Matt Kemp and the Petco Park Problem

Well, the final 24-36 hours of the Winter Meetings sure were fun-filled. The Red Sox rebuilt their rotation, importing a slew of ground ball guys to help combat the Fenway effect. Their rotation, however, will not include Jon Lester, who was lured away by the Cubs and a nifty $30M signing bonus, among other amenities. Perhaps most notably, the baseball world got an inkling of what the Dodgers’ new braintrust is all about, as they largely re-made their club, installing a brand new middle infield, adding to their already imposing starting rotation, deploying multiple clubs’ prospects here, there and everywhere, and moving slugger Matt Kemp within the division, to the Padres. It’s all pending physicals, of course. How will moving from Dodger Stadium affect Kemp? Are the Padres getting an impact bat, or something less than that?

First, some background. Kemp was not your typical blue chip, top of the draft high school prospect. He was an absolute steal in the 6th round out of Midwest City HS in Oklahoma, a huge feather in the cap of then-Dodger Scouting Director Logan White, who is perhaps not coincidentally currently employed the Padres, a recent hire of new GM A.J. Preller. In 2002-03, his first two years making the selections, White secured some kind of haul for his club.

In the former draft, he tabbed James Loney in the 1st round, Jonathan Broxton in the 2nd, James McDonald in the 11th, Eric Stults in the 15th, and some guy named Russell Martin in the 19th. In the latter, he got Chad Billingsley in the 1st round, Xavier Paul in the 4th, Wesley Wright in the 7th and A.J. Ellis in the 18th, in addition to Kemp, the clear pick of that litter.

For a high school draftee, Kemp tore through the minor leagues, batting .310-.357-.516 reaching the majors at age 21, and fully entrenching himself as a Dodger regular by the age of 22. Each year, I compile an ordered list of minor league prospects based both on offensive production and age relative to the league. Kemp ranked in the top 71 prospects in the game in each of his four minor league seasons, the top 41 in three of them, the top 25 in two, and peaked with a #4 ranking in 2006. Players with such a dominant minor league track record have a very high incidence of significant MLB success.

And Kemp certainly was quite successful over his tenure with the Dodgers. He finished second in the NL MVP voting in 2011, when he missed a 40/40 season by a single homer, leading the league in homers, RBI and runs scored while batting .324. Injuries have tarnished him quite a bit since, as he’s stolen 26 bases in three seasons since swiping exactly 40 in 2011, and the early returns on him defensively as a corner outfielder — let alone a center fielder, where he is now a disaster, to be honest — aren’t very good. Matt Kemp’s value, at the relatively tender age of 30, is already squarely centered in his bat.

Today, let’s attempt to evaluate the strength of that bat, and how the move from Dodger Stadium to the notoriously pitcher-friendly Petco Park might affect him. Each year, I prepare my own park factors based on granular batted-ball data. I compare the actual results of each batted ball hit in each park to what each batted ball of that approximate speed and angle would have produced in a neutral environment. Give each actual and projected event a run value, and divide the actual average run value per 27 outs to the projected average run value per 27 outs, and voila, you have your park factors, adjusted for batted ball quality. Below you will find the 2014 overall and fly ball-specific park factors for Dodger Stadium and Petco Park, as well as both park’s overall single, double, triple and home run park factors. The Dodger Stadium figures might surprise you:

OVERALL	LF	LCF	CF	RCF	RF	ALL
Dodger Stad.	93.3	121.5	110.4	126.8	86.1	104.8
Petco	87.2	82.5	79.0	78.6	90.1	85.3
—————-	———-	———–	———-	———-	———-	———-
FLYBALL	LF	LCF	CF	RCF	RF	ALL
Dodger Stad.	108.4	177.1	163.8	163.6	71.8	125.1
Petco	73.5	69.7	45.7	60.9	96.2	75.1
—————-	———-	———–	———-	———-	———-	———-
HIT TYPE	1B	2B	3B	HR
Dodger Stad.	94	116	60	122
Petco	93	91	100	90

Obviously, Petco grades out as a fairly extreme pitchers’ park, Its 2014 overall 85.3 park factor was the third lowest in the majors, ahead only of noted hitter graveyards AT&T Park and Safeco Field. Its 75.1 fly ball park factor ranks fourth lowest, ahead only of those two plus Angels Stadium. Interestingly, Petco had the lowest singles park factor and fifth lowest doubles park factor last season. Its homer park factor of 90, while well below average, was higher than that of 10 other parks last season, largely since the fences were brought in fairly significantly a couple years back.

