At this point it’s impossible to reasonably argue that Clayton Kershaw didn’t have an amazing month of June. This is breaking news. (How did you not know already?) Kershaw threw a no-hitter, which was nearly the second perfect game in Los Angeles Dodgers history; struck out 15 in that game, making it the second rated game all-time according to Game Score; went 6-0 as he struck out 61 in 44 innings of work, which works out to a 12.48 K/9; and had an ERA of 0.82 with a FIP of 1.18. There are teams that might consider six wins a success (Colorado had eight in June while Texas had 9), so when Kershaw dominates as thoroughly as he did, he deserves every last bit of recognition heaped upon him.
I wonder, though, did Kershaw even have the best month of June for a starting pitcher? Did you see what Felix Hernandez did in Seattle this past month? Better yet, was Kershaw even the best starter in the NL? Because, you know, the Cubs Jake Arrieta and the Nationals Jordan Zimmermann pitched pretty well too.
Before breaking out the traditional metrics, which may or may not give us a clear idea just how great these four pitchers were, I’d like to first provide a graphic illustrating just how amazing they were compared to their contemporaries.
Notice those three circular outliers to the far right? From left to right, those three circles represent the accumulated fWAR in the month of June for Zimmermann (1.7), Arrieta (1.8), Hernandez and Kershaw (2.0 each). I like this image for a couple of reasons: 1) I’m learning R and can now make colorful images to put into these posts, and 2) a visual such as this can show you just how far beyond the median and upper quartile the four pitchers were during the month. Impressed? When you consider the middle 50% earned between 0.2 and 0.7 fWAR, these four guys were unbelievable in their respective six starts.
Let’s move onto some more traditional stats, like innings pitched, hits allowed, and see how each of the four fared.
June Traditional Stats
I believe in neither W-L records nor ERA, but the traditional measurements are fun to look at regardless even if they provide little context. For instance, Zimmermann went 3-2 in June, but in those two losses the Nationals were shutout in both games, recorded a grand total of five hits (two and three), and Zimmerman allowed three runs (one and two), and if not for coming back against the Brewers, another brilliant Zimmermann outing would have resulted in a loss, making him 3-3 in June. Sometimes numbers tell us little, but in some other cases, numbers can tell us quite a bit about a player’s performance.
Just to put some of these numbers into perspective, Yu Darvish currently is the all-time leader in both K/9 by a starter at 11.15 and K% at 30%, but if I restrict my criteria to pitchers who have more than 100 career starts, the all-time leader becomes Randy Johnson at 10.57 and 28.5%. Kershaw, Hernandez, and Arrieta morphed into a more monstrous Randy Johnson this past month while Zimmermann became Justin Verlander (career 8.39 and 22.5% respectively). Kershaw’s 37.7% is amazing, even in an era of high strikeouts, but it still is only slightly better than the season record of 37.5% held by Pedro Martinez in 1999.[i] Does that make it any less impressive? Not at all.
In three separate months, Martinez had K/9 of 14.11, 14.14, and 15.37 with the representative K% at 38.3, 40.4, and 46.7. Looking at the table above, I thought that the numbers certainly seemed impressive in regards to today’s game and today’s pitchers, but how did the four players above compare to a historically great season?
Well, there you are.
If you look at the numbers like this, it’s not such a clear cut case for Kershaw to be crowned June’s best pitcher. Of course, by fWAR Hernandez is tied with Kershaw, and for the month Arrieta had a better H/9 and batting average against while being only ever so slightly worse in terms of WHIP. In fact, look at that same chart above but in terms of how each pitcher performed compared to the median:
% Improvement Over Median
I can of course pick and choose, trying to fit an argument into whatever criteria I deem best at the moment, but the point isn’t to argue which statistic is the most valuable for a pitcher or what defines most dominant. My point is that a case can be made for each of the four as June’s best, and nitpicking over an inning here or a strikeout there doesn’t seem to take the argument any further than I’ve already gone. I should just move onto individual games and see who provided the most consistent performance.
