The Ever-Changing Matt Carpenter

Editor's Note: Please welcome Ben Markham as a writer at VEB. He is an active member of the community and written 13 Fanposts, which is a great way to involve yourself in the VEB Community. This deep dive on Matt Carpenter is a good example of Ben's passion and insight into statistical analysis and the Cardinals.

Over the last three years, Matt Carpenter has arguably been the single most important player for the Cardinals. Each of the last three years, the Cardinals have ended the year in first place in the NL Central, and the Pirates have come in second. In each of those seasons, Matt Carpenter's fWAR has been larger than the difference in record between the two teams. In 2013, Marp's 7 WAR performance likely made the difference between the Cardinals' 97 win team and the Pirates' 94 wins. Same for his 4 WAR season in 2014 when the Cards bested the Pirates by 2 games. This year we again topped the Pirates by just 2 games, and Carpenter had another great year, turning in a little more than 5 fWAR.

It would be too simplistic to say that without Carpenter, we wouldn't have won the division either of the last 3 years. With a different team, Mozeliak and company would have made different moves. Perhaps they would have had similarly strong teams but they just had to spend a little more to get there. But things definitely wouldn't have been as great, and we wouldn't have still been in the position we are now, with financial flexibility and an OK-if-lacking-upside minor league system. Not bad from a guy taken 399th overall in the 2009 draft as a college senior.

So Matt Carpenter has been really, really good. Yeah, real groundbreaking stuff. You know that already. What makes it interesting to me is how Matt Carpenter has done it. While contributing almost all of his value on offense, he has had a distinctively different offensive profile each of the last three years. Let's begin by looking at his core stats from both 2013 and 2014:

	BB%	BB+	rank	K%	K+	rank	ISO	ISO+	rank	BABIP	BABIP+	rank	wRC+
2013	10	123	43/140	13.7	71	108/140	.163	112	66/140	.359	121	10/140	146
2014	13.4	172	6/146	15.7	79	91/146	.103	78	121/146	.318	106	52/146	117

Here we see his BB, K, ISO (Isolated Slugging), and BABIP (batting average on balls in play) numbers. I calculated a "+" stat for each, which simply adjusted his numbers to league average so that average was 100. For instance, Matt's 123 BB+ in 2013 means he walked 23% more than average, whereas his 71 K+ means he struck out 29% less than average. Rank shows where Carpenter ranked among qualified hitters. The "+" stat and the ranking are provided to give the reader an idea of exactly how much Carpenter changed year to year.

In 2013 he was above average in each of the 4 core stats. He was great at taking walks and limiting strikeouts, solidly above average power and a great, if fluky results on balls in play. Marp had the tenth highest BABIP in the league in 2013, a pretty sure sign that he would regress in the following season.

While 2013 Marp was patient, 2014 Marp took it to a whole new level. While 3 percentage points and some change doesn't seem like much, Marp went from the back of the top third at taking walks, to the top 5% at doing so. His strikeout rate ticked up a bit, but nothing that concerning. His BABIP regressed as expected, though still above average. However, Carpenter's power fell off a cliff. He went from the top 40% to the bottom 20%. Ultimately, that is what caused him to drop from one of the best hitters in 2013 to a merely solidly above average hitter in 2014.

So why did these changes occur? Well, you can't talk about Matt Carpenter without bringing up his plate discipline numbers. Let's look at a few of those in the same manner we did with his core stats:

	O-swing	O-S+	Rank	Z-swing	Z-S+	Rank	Contact	Contact+	Rank	Zone	Zone+	Rank
2013	22.1	71	136/140	55.6	85	135/140	88.8	112	13/140	45.6	102	38/140
2014	19.3	62	146/146	49.4	75	146/146	89.9	113	10/146	45.7	102	48/146

(for an intro on plate discipline stats, click here)

Carpenter was already one of the best at laying off pitches outside the zone, and in 2014 he was THE best at it. But it came with a similarly sized uptick in pitches taken in the zone as well. As you can see his Z-Swing percentage was also the lowest in the league, so this wasn't a matter of Carpenter getting even better at strike zone judgement as much as it was him being extremely selective. The contact, as well as the percentage of pitches opposing pitchers threw in the zone, stayed steady.

So, from 2013 to 2014 Marp got a lot more selective, taking pitches both in and out of the zone at an extraordinary rate. This doesn't explain the power drop off though. Let's look at his Batted Ball profile from both years:

	LD%	LD+	rank	GB	GB+	rank	HR/FB	HR/FB+	rank	HH%	HH+	rank
2013	27.3	129	6/140	38.7	87	104/140	6.10	58	124/140	33.5	110	62/140
2014	23.8	114	28/146	41	92	90/146	4.70	49	127/146	33.3	114	53/146

In 2013, Matt's LD% (Line Drive percentage) was elite. He was in the top 5% at churning out line drives. To an extent, that explains the very high BABIP. He also kept the ball off the ground pretty well, with 13% less ground balls than average. That's important because ground balls are very unlikely to turn into extra base hits. He also hit the ball hard more often than his peers, as indicated by the above average Hard Hit percentage. These are all good reasons for why Matt was able to post a very high BABIP and an above average ISO. So good in fact that he was able to survive a very poor Home Run to Fly Ball ratio (HR/FB).

In that context, it's also easy to see why Carpenter's ISO and BABIP dropped so hard in 2014. The line drives trended down, the ground balls trended up, and the HR/FB rate dropped to even more pitiful numbers. Overall, Marp was a solidly above average hitter in 2014, but the lack of power was concerning. For a player to put the ball in the air as often as Carpenter does, you'd like to see more of those balls get over the fence.