Then there’s Dodger Stadium, which most people have pegged as pitcher-friendly. Not so fast. Its 2014 overall 104.8 park factor was eighth highest in the major leagues. Its fly ball park factor of 125.1 was 7th highest, and it ranked as the third most double-friendly (116 park factor) and seventh most homer-friendly (122) park in the majors last season. Like Petco, it was stingy with singles, with its 94 park factor ranking third lowest in MLB.

Now let’s focus on the middle of the field, the LCF, CF and RCF sectors. Petco’s overall park factors in those three sectors were 82.5 (25th), 79.0 (28th) and 78.6 (30th and last). Dodger Stadium’s were 121.5 (4th), 110.4 (7th) and 126.8 (3rd). With regard to fly balls only, Petco’s were 69.7 (25th), 45.7 (28th), and 60.9 (29th) as compared to Dodger Stadium’s 177.1 (3rd), 163.8 (4th), and 163,6 (3rd) in those sectors.

Now let’s put the focus back on Kemp, looking at his 2014 plate appearance outcome frequency and production by BIP type data. As part of this analysis, we’ll pay special attention to how much of his damage, especially in the air, has been done to those middle three sectors of the field. First, the frequency data:

FREQ – 2014
Kemp	%	REL	PCT
K	24.2%	119	84
BB	8.7%	114	66
POP	3.6%	47	10
FLY	31.8%	113	74
LD	24.0%	115	76
GB	40.6%	94	41

Any discussion of Kemp has as a hitter has to begin by looking at his K and walk rates. Kemp hits the ball quite hard, as we shall shortly see. 10 players in the NL who, based on a combination of metrics, hit both their fly balls and line drives a full standard deviation or more harder than the league average last season. Those 10, in random order: Kemp, Evan Gattis, Jarrod Saltalamacchia, Giancarlo Stanton, Marcell Ozuna, Jay Bruce, Khris Davis, Andrew McCutchen, Mark Trumbo, Paul Goldschmidt and Michael Morse. Three names — Stanton, McCutchen and Goldschmidt — stand out as legitimate stars, and the difference between them and the others can be found in their respective walk rates; Stanton’s was 14.7%, McCutchen’s 13.0%, and Goldschmidt’s 13.4% in 2014.

The rest of those guys, including Kemp, swing hard in case they hit it, and had sub-10% walk rates last season. Of that group, Ozuna is the one to watch. He’s just a puppy, and has plenty of time to nudge his walk rate up and his K rate down, and become a star in the process. Khris Davis too has a puncher’s chance, Bruce and Morse have had their day and are beginning to trend downward, and the rest likely are who they are moving forward.

Kemp, like Bruce and Morse, has had his day. His 2014 walk rate was the best of the rest, and his walk rate percentile rank of 66 is above MLB average, but simply isn’t good enough given his high K rate (84 percentile rank) to push him into elite offensive performer range. Kemp has never posted a K rate percentile rank below 80 in his career, and only once has exceeded his 2014 walk rate percentile rank. He will need to rely on exceptional batted ball authority to overcome his unfavorable K/BB situation.

There is some pretty exceptional stuff in Kemp’s batted-ball frequency profile, that gives you a better idea how his 2011 mega-season — the one in which he drew a bunch of walks — came together. Kemp doesn’t pop up. His 2014 popup percentile rank of 10 is not an anomaly; his career high is 22. His liner percentile rank of 76 isn’t an outlier, either; his career low, amazingly, is 72. Kemp has a knack for squaring up the baseball.

At one time, Kemp actually hit too many fly balls. Lots of good things drove his 2011 campaign, but one very temporary negative aspect was among them. Hitting more fly balls than grounders is actually quite rare, and isn’t a harbinger of continued excellence. Only 10 or so guys do this in either league in a given year, and invariably, the vast majority of those players decline the next season. Some of them are old players who are solely focused on pull power in the air; think Raul Ibanez, 2013. Very occasionally, they are players like 2011 Matt Kemp and 2014 Mike Trout who either win MVPs or come close to doing so. Their homer totals are bolstered by the extra fly balls they hit, but holes are opened in their offensive games that are then exploited going forward.