Bill James devised a method of scoring pitching performances called Game Score that can be used here to track each pitchers performance per game, attempting to find who was the most consistent for each of their respective six starts. Here, I’ll use the summary function in R to provide the mean, median, and quartiles.
|Player||Min||First Qu.||Median||Mean||3rd Qu.||Max|
R Summary Game Scores
I fully came into this expecting Arrieta to have the least variance in the middle, and he met my expectations. That sort of thing happens when you have games where you allow three hits twice, two hits once, and one hit once. Strike out people, don’t allow hits, and your Game Score is going to look impressive. I’ve written about Kershaw’s absolutely brilliant outing against the Rockies, and that game happened to be the second highest rated game in regards to Game Score in MLB history. So, seeing Kershaw’s mean top the list doesn’t surprise. In terms of variance between the middle 50% Kershaw is second to Arrieta. Hernandez had the highest 3rd quartile score, owing to monster outings against Tampa Bay where he struck out 15 and this past Sunday against Cleveland where he allowed just one hit.
Then there’s Zimmermann. Next to Kershaw’s near perfect game, Zimmermann pitched the best game of June. In that game he struck out 12 Padres while allowing just two hits. Game Score tends to reward high strikeout pitchers, so Zimmermann’s scores may not reflect just how dominant he was this past month. In relation to the other three men listed, his numbers might seem somewhat pedestrian, but that is far from the case. I watched enough Nationals games in June to know that hit corners like a Pong champion and didn’t win NL Player of the Week for June 2-8 by accident.
What have I learned so far? Clayton Kershaw and Felix Hernandez were the best out of the four in terms of fWAR accumulated in the month; Kershaw dominated the strikeout categories; Jake Arrieta was the most difficult to hit in terms of batting average and H/9; and in terms of Game Score, Kershaw had the highest, Zimmermann was second, Hernandez’s middle 50% was the highest, and Arrieta had the smallest range between the 1st and 3rd quartiles. An argument could be made for each pitcher for June’s best, though I would select Kershaw as MLB’s most dominant in that time span. That doesn’t mean I’m taking a real chance with that argument. The numbers just swing that way.
There is another interesting fact. June 2014 was far and away the best month in terms of fWAR for any of the pitchers in their respective careers. This begs the question: how good was each pitcher in relation to his past performances? I went back through their monthly splits, finding the fWAR for each month since each pitcher came into the Majors. One caveat: I disregarded their rookie seasons. I did this because a rookie pitcher is brought up at different times, left in to get experience, taken out to minimize psychological damage, shut down early, etc. I didn’t feel like trying to account for 100 different variables that wouldn’t provide any more insight into what’s supposed to be an enjoyable experience. Anyway, with that being said, here’s how the numbers came out for our four hurlers:
|Player||June ‘14||Prev High||Prev Low||Avg||Standardized|
Personal Standardized Scores
Standardized score was derived by subtracting the average from June’s fWAR then dividing by the standard deviation.
I think it’s safe to say that Arrieta’s June is the most surprising compared to past performance. If we look at his previous high (twice, May in both 2014 and 2012) his 1.8 is 200% better and 600% better than his average month. Arrieta’s standardized score clearly reflects that as well, seeing that he’s scored well higher than the others. Up until this point in his career, Arrieta has been more about potential than realized performance. Well, so far in 2014 he’s begun to live up to the hype that once made Orioles’ fans so intrigued.
Once again, look at Kershaw leading this celebrated group with his average of 1 fWAR per month over his career. I don’t know if there’s anyone on the planet that still requires proof that Kershaw is one of the best pitchers out there, but if there is such a person, I doubt if this extra bit of information convinces him/her. Things like this continually amaze me. I also understand that you get it too, so I won’t belabor the point further.
Just for fun, I’ll include the R summary for their respective months.
|Player||Min||First Qu.||Median||Mean||3rd Qu.||Max|
R Summary fWAR
Yawn. Kershaw and Hernandez are really good, Zimmermann is underrated, and Arrieta is a player that up until this year had great stuff but never put it all together. I’m also running out of superlatives for the group, so I’ll just end it with the table above. This was fun. Let’s do it again for July.
[i] Pedro Martinez in 1999 owns the all-time record for a season in both K/9 and K% when he struck out 13.22 per nine innings with a 37.5 K%. Randy Johnson in 2001 is number two in both categories, coming in at 13.2 and 36.7% respectively. Why is this important? Kershaw just killed the month of June, and it still falls short of Pedro in ’99 and Johnson in ‘01.