It's in that backdrop that Matt and the Cardinals entered 2015. He had without a question elite strike zone management, but would the quality of balls put in play limit his offensive profile? Well, let's look back at his four core stats, this time including 2015:

	BB%	BB+	rank	K%	K+	rank	ISO	ISO+	rank	BABIP	BABIP+	rank	wRC+
2013	10	123	43/140	13.7	71	108/140	.163	112	66/140	.359	121	10/140	146
2014	13.4	172	6/146	15.7	79	91/146	.103	78	121/146	.318	106	52/146	117
2015	12.2	156	15/141	22.7	114	27/141	.233	152	19/141	.321	107	47/141	139

Matt Carpenter with power? That is one scary sight for opposing pitchers! In a single season, Carpenter went from the bottom 20% in power to the top 20%. The trade off was a big change in his K rate, as he went from much better at limiting strike outs than his peers to posting his first K% above average. That's OK though, strike outs are the least important part of a player's core stats. Marp only lost a little bit of his walk rate, enough to take him from the top 5% to just outside the top 10%. But that's still elite.

So what did he change? Let's look at the plate discipline stats again:

	O-swing	O-S+	Rank	Z-swing	Z-S+	Rank	Contact	Contact+	Rank	Zone	Zone+	Rank	SwStk	SwStk+	Rank
2013	22.1	71	136/140	55.6	85	135/140	88.8	112	13/140	45.6	102	38/140	4.1	44	133/140
2014	19.3	62	146/146	49.4	75	146/146	89.9	113	10/146	45.7	102	48/146	3.3	35	145/146
2015	22.8	73	135/141	60.1	90	129/141	80.2	102	77/141	44.2	98	75/141	7.7	78	93/141

This time I included Swinging Strike percentage as well, and you can see why. After being one of the best for two years at avoiding swing and misses, in 2015 the number jumped up. He was still much better at avoiding swing and misses than average, but he could no longer be considered elite at the skill. This ultimately is a good thing, it means he was being more aggressive, taking more chances at hitting the ball hard.

He was also no longer trying his best to minimize swinging. His O-swing trended back to 2013 levels. He was much more aggressive with pitches in the zone, posting his highest Z-swing% of the three year period (despite still being below average overall). The contact% slipped to merely average, but for 2015, Matt wasn't great because of quantity of contact, but because of the quality of it. Notice also that the power seemed to lead to a small, but not insignificant decrease in zone percentage. Pitchers were a little more weary of Marp's newly found power, and pitching him more carefully. That's why even though he was less patient than his 2013 self, he still walked more often. With increased power often comes increased walks, and that's indeed what happened for Marp in 2015.

But how did he generate that extra power? A look at the Batted Ball stats:

	LD%	LD+	rank	GB	GB+	rank	HR/FB	HR/FB+	rank	HH%	HH+	rank
2013	27.3	129	6/140	38.7	87	104/140	6.10	58	124/140	33.5	110	62/140
2014	23.8	114	28/146	41	92	90/146	4.70	49	127/146	33.3	114	53/146
2015	28.5	136	2/141	29.7	66	140/141	15.80	139	34/141	36.8	129	22/141

Career high LD%, career high HH%, career low GB%, career high HR/FB. Yep, that'll do it. Notice that among qualified hitters, Marp had the second lowest GB rate in the game. Only Lucas Duda was better at getting the ball in the air in 2015. While Marp was always better than average at getting the ball in the air, he took it to elite levels in 2015. At the same time, he made big strides in hitting ball hard, going from the top 40% in the prior two years to the top 20% in 2015. More hard hit balls certainly had an effect on his HR/FB rate, as Marp finally busted out of his discouraging HR/FB rates of the past. He didn't forget about Line Drives either, his LD% came back with a vengeance in 2015, with Brandon Belt the only hitter in baseball able to top Carpenter in the stat.

In 2014, Matt altered his plate approach to max out OBP. That's not a bad route, OBP is important and he was the lead-off hitter after all. But you know what's even better? Maxing out power. Especially since the power will lead to being pitched more carefully and thus more walks anyway. What's in store for Matt in 2016? Let's start with his projection from Steamer compared to the last three years:

	BB%	K%	ISO	BABIP	wRC+
2013	10	13.7	.163	.359	146
2014	13.4	15.7	.103	.318	117
2015	12.2	22.7	.233	.321	139
2016 Steamer	11.9	19.1	.156	.319	120

This is about what I would expect from a projection system. Regress the career high K% and career high ISO towards previous years. But boy is that ISO regressed hard. However, a projection system isn't going to instantly believe such an outlier year, even if it comes with other stats that indicate a change. Matt is unlikely to be the second-best in baseball at getting the ball in the air again, and he was pretty bad at getting those fly balls over the fence before last year. So it's reasonable to assume that a drop in FB% and HR/FB rate will cut into his ISO. But I would expect Marp to continue his same approach next year after such a successful season, so dropping it 77 points seems like a bit much. I would take the over on that one, even if I think Steamer is correct to assume some regression. For the same reasons, I think the K% is regressed a bit too much as well. If Marp continues with the same approach, the above average strikeout rate is probably here to stay. That's fine though, because his above average walk and power numbers more than make up for it.

Matt Carpenter has been a great hitter, and one of the few capable of pulling off a complete change in approach year to year. At different times throughout his career, he has had an elite ability to lay off pitches, elite contact ability, elite ability to hit Line Drives, elite ability to keep the ball off the ground, and an elite ability at hitting the ball hard.

Because Matt is continuously changing his approach, it's hard to say what he'll look like in 2016. Perhaps he continues to increase the percentage of swings he takes on pitches in the zone, or his K% drops a little bit in his second season of swinging for the fences. But one thing's for sure: He will once again be a great hitter.