Jake Arrieta’s Argument for the Best Season Half Ever

Sunday night, Jake Arrieta came within sniffing distance of doing the almost unthinkable. By which I mean, Arrieta made a serious bid to hit two home runs. He also, at the same time, flirted with a perfect game against the Pirates, but that part is very thinkable. I don’t know how many times this year Arrieta has grabbed attention for taking a no-hitter or a perfect game deep, but it numbers somewhere in the “a lot”s, with Arrieta more or less existing on the verge of history. It doesn’t take a no-hitter bid to put him in that position — the bid is practically a foregone conclusion.

Eventually, Arrieta gave up a hit and put multiple people on base, but none of those people happened to score, Arrieta spinning another seven shutout innings. Two batters of a total of 22 reached, and one of them only did so because Arrieta did him the privilege of hitting him with a pitch. The outing was timed well, what with the Pirates being a rival of the Cubs. The outing was timed well, what with Arrieta in the running for the Cy Young award. And the outing furthered Arrieta’s case for maybe having the best season half that ever there was. However arbitrary season halves are, we’ve been splitting seasons at the All-Star break forever, and what Arrieta has done since the break legitimately defies belief.

The math: 14 starts, covering just over 101 innings. All of them have counted as “quality starts.” All but one, the Cubs have won, and in the loss they got shut out. Arrieta’s given up a dozen runs. Nine have been earned. He really has averaged less than a run allowed an outing.

For the sake of comparisons, we can make use of the Baseball-Reference Play Index. It makes everything that follows pretty simple. Let’s get started, looking for the best ERAs ever. I know we don’t love ERA, but that’s why this is a starting point. You can’t not acknowledge it somewhere. I decided to define a qualifying season half as one in which the pitcher started at least 10 games. I think it’s a good-enough cutoff, and now here’s a top 10:

Top 10 Season-Half ERAs
Pitcher Half Year IP ERA
Ferdie Schupp 2nd Half 1916 113.3 0.71
Walter Johnson 1st Half 1918 166.3 0.76
Jake Arrieta 2nd Half 2015 101.3 0.80
Dutch Leonard 1st Half 1914 149.3 0.90
Kris Medlen 2nd Half 2012 95.3 0.94
Roger Clemens 2nd Half 1990 92.7 0.97
Bob Gibson 1st Half 1968 160.7 1.06
Tom Seaver 2nd Half 1971 139.3 1.10
Pete Alexander 2nd Half 1915 178.3 1.11
Spud Chandler 2nd Half 1943 121.0 1.12
SOURCE: Baseball-Reference

Just starting with ERA, we find second-half Arrieta in third place. Baseball has taken place over many years! So, while Arrieta would have more company if his numbers were worse, instead his numbers are fantastic, and his peer group is small. I can’t speak to any players the search might’ve left out, perhaps because their data is incomplete, but consider this a post about statistics we know. No use involving players who don’t have full statistical records.

The obvious next step from ERA — raw RA, folding in unearned runs. That top 10:

Top 10 Season-Half RAs
Pitcher Half Year IP RA
Jake Arrieta 2nd Half 2015 101.3 1.07
Kris Medlen 2nd Half 2012 95.3 1.13
Roger Clemens 2nd Half 1990 92.7 1.17
Ferdie Schupp 2nd Half 1916 113.3 1.27
Tom Seaver 2nd Half 1971 139.3 1.29
Dutch Leonard 1