The AL Cy Young Award Battle

Award season is upon us, and as far as the biggest pieces of individual hardware go, there will be relatively little debate this season. Clayton Kershaw would appear to be poised to sweep the NL Cy Young and MVP awards, and Mike Trout would seemingly be in line to finally (?) land his first AL MVP. That leaves us with the AL Cy Young, a heavyweight matchup between perennial contender and one-time winner Felix Hernandez and this year’s breakthrough pitcher, Corey Kluber. Conventional wisdom seems to hold that Hernandez would be the choice of traditionalists, while Kluber would gain the sabermetric vote. Should that be the case? Let’s take a deeper look.

Felix “The King” Hernandez has been baseball royalty since arriving on the scene as a teenager in 2005. He first made a run at the Cy Young in 2009 at age 25, going 19-5, 2.49, and finishing second behind Zack Greinke. He won it in 2010, in what was seen as a seminal moment in the ongoing acceptance of sabermetrics by the traditional baseball media. Felix went 13-12 with an AL-leading 2.27 ERA that season, and voters looked past his meager win total and rightly rewarded him. While Felix has been strong since then, garnering two more Top 10 Cy Young finishes along the way, he didn’t truly recapture his 2010 form until this season. His 15-6 record continues to belie his ability, and while his 2.14 ERA led the AL, it took a last-minute scoring change, docking four earned runs allowed from his total, to do so.

While Felix had been recognized as a future great from a very young age, Corey Kluber has taken a more scenic route to excellence. He was selected on the 4th round of the 2007 draft by the Padres out of Stetson University. He was moved to the Indians in a 2010 trade that sent Jake Westbrook to St. Louis and Ryan Ludwick to San Diego. He was never seen as a top-shelf prospect – while he did strike out just over a batter per inning in his minor league career, he also allowed over a hit per inning and posted a less than spectacular 4.42 career minor league ERA. After a couple of nondescript stints at the major league level in 2011-12, Kluber cracked the Indians’ rotation to stay in 2013, posting a strong 136/33 K/BB, and then took it to the next level this season.

To cut through the noise in their numbers and get to their true performance level, Let’s take a look at their outcome frequency and relative production by BIP type data compared to MLB average for 2014, both before and after adjustment for context. First, the frequency info:

FREQ – 2014
Hernandez % REL PCT
K 27.2% 133 85
BB 5.0% 66 9
POP 4.6% 59 3
FLY 22.4% 80 7
LD 17.9% 86 4
GB 55.2% 127 99
———— ———— ———– ———–
Kluber % REL PCT
K 28.3% 139 88
BB 5.4% 71 16
POP 6.4% 83 22
FLY 25.5% 91 24
LD 21.7% 104 73
GB 46.4% 106 80

First and foremost, both Hernandez and Kluber excel at the two behaviors that key pitcher performance, whether you measure it by ERA or FIP – they maximize strikeouts and minimize walks. Hernandez’ K rate is in the 85th percentile, while his BB rate is in the 9th percentile. Kluber is right there with him, with K and BB percentile ranks of 88 and 16, respectively. Such maximization of free outs and minimization of free baserunners gives pitchers significant margin for error in their management of batted ball contact.

Felix’ BIP frequency portfolio is a thing of beauty. His GB rate is in the 99th percentile, with all of his other BIP percentile ranks in the single digits. He has always been a ground ball guy, but his previous high grounder percentile rank was 92, and that was in his Cy Young season of 2010. His previous low fly ball percentile rank was 13, also in 2010 – it was 7 in 2014. His previous low liner percentile rank was 12 in 2009 – it was 4 in 2014. While a low liner rate alone is nothing to get too excited about – they fluctuate quite a bit from season to season for most pitchers – the full package shown in Felix frequency profile most certainly is.

Kluber’s BIP profile isn’t too shabby itself, though it does suffer when posted alongside the King’s. He too is a ground ball generator (80 percentile rank) though not to as extreme an extent. His liner rate is quite high (73 percentile rank), though it’s down quite a bit from his 2013 97 liner percentile rank.

