jason Frasor, who took over, has pitched well enough to rank 14^th in the league in WXRL. Few of the other relievers have distinguished themselves, however, which helps explain why the Jays are 17-26 in one-run games and 15-28 in games decided by two or three runs, but 18-12 in games decided by six or more. These Jays might become members of the select subset of this group to outscore their opponents; prior to this year, just five teams have done so.

• The Diamondbacks rank 14^th in the NL with 1.9 WXRL. Their stumble out of the gate prompted the early July trade of top set-up man Tony Peņa, then in late August they traded Jon Rauch and lost Chad Qualls for the year. None of them had been terribly effective to begin with, but neither were they replaced by anyone who did much better, though rookie Juan Gutierrez has gone five-for-five in converting his save opportunities since Qualls went down. Elsewhere, the team has two of the five worst relievers in the league in terms of WXRL in Esmerling Vasquez and Daniel Schlereth, both with around -1.2 WXRL. The Snakes are 20-27 in one-run games, 12-15 in two-run games, and 15-17 in games decided by six or more runs, though they're 3-0 in games decided by 10 runs or more. Their additional mark of distinction is that they're in the black as far as their third-order run differential is concerned; prior to this year, just six teams had finished with positive third-order differentials while still falling at least 10 games shy of their projected record.

• The Nationals are dead last in the NL with -1.2 WXRL, a dismal showing that would rank 53^rd for all-time badness. It's been one debacle after another for the poor Nats, who with one more loss will reach the century mark for the second year in a row. They left the gate with closer Joel Hanrahan, who blew three of his first five save opportunities and compiled a 7.71 ERA and -0.6 WXRL before being mercifully exiled to Pittsburgh. By that point, in late June, the team had fallen into a double-digit third-order deficit for good, making them possibly the fastest team to do so. Hanrahan had plenty of help, of course: top set=up man Saul Rivera, who'd been a durable and reliable reliever from 2006 through 2008, was torched for an 10.24 ERA in April and sent to Triple-A for two months to get his act together; his -1.2 WXRL is the third-worst mark in the majors; teammate Jesus Colome is seventh at -1.1. In all, nine Washington relievers with at least 10 appearances have negative WXRLs, including the usually competent Julian Tavarez, Ron Villone, and Joe Beimel. Meanwhile, Mike MacDougal has sort of resurrected his career by converting 16 out of 17 save opportunities, albeit with a 29/35 K/BB ratio and an ugly September in which he's yielded runs in five of seven appearances, doubling his season ERA to 3.88. All told, the Nats are 17-23 in one-run games, 10-22 in two-run games, and 9-15 in games decided by six or more runs, including last night's 12-run drubbing by the Dodgers.

It's also worth noting that offense has played a part in these teams' underachievement as well. The Jays, Diamondbacks, and Nationals have all hit worse than the major league average (in terms of OPS) with runners in scoring position, and in the Close and Late category, while the Snakes, Nats, and Rays are all below average with two outs and runners in scoring position. Exploring the link between Pythagorean over- and underachievement and the hitting end of things remains is something I'm planning to do more fully once the season is over. In the meantime, though the playoff races are more or less sewn up now, between the Angels and these four teams, there's plenty to watch on our Adjusted Standings page.