Friday, August 21, 2009

Explaing Albert's "Albatross"

On Circling the Bases this morning, contributor Matt Casey pondered Albert Pujols’ second-half performance, which is a significant drop-off from his amazing start. Pujols’ season line has dropped from . 332/.456/.723 to .321/.445/.675 thanks to a .280/.401/.525 “drought”. Casey writes,
“Are we ready to blame the Home Run Derby for squashing Albert Pujols' chance at the Triple Crown, or is that just an easy excuse? We remember players from the past flaming out after taking part in the Derby - David Wright hit only 6 HR after the break in 2006, and of course Bobby Abreu hit a mere 6 dingers in the second half after dominating the slugfest the year before….Did the HR Derby mess up his swing a bit?”

Casey’s article equivocates a great deal, and never really delves deeply into Albert’s performance beyond looking at the surface numbers and wondering aloud whether his two rounds of homerun derby (in which he swung at something like 35-40 pitches) has messed with the man. Of course we can’t be certain, even if Pujols were to complain about the Derby’s effect on him (which he hasn’t), since the data set is small and the subject of the study is probably less reliable a source than The Common Man would like to have. That said, The Common Man hopes that some intrepid soul is going through tape of Albert’s ABs before and after the All Star Break to see if there’s any difference in his swing (ah, the glory of our times, when such things are possible).

The Common Man prefers to look at the issue as a whole. Understanding that even great players have rough patches, it’s important to look at the alleged phenomenon Casey and others have used to explain the performances of Abreu and Wright, and that other players have used to excuse themselves from the Derby. Is this effect real? Does the Home Run Derby unduly affect its participants, as many have been led to believe.

First, a caveat: There is no way for The Common Man to determine with any reliability whether the Home Run Derby has affected the performance of an individual player. It certainly could have. But significant analysis of swing mechanics is probably necessary to address that.

Now then, since 2000, there have been 54 individual participants in the Home Run Derby, but 80 player-seasons are in our data set (8 participants per derby, 10 derbies). Of those 80 seasons, 47 experienced a drop in batting average from the first half to the second, losing an average of .0093 points off of their BA. Forty of the player-seasons experienced a drop in on-base percentage, and the cohort lost an average of .0045 points from their OBP. Forty-five seasons experienced a drop in SLG in the second half, losing an average of .037 in the process. Finally, between the first and second halves, players lost an average of 7.8 homers.

So at first glance, it appears that there is a drop, but there’s some noise in our data. First, the two halves, despite their names, are not the same duration. Teams tend to play somewhere around 90 games in the first “half” (Pujols’ Cardinals played 91 games before the All Star Break this year), meaning that there are fewer available at bats in the second half. This accounts for a great deal of the difference in total homeruns hit, such that the actual difference in terms of homeruns/game ends up being closer to 6.8 HR lost over the course of the season. Second, our data also includes players from 2009, whose HR are drastically down simply because they’ve only played 40 games or so since the break. If we remove these individuals from our analysis of the HR total, we find the difference in HR is something like 4.6 HR lost.

Finally, there’s the case of Ivan Rodriguez and Brandon Inge, both of whose rate stats and home run totals were and have been gutted by injury. In 2000, I-Rod was hitting .366/.393/.708 at the All Star Break, with 26 home runs, on his way to what could have been the greatest season by a catcher ever (perhaps even better than Mauer’s year this year). Eleven games into the second half, however, Pudge broke his right thumb on Mo Vaughn’s bat (is there nothing that behemoth can’t ruin?) as he tried to throw to second base, and missed the rest of the year. He finished the second half hitting .184/.225/.316 with 1 HR. Inge has suffered from knee problems unrelated to his Derby appearance since the break that have robbed him of a great deal of effectiveness. While he continues to play, Inge is hitting .170/.257/.250 in 31 games, and is probably not a good example of typical post-Derby play. I-Rod’s removal (Inge was previously removed from the analysis because he participated in 2009), the difference, adjusted for approximate games played, is approximately 3.8 homers.

Once we remove these significant noisemakers from the data, we see that, in 71 players-seasons, there have been 42 drops in BA, 35 in OBP, and 40 in SLG. This translates to an average drop of .0066/.0006/.029, a very small blip. Indeed, when one considers that the players likely to be part of the Derby had already performed at a level above their normal ability level, and the difference in HR hit, is likely attributable to players finding their truer level, or regressing to the mean, or nagging injuries that weren't there in the first half, rather than some kind of post-Derby swing hitch. So let’s stop wondering about Albert Pujols, who is probably fine aside from his seemingly constant throwing and foot problems. The man, while most uncommon (far more so than The Common Man, I assure you), is still just a man and prone to regression like everyone else. Unlike, for instance, Joe Mauer, who has upped his game since the All Star Break (during which you’ll remember he participated in the Derby), hitting .392/.449/.622.



