November 9, 2007
Using Batted-Ball Data and PECOTA
As this is my first entry into fantasy baseball coverage here at Baseball Prospectus, it makes sense to explain what it is I would like to accomplish. There are countless fantasy baseball columns out there, in magazines or on websites, but none of them focus on the kinds of statistics that we have at our disposal here at BP. We do our best to help you out before draft time through the release of the annual, fantasy versions of the PECOTA forecasts, articles from our friends at Rotowire.com, and the excellent Team Tracker and Player Forecast Manager applications available to subscribers, but there is always room for more assistance in the competitive world of fantasy baseball.
The focus of this column will be to implement the various statistics and applications that we have throughout the season, instead of just at the beginning of the year. PECOTA can be referenced throughout the season in order to assist in your decision making when it comes to acquiring or dumping players. BP's advanced pitching and hitting statistics are invaluable when it comes to evaluating performance in context, and this column will serve as an outlet for that information's weekly dissemination. In addition, as those of you who read the weekly Player Profiles know, batted-ball data and BABIP effects will be implemented into analysis when necessary in order to paint a more in-depth picture of a player's performance. Essentially, we want to give you the Baseball Prospectus perspective on fantasy baseball, year round.
For this first piece, we will take a look back at some of PECOTA's best 2007 forecasts with an eye towards helping you assemble your keeper lists over the winter. With many leagues using the offseason to make their own deals in order to prepare for the upcoming year, having a way to check out forecasts before they are available is a plus.
We can go about this by finding players who matched or came very close to their weighted mean forecasts; players who hit their weighted means are likely to have their new PECOTA projections resemble their 2008 lines from the 2007 five-year forecasts. By the same token, players who missed their weighted mean due to bouts of poor luck on balls in play during the season can have their lines corrected and match up the same way.
For those less familiar with batted-ball data, you can approximate where a player's AVG/OBP/SLG should have been using their line-drive rate and BABIP. This is done by adding .120 to a player's line-drive rate, generating a figure called eBABIP-expected BABIP. The difference between a player's eBABIP and BABIP can be added or subtracted to a player's line to get a better idea of where they player should have been without either poor or bountiful luck.
First off, we have Alfonso Soriano, who should qualify as both a left fielder and a center fielder in many fantasy leagues, for 2008 at least. Soriano was projected to hit .287/.349/.569 in his first year with the Cubs, and he managed to .299/.337/.560 with 33 home runs and just 19 steals after 46 bombs and 41 bags the previous season. Despite the drop in steals, Soriano still has almost absurd offensive value-as long as he qualifies as a center fielder. His 2008 PECOTA projection from last year's five-year forecast is .285/.349/.564, which should be more than enough reason to keep him or grab him early on in your draft. Taking a look at his batted-ball data doesn't mess with this forecast that much, as his .320 eBABIP was not all that far off from his actual .337 BABIP. If you kept Soriano around for his steals, then you may not be as happy as you could be, but otherwise he is still a fantastic fantasy option that you want to hold on to.
Derek Jeter's PECOTA forecast of .322/.390/.452 was eerily similar to his actual line of .322/.388/.452. This helped make Jeter one of the most valuable offensive shortstops around, finishing third in the positional VORP rankings at 53.3, also good for first in the AL. If you look at his '08 forecast though, you may want to consider dealing him to someone who hasn't had the chance to see it. It isn't that a .308/.375/.434 line lacks value-for a shortstop these days, that's still high quality-but if you're looking to gain more back through a trade, it's better to deal him a year early than a year too late. He still looks steady for the next few years, though there is an issue with his batted-ball data. Jeter's BABIP was a lofty .369, down from an even more ridiculous .394 figure in 2006. His eBABIP was just .319, meaning you could lop off a significant chunk of batting average from his season line. That's something to watch out for, as a drop in his slash stats of nearly 45 points would put a serious dent in his value.
After winning the AL MVP Award in 2006, Justin Morneau saw his numbers take a dive off a cliff in 2007. His BABIP dropped 59 points from .335 to .271, and his line-drive rate dropped with it as he finished at .271/.343/.492. This is a bit off from Morneau's weighted mean forecast of .293/.362/.526, but those lost hits can be recovered with some adjustments. Morneau typically hits somewhere around 18-20 percent of his batted balls for line drives each year, but he managed only about 16 percent in 2007. His BABIP reflected the change-liners correlate BABIP better than any other batted-ball type-so his numbers fell. If Morneau is able to tweak his swing enough to recover those lost liners, we should see his averages climb back up. The .290/.359/.524 line expected out of him in 2008 from his 2007 PECOTA sounds close to what an adjustment would give him, so work off of that in your evaluation of the former MVP's value in the season to come. This does not give him keeper caliber numbers unless you're in a very deep league, so attempt to deal him accordingly.
Carlos Beltran's 2007 has been viewed as a disappointing follow-up to his MVP-caliber 2006 season, but PECOTA was expecting the dip in production-his forecast of .282/.377/.530 was close to his actual .276/.353/.525. Given that his liners and BABIP aren't too far off from his eBABIP, it's safe to expect Beltran to perform near his 2008 forecast of .278/.377/.527, excellent production out of center field. Despite losing 71 points of ISO from 2006, Beltran finished fifth in VORP for center fielders, and second in the NL. His combination of power and speed, though not rare for center, is nevertheless still among the best in today's game, and for that reason he should either be kept or one of the top picks on draft day.
Fantasy players have made Carl Crawford one of their top picks almost every year since his debut, due to his speed and batting averages. The 2007 season was no different, with Crawford matching PECOTA's .309/.351/.476 projection with an actual .315/.355/.466 showing, along with the fourth season with 50-plus steals of his career. Crawford should be entering his offensive peak right now at age 26, and his PECOTA projection from the five-year forecast is almost identical to his 2007 figures: .310/.351/.476. Combined with his massive steal totals, if you have Crawford, you have one of the top fantasy players around. There is one issue though, and that's Crawford's BABIP of .375. His liner rate of 23.5 percent is by far the highest of his career, and probably isn't sustainable, given that his 2004-2006 rates ranged from 17.5 to 19.6 percent. With a dip in liner rate next year, Crawford could see his batting average fall by roughly 20 points or so, but with everything else he brings to the table, he's more than worth that risk.