October 7, 2004
The Shape of Seasonal Performance
To find out, the season has to be broken down into smaller samples of data. When analyzing the season, however, dividing the year into pieces for study can yield warning calls from the wings. "Small Sample Size!" they yell, and "Arbitrary date constraints!" they cry. It becomes difficult to carve the season down into pieces to examine how players rise and fall as half-seasons or less of information don't lend themsleves to quality analysis. Even an entire season isn't long enough to truly differentiate among the great majority of players in terms of their true ability.
To keep the sample sizes large enough, I'll break down the season into just two pieces, conveniently splitting things up at the All-Star break when three days of fishing, golfing, or whatever players do during their only vacation of the summer provides a nice natural split. To weed out the players who have turned things around the most this season, let's look at a quick estimate of runs generated using the handy formula 27*OBP*SLG/(1-AVG) for all players who accumulated at least 100 plate appearances in each half. (While we could take PA into account, the fact that there are more games in the first half skewed the results, so these results are prorated over 27 outs.)
The most positive turnarounds this season:
PLAYER PRE-ASB POST-ASB DIFF J.T. Snow 5.1 14.1 9.0 Ichiro Suzuki 5.8 11.3 5.5 Carlos Delgado 4.8 9.9 5.1 Ross Gload 4.6 9.7 5.1 Craig Monroe 4.3 9.4 5.1 Juan Rivera 4.2 9.2 5.0 Luis Gonzalez 4.3 9.2 4.9 Ben Broussard 4.7 9.0 4.3 Adam LaRoche 4.0 8.2 4.2 Adrian Beltre 8.1 12.1 4.0 Jim Edmonds 8.7 12.7 4.0 Albert Pujols 9.3 13.2 3.9 Kevin Millar 5.4 9.2 3.8 Aubrey Huff 5.3 9.0 3.7 Ryan Klesko 5.0 8.6 3.6This is quite a diverse group of players, headed by a wide, wide margin by J.T. Snow. In fact, Snow's turnaround in the second half is the third biggest since 1972, but every other player who increased his runs by more than eight did so in a strike year. Barry Bonds (11.6) and Jeff Bagwell (10.8) where ahead of Snow's pace in 1994; Mike Stanley (8.6)--also in 1994--and Jason Thompson (8.0) in 1981 were just behind him. The common thread among all these players is just clearing the 100-PA threshold set at the bottom of this study. Save for Snow, none of these players had more than 138 plate appearances in the second half of their season. This year, Snow had 193 PA in the first half and 216 in the second, lending a bit more validity to his record-breaking turnaround.
The rest of the group is an amalgam of labels. There's a rookie (Adam LaRoche) and a career minor leaguer finally getting playing time (Ross Gload) who appear to have figured things out. There's the unclassifiable Ichiro Suzuki, überprodigy Adrian Beltre's emergence, the improbable improvement by two of the Cardinals' three heads of Cerebus, Carlos Delgado's terrible first half, and a smattering of role players enjoying hot streaks. There certainly doesn't seem to be much rhyme or reason in this group.
And the worst:
PLAYER PRE-ASB POST-ASB DIFF Sammy Sosa 7.9 5.0 -2.9 Mike Lowell 8.5 5.4 -3.1 Hank Blalock 8.2 4.9 -3.3 Jim Thome 10.1 6.7 -3.4 Jose Valentin 6.4 2.9 -3.5 Danny Bautista 6.8 3.2 -3.6 Sean Casey 10.2 6.4 -3.8 Charles Johnson 6.8 2.7 -4.1 Hee Seop Choi 7.5 3.2 -4.3 Laynce Nix 7.4 3.1 -4.3 Mike Piazza 7.5 3.2 -4.3 Ivan Rodriguez 9.9 5.6 -4.3 Lyle Overbay 9.3 4.8 -4.5 Troy Glaus 10.3 4.8 -5.5 Manny Ramirez 12.3 6.7 -5.6This group, however, appears largely made up of slower players, many of whom are playing first base or catcher. Catchers are often lamented as baseball's walking wounded, and while the numbers here bear that out, they don't do so as obviously as they do for first basemen (with the exception of the aforementioned Snow). Jim Thome, Sean Casey, Hee Seop Choi, Mike Piazza and Lyle Overbay highlight our list of players who struggled through the stretch run.
Manny Ramirez's second half (.266/.349/.536) only looks bad when compared to his first half (.344/.437/.682). With Vladimir Guerrero and Gary Sheffield both holding steady with 0.0 change, Ramirez's deceleration in the second half has likely dropped him from serious MVP consideration. His decline isn't anywhere near the biggest in recent memory. Chris Stynes had a hot start in 2000 (.449/.477/.719) in just 109 plate appearances and came back to earth in the second half (.288/.343/.417) for a net of -11.3.
If instead of looking at individual players we look at how positions tend to do over the course of the season, we can see if there's any truth to ideas like catchers wearing down more over the course of the season than others do. Here's how the various positions have done in 2004:
POS CHANGE C -0.11 1B 0.35 2B 0.06 3B 0.32 SS -0.17 LF -0.11 CF 0.00 RF -0.50 DH -0.16(For the purposes of the study, players were counted only at the position at which they accumulated the most PA in a season. Thus, for example, Ross Gload is listed as a first baseman instead of a right or left fielder because he registered more PA at first base.)
While catchers have shown a negative trend this season, it's been nothing compared to the collapse of right fielders, typically one of baseball's top offensive positions. Looking at the individual performances, Sammy Sosa and Danny Bautista stand out as examples of extreme negatives, but they should be counterbalanced by Ichiro and Juan Rivera. The answer lies in the middle group of players who made neither list. Jermaine Dye, Bobby Abreu, Jose Cruz Jr. and Richard Hidalgo, among others, all showed significant dropoffs in the second half of the season, pulling down the average for right fielders across the board.
First base, however, has improved as the season has gone on, despite the struggles of five of its biggest stars. Like right field, first base was buoyed by Snow, Kevin Millar, Delgado's turnaround and Albert Pujols' remarkable improvement on his awesome first half. With Shawn Green and Ryan Klesko also improving in the second half, first base actually came out ahead.
This improvement, however, is far from typical. If we add the difference between first- and second-half performance for each year from 1995-2004, we get the following totals:
POS CHANGE C -1.98 1B -1.31 2B -2.05 3B -1.13 SS -0.22 LF -0.72 CF 0.01 RF -1.60 DH -1.02Only center field--and just barely so--has shown improvement from the first to the second half at the plate. While catchers certainly do have a very hard time, second basemen, perhaps due to players sliding into them turning the double play, show the greatest decline over the course of the season. (It should be noted, however, that catchers actually improved in the strike-shortened 1995. Excluding that season gives them a net total lower than second base.)
Next on the list is right field, the same position that's having such a hard time this season. While the extreme decline this year makes up some of the difference over the last 10 years, it still doesn't explain the degradation performance right fielders see as the season goes on. Of any position on the field, right field seems like the one least likely to increase the chances of injury, so it's hard to place the blame for their struggles on defensive difficulties or injuries. Perhaps right fielders, many of whom seem like first basemen who simply have to run, aren't of the body type to deal with a full season. I have no idea how to explain it.
Next spring, though, when your favorite right fielder, catcher or any other player gets off to a hot start, remember that it's more than likely he'll taper off as the season winds through the summer. For this year, especially considering the general trend for hitters to wear down in the second half, J.T. Snow's turnaround remains one of the season's most improbable events. Of course, he's still no where near as good as that other guy on his team.