Note: Other fun stuff in the data (I’m happy to send a spreadsheet around to anyone interested):

Barry Bonds’ line in the second half of 2002 was .404/.608/.825. An OBP of almost 61%? Holy God.

Somehow Hee Seop Choi slipped into a Home Run Derby in 2005 for the Dodgers. It was Choi’s last season in the majors, and he was hitting .236/.318/.458 at the time. He was not on the All Star Team, as a particularly bad Dodgers squad was represented by Jeff Kent and Cesar Izturis.

Another unlikely participant was Pudge Rodriguez in 2005. Pudge had 6 homers at the All Star Break, and would finish with just 14.

The biggest drop in BA was Jim Edmonds in 2003, who hit .303 in the first half and .214 in the second. Based on his playing time, and injury history, he very well may have been hurt. His OBP and SLG dropped as well, but were still perfectly acceptable in the second half (.357/.507). The largest jump in BA was Junior Griffey in 2000, who went from .238 to .317. Griffey’s OBP, however, ticked up just two points from .386 to .388, and his slugging from .550 to .564.

Ryan Howard’s OBP jumped from .341 in the first half of 2006 to .509 in the second half once pitchers stopped throwing him the ball over the plate, the biggest jump in OBP in this study. On the opposite end, Carl Everett dropped .074 in 2000.

The biggest drop in HR between halves (non-I-Rod edition) was Carlos Delgado in 2000, who hit 30 bombs in the first half, but just 11 after the break. Not so coincidentally, Delgado’s drop in SLG (.225 was also the highest in the data set.

The smallest difference between halves in the three rate categories in the data set belonged to Miguel Tejada (who hit .311 in both the first and second halves of 2004), Vlad Guerrero (who slugged .547 in the first half of 2007, and .548 in the second half), and Garret Anderson (.345 and .344 OBPs in 2003) and Grady Sizemore (.374 and .375 in 2008).

7 comments:

Bill said...

That's really interesting.

One thing I'd like to see, if you or someone had about 5,000,000x more time, is a breakdown of the derby participants' pre- and post- splits against all other full-time, full-season starters. I bet there's a natural tendency to slow down over 162 games. It doesn't show up in the league-wide numbers, but that could be because your seriously underperforming guys (your Chris Davises) drag down the first-half numbers and are out of the league by the time you get to the second half. Just a curiosity...

The Common Man said...

While I think you're right that there's a tendency to slow down over 162 games, Bill, I think that tendency exists in pitchers as well (especially give the attrition of pitchers good enough to make the opening day rosters to injury over the course of the year). As such, I imagine the topic is made infinitely more complex. Maybe Tango, et al can include it in The Book II.

Bill said...

I feel like it would be likely to affect pitchers less, with the exception of a Sabathia or Halladay who just gets ridden into the ground every year, because pitchers' fatigue is so heavily (if stupidly and homogeneously) monitored, while most position players' is more or less ignored. But you're probably right, I don't know.

lar said...

Good work, TCM. Glad to see something from you.

I totally believe the conclusion that you came to. The whole "home run derby screws up your swing" argument is pretty silly, and I'm surprised that one of the Circling the Bases guys actually pushed it as a possibility.

I can't remember where right now, but I remember reading a similar study to what you just did here right before the All Star Break, looking at pre- and post-break stats, etc. Whoever it was found pretty much the same thing you did, and may have even tried to account for the things Bill brings up. Still, it's nice to see a second look at the data, just to verify the same thing. Plus, this is just the type of thing that needs to be constantly brought back up when silly people make silly arguments.

I think the best argument for Pujols not being affected by the Home Run Derby, though, is that only mortals can get worn out or otherwise adversely affected by a little effort, and Pujols is no mortal. That may not be subjectively verifiable, though ;-)

The Common Man said...

Thanks Lar, I'm happy to be back. I may have missed the study you're talking about. But this is what happens when no one else is in the office on a Friday in the summer, 1200+ words or so. :) I like your new Bloquin site.

lar said...

Yeah, I know exactly what you mean. I really wish I had the retrosheet stuff set-up in my office for days like that. I have the Lahman db, which is nice, but doesn't give you anything for play-by-play.

And thanks. I really do like the look of the Bloguin site too. Still getting used to it's layout and controls, though...

Brad Templeman said...

Very interesting. Its something said every year, mostly by people who haven't analyzed all of the numbers. It's possible it did affect one player, but we don't know who it is or for sure that it happened. If there is anyone who would not allow something as silly as the home run derby to mess their swing up, it's Albert Pujols.