Kemp’s fly ball percentile rank crested at 96 in his peak season of 2011, but has declined each season since. He’s now in a fairly balanced place with regard to his fly ball and grounder rates. He doesn’t hit many “high” fly balls or “low” grounders compared to his peers, either. MLB hitters batted just .094 AVG-.224 SLG on high fly balls — they are separated by physically drawing a line halfway between the popup and line drive boundaries — and .372 AVG-.960 SLG on low fly balls in 2014. For Kemp, 21.31% of all fly balls fit into the high group, 78.69% into the low. This is truly exceptional. Probably unsustainable, but exceptional nonetheless, and indicative of Kemp’s solid all-around ball-striking ability.

The frequencies only tell us so much; let’s delve deeper into Kemp’s production by BIP type, which is largely driven by batted-ball authority:

PROD – 2014
Kemp	AVG	OBP	SLG	REL PRD	ADJ PRD	SD PRD
FLY	0.426		1.148	256	190	172
LD	0.750		0.967	127	118	114
GB	0.186		0.212	60	138	118
ALL BIP	0.385		0.683	164	167	153
ALL PA	0.283	0.345	0.501	140	142	133

Kemp’s actual production on each BIP type is indicated in the AVG and SLG columns, and it’s converted to run values and compared to MLB average in the REL PRD column. That figure then is adjusted for context, such as home park, luck, etc., in the ADJ PRD column. For the purposes of this exercise, SH and SF are included as outs and HBP are excluded from the OBP calculation. Also for the purposes of this exercise, an extra column has been added to his full-season data, translating it as if half of his games were played in Petco Park, his new 2015 home.

First of all, Kemp did massive damage in the air last season, batting .426 AVG-1.148 SLG, for REL PRD of 256. Only four players in the NL did more actual fly ball damage last season; Goldschmidt, Stanton, Devin Mesoraco and Troy Tulowitzki. As we have already seen, the first two are among the three elite all-around ball-strikers in the NL, Tulo has Coors helping him out, and Mesoraco is one of those “more fly ball than ground ball” outliers discussed earlier who is likely due for a downtick in 2014.

Kemp was helped greatly by his home park on fly balls last season, and did most of his damage in those middle three field sectors where Dodger Stadium was most vulnerable. Fully 80.3% of Kemp’s fly balls were hit to those middle three sectors; he only turned and pulled to LF in the air 13 times all season. He hit 17 homers to LCF, CF and RCF combined in 2014; 13 of them were hit at home. After adjustment for context, Kemp’s fly ball ADJ PRD drops to 190, which is still quite good, but not quite elite. Kemp’s fly ball pull rate — (fly balls to (LF + LCF)/(RCF + RF)) — of 0.77 was a little under the MLB average for a righthanded hitter.

Kemp also did fairly extreme damage on line drives, posting a 127 REL PRD, just outside the top ten in the NL. Most of that is on merit, as the contextual adjustment drops him down only slightly to 118. On grounders, however, Kemp was a poor performer, batting only .186 AVG-.212 SLG, for a 60 REL PRD. His hard/soft grounder rates would seem to suggest that he should do quite better, hence the 138 ADJ PRD figure. This is where we need to dig a little deeper, however.

Kemp is one of the most extreme ground ball pullers in the game. Though his fly ball pull rate was low, suggesting that there is leeway for him to selectively pull more frequently in the air as he ages, enhancing his chances for power in his decline phase, his grounder pull rate was an excessive 10.00 in 2014, well over the level where overshifting becomes automatic, even for a righty hitter. This is quite a unique swing, with such a disconnect between his fly and grounder pull tendencies. This means that his 60 REL PRD on grounders is more likely to be the norm going forward than his 138 ADJ PRD.

Let’s put all of this together in the context of Petco. Move his fly ball and liner SD PRD figures down to 172 and 114, respectively, and make Kemp own that 60 REL PRD on grounders. You are now looking at a .251-.314-.464 true talent player, adjusted for his new home and his excessive grounder pulling tendency. That is a 118 SD PRD true talent, instead of the 133 figure above, which doesn’t take the estimated overshifting impact into account. And that’s with almost no popups and a ton of liners. He’ll age from that baseline. That’s simply not enough production from a player with limited complementary skills, at best.

Kemp hits the ball hard, and he hits it true, rarely too “high” by popping up, or too “low” by chopping it. Still, his significant K/BB imbalance and especially his extreme grounder pulling tendency presents him with a difficult calculus. He is going to be around awhile, a guy who should hit 20+ homers deep into his thirties, but he will not be a star caliber bat again unless he at least partially addresses one or both of those weaknesses.