We can learn only so much from looking at the frequencies – we need to get a better feel for the batted ball authority allowed within the major BIP types. To do so, let’s take a look at the production by BIP type allowed by both pitchers, both before and after adjustment for context:

PROD – 2014
Hernandez AVG OBP SLG REL PRD ADJ PRD ACT ERA CALC ERA TRU ERA
FLY 0.227 0.586 65 103
LD 0.627 0.912 98 98
GB 0.222 0.248 86 69
ALL BIP 0.279 0.423 73 75
ALL PA 0.199 0.239 0.301 59 61 2.14 2.21 2.29
———— ———— ———– ———– ———– ———– ———– ———– ———–
Kluber AVG OBP SLG REL PRD ADJ PRD ACT ERA CALC ERA TRU ERA
FLY 0.276 0.635 84 67
LD 0.722 0.970 121 99
GB 0.229 0.264 94 86
ALL BIP 0.331 0.490 100 84
ALL PA 0.231 0.273 0.342 77 66 2.44 2.87 2.45

The actual production allowed on each BIP type is indicated in the AVG and SLG columns, and is converted to run values and compared to MLB average in the REL PRD column. That figure is then adjusted for context, such as home park, team defense, luck, etc., in the ADJ PRD column. In the three right-most columns, their actual ERAs, calculated component ERAs based on actual production allowed, and “tru” ERAs, which are adjusted for context, are all presented. For the purposes of this exercise, SH and SF are included as outs and HBP are excluded from the OBP calculation.

A couple of the significant contextual factors that are sure to creep into the ongoing comparisons between Hernandez and Kluber are the extreme pitcher-friendliness of Felix’ home park, Safeco Field, and the caliber of the defense playing behind Kluber, which has received low grades from advanced metrics. By poring through the production by BIP data above, we can get a better feel for how these factors should be weighted.

Hernandez allowed well below MLB average production on fly balls (.227 AVG-.586 SLG, 65 REL PRD), though that is heavily influenced by his home park. In actuality, he allowed slightly harder than average fly ball contact (103 ADJ PRD). He allows very few fly balls, as we saw above, so in totality, Felix is affected by Safeco much less than other Mariner starters. On the other hand, Felix allows tons of grounders, and yielded .222 AVG-.248 SLG production on them, for 86 REL PRD. After adjustment for context, however, he should have allowed materially lesser production – a 69 ADJ PRD figure that is seemingly impossible on its surface.

Scott Feldman’s 79 mark was the lowest among 2013 ERA qualifiers. Fully 28.2% of the grounders allowed by Hernandez in 2014 – that would be 28.2% of a very large number – were hit at 55 MPH or lower. This, in a league where pitchers don’t bat. Those grounders are basically outs. 59.2% of all grounders allowed in the majors this season were hit at between +5 and -10 degrees off of the bat – only 49.5% of Felix’ were. Felix allows a ton of grounders, and they’re hit much more weakly, and at a much lower exit angle compared to those allowed by his peers. These facts are the primary driver behind his 73 REL PRD on all BIP, and the relatively small – for Safeco – upward contextual adjustment to 75 ADJ PRD.

All in all, the run-preventing effect of Safeco on the fly balls allowed by Felix is basically offset by the bad luck he experienced on grounders. One does not need to make a significant inflationary adjustment to his traditional ERA of 2.14 to come up with his “tru’ ERA of 2.29. You basically need to add back the earned runs taken away from him by the end-game official scoring adjustment.

As for Mr. Kluber? The “contributions” of his defense can easily be seen in his production by BIP type data. He allowed .276 AVG-.635 SLG on fly balls (84 REL PRD), but based on the relative authority of those fly balls, he should have allowed much less damage (67 REL PRD). He also allowed much more damage than he “should” have on liners (121 REL PRD, 99 ADJ PRD) and grounders (94, 86). Overall, he allowed exactly MLB average production on balls in play (100 REL PRD), but adjusted for context, should have allowed 84 ADJ PRD.

