Wednesday, July 08, 2009
The control hitters have over LD%
Posted by Derek Carty at 7:00amI thought we'd do something a little bit different today, pulling an interesting question out of the 'ole mailbag to answer:
I understand that pitchers are not in control of their LD%s. Are hitters, and to what extent? Should I use LD% when evaluating hitters? To take an example from this year that's bothering me, Brendan Ryan has a LD% of over 22%. Albert Pujols is under 16%. I find it difficult to believe, to say the least, that this is the product of skill-driven results. Furthermore, Pujols has a below-average BABIP (in the .275 range last I checked), while Ryan's is above-average (in the .350 range last I checked). Arguably the differential in BABIP could be explained by LD%. Or, should I be looking at it as a case where the LD%s will correct, and the BABIP with it?
I've expressed, in passing, my dislike for the assumption that line drive rate is a repeatable skill, but I haven't really delved too deeply into it for quite some time. Thanks very much to reader Todd for bringing this question up and questioning something that many analysts still believe to be true.
The short answer to your question is that, no, hitters don't have very much control over their line drive rates.
Why are line drives good?
Let's take a quick step back. In case you haven't seen these numbers yet, line drives are very good. They become hits roughly 70 percent of the time, while groundballs fall in for hits just 25 percent of the time and outfield flies 15 percent of the time. Line drives are hit in the air, on a lower plane than outfield flies so that they land sooner, and they are often struck harder than outfield flies.
Line drives have a higher correlation with a hitter's Batting Average on Balls in Play (BABIP), a huge component of batting average, so line drive rate has gotten a lot of play over the last couple years. Unfortunately, many see the high correlation and assume that a high line drive rate now will equal a high BABIP in the future. Many will even refer to and arrive at conclusions based on the "LD%+.120" formula for "expected BABIP", which I have expressed my disdain for (as a forward-looking metric) in the past. I've even shown that this formula is worse at predicting BABIP than the hitter's actual BABIP from the previous season. But why? If line drive rate correlates so well with BABIP, why doesn't it do a good job of predicting it?
Line drive rate is not very stable
That's why. If I look at all hitters from 2004 through 2008 who amassed at least 300 at-bats in adjacent seasons (and played on the same team both years*), the year-to-year correlation is 0.28. That's a little worse than what we might call a medium correlation and essentially means that 8 percent of a hitter's line drive rate can be explained by his rate in the previous season.** This is certainly significant, but it's not too terribly high. To compare, roughly 60 percent of a player's ground ball rate can be explained by his previous season rate.
*It's been shown that balls in play are classified as line drives differently from park to park, so I limited myself to players who remained on the same team as a quick way of eliminating some of this bias.
**This isn't exactly true since I'm opening myself up to some bias by including some arbitrary cut-offs, but it serves our purposes well enough.
Check out this table, which shows the best and worst line drive hitters in 2007 and how their rates fared in 2008.
+--------------+----------+-------+-------+ | LAST | FIRST | 07LD% | 08LD% | +--------------+----------+-------+-------+ | Young | Michael | 27.20 | 22.51 | | Figgins | Chone | 26.45 | 23.76 | | Atkins | Garrett | 24.47 | 22.07 | | Howard | Ryan J | 24.33 | 22.30 | | Polanco | Placido | 23.92 | 18.73 | | Cust | Jack | 23.21 | 20.83 | | Wright | David A | 23.19 | 25.63 | | Hall | Bill | 23.10 | 20.92 | | Sanchez | Freddy | 22.47 | 24.31 | | Aurilia | Rich | 22.18 | 17.95 | +--------------+----------+-------+-------+ | Kendrick | Howie | 15.94 | 20.00 | | Guerrero | Vladimir | 15.64 | 17.09 | | Uggla | Dan C | 15.64 | 15.75 | | Snyder | Chris R | 15.33 | 18.22 | | Uribe | Juan | 15.10 | 20.46 | | Young | Chris B | 15.09 | 19.13 | | Punto | Nick | 14.56 | 20.51 | | Buck | John R | 13.41 | 16.19 | | Matthews Jr. | Gary | 12.89 | 14.46 | | Laird | Gerald | 12.15 | 21.53 | +--------------+----------+-------+-------+
Keep in mind that a league average line drive rate is roughly 19 percent. As you can see, almost all of the leaders and trailers regressed toward that 19 percent mark the following season, often quite heavily. In fact, just two of the 20 failed to move closer to league average in 2008.
If you need further validation, you can check out this list of projected line drive rates (leaders and trailers), courtesy of David Gassko and Chris Costancio's THT Projection System:
+------------+----------+-------+ | FIRST | LAST | pLD% | +------------+----------+-------+ | Garrett | Atkins | 21.6% | | Todd | Helton | 21.3% | | Freddy | Sanchez | 21.3% | | David | Wright | 21.3% | | Michael | Young | 21.0% | | Bobby | Abreu | 20.9% | | Manny | Ramirez | 20.9% | | Ryan | Ludwick | 20.8% | | Miguel | Cabrera | 20.7% | | Mark | Loretta | 20.7% | +------------+----------+-------+ | John | Mayberry | 16.9% | | Emmanuel | Burriss | 16.9% | | Chad | Tracy | 16.9% | | Lou | Marson | 16.8% | | Robinzon | Diaz | 16.8% | | Nick | Evans | 16.8% | | Cameron | Maybin | 16.8% | | Luis | Castillo | 16.7% | | Laynce | Nix | 16.6% | | Alexi | Casilla | 16.1% | +------------+----------+-------+
As you can see, the top projected line drive hitter for 2009 was Garrett Atkins at 21.6 percent, only a couple percentage points above league average. If we look in the other direction, Alexi Casilla is the only player projected to be much under 17 percent. This is a pretty tight range and provides further evidence that we shouldn't put much weight into wide swings in a hitter's line drive rate — sophisticated projection systems obviously include a great deal of regression to the mean for hitters.
The moral of the story
The moral of the story is that, like BABIP, line drive rate is prone to swings in luck. While some players are better line drive hitters than others and can post above average rates more often than not (see: Michael Young), hitters in general don't have a ton of control over this stat — or, at least, it takes several seasons to get a really good read on their ability. If a player like Pujols has never posted a line drive rate this low before, we should expect it to rise going forward (and like Todd suggested, his BABIP will often go with it). The reverse goes for a guy like Brendan Ryan, who has already seen his line drive rate drop to 21.5 percent since this e-mail was received a week or so ago.
Concluding thoughts
Any questions, as always, feel free to comment or e-mail me.
Derek Carty, 23, has also been published by NBC's Rotoworld, Sports Illustrated, FOX Sports, and USA Today. This season, he'll be contributing to FanDuel and will be linking to all of his work at DerekCarty.com. In his three years competing in expert leagues, he has won 2 titles with 4 top three finishes, including a LABR NL title in 2009, making him the youngest person to ever win a major expert league title. Derek is a proud graduate of the MLB Scouting Bureau's Scout Development Program and is a firm believer in the importance of combining stats and scouting. He welcomes questions via e-mail, Facebook, or Twitter.
Thanks for this. I’ve been wondering how stable LD% is for some time, and I see some analysts treat it like it’s pure skill.
Which brings me to Carlos Quentin. He was a pre-season favorite to be overvalued (right on so far), but I looked at him and saw some batting average upside with his relatively low K% and a 15% LD rate. In fact, despite his dismal start this year, THT’s PrAVE has him at .309
So while his power last year was probably a bit over his head, isn’t he closer to a .300 hitter than we may think?