Kluber’s calculated component ERA of 2.87 is puffed up quite a bit by extra hits yielded by the relatively poor defense behind him – fairly optimal sequencing allowed him to record a lower actual ERA of 2.44. The impact of those two factors – the defense and the sequencing – offset almost exactly, as his “tru” ERA of 2.45 is virtually the same as his actual mark.

There’s quite a bit of irony here. In this day and age of more sophisticated thought brought on by the introduction and advancement of sabermetrics, actual, traditional ERA – by accident – seems to have gotten this one just right. Yes, Safeco deflates offense, but it affects Felix much less than most because of his prolific grounder-inducing ability. Kluber’s subpar defense did hurt him, but his ability to work out of jams prevented it from negatively impacting his ERA. These are two clearly Cy-worthy seasons – but Felix Hernandez was a little bit better. He would get my vote.

We hoped you liked reading The AL Cy Young Award Battle by Tony Blengino!

Please support FanGraphs by becoming a member. We publish thousands of articles a year, host multiple podcasts, and have an ever growing database of baseball stats.

FanGraphs does not have a paywall. With your membership, we can continue to offer the content you've come to rely on and add to our unique baseball coverage.

Support FanGraphs




newest oldest most voted
Luke
Guest
Luke

Interesting. I was pro-Kluber before reading this.

One factor left out here, though, is pitch-framing. We know Zunino is substantially better at that than Gomes. Do you think this effect would be large enough to tilt the scales in favor of Kluber?

deflated
Guest
deflated

How many percentage points do you think better framing would move from Kluber’s BBs to Ks? It’s not going to affect the batted ball data and Felix wins there and I just don’t see where a few extra strikes are going to make that big a difference.

vivalajeter
Guest
vivalajeter

Why wouldn’t it impact batted ball data? I would think that it would help put the pitcher in more pitchers counts, rather than falling behind in the count, and hitters perform worse in pitchers counts.

deflated
Guest
deflated

I should have said ‘no statistically significant change’ rather than ‘won’t affect’, sloppy phrasing on my part. Changing from Gomes to Zunino we’re talking about 0.9 extra strikes a game; over the course of a season we’re talking about less than 20 at bats were Kluber may have been in a better count with a different catcher and many of those at bats ended in a favourable outcome for Kluber anyway.

I think Gomes’ better arm and lower SB rate would have more of a factor than framing.

haishan
Guest
haishan

StatCorner estimates Zunino as providing 22 RAA by framing, versus 8 RAA for Gomes. Let’s assume that their framing value was independent of the pitcher, and also that those were the only catchers for Hernandez and Kluber, because that makes calculations easier.

Zunino caught 1121 innings and Gomes 1082; Hernandez and Kluber each pitched 236 innings, basically. So Zunino’s framing saved about 4.6 runs for Hernandez and Gomes saved 1.7 for Kluber.

King Felix’ “tru ERA” comes out to almost exactly 60 ER over the course of the season. Kluber’s comes to 64. Add in the 2.9 extra runs of catcher framing that Zunino saved and it’s a virtual tie. If you want to discount the framing numbers some, that’s fine, but you can’t get Felix’ ERA over Kluber’s, nor can you turn this into a blowout for the King. Which is probably as it should be.

Luke
Guest
Luke

Thanks haishan. Good answer.

Umpires Love Felix + Zunino
Guest
Umpires Love Felix + Zunino

Baseball Prospectus gives data by batteries.

Kluber’s catchers saved him 0.4 runs total.
Hernandez’s catchers saved him 8.3 runs total.

Quite the difference, huh?

(Felix + Zunino was the 2nd best framing duo in the entire major leagues!)

haishan
Guest
haishan

aha, I didn’t realize battery data was available. Google failed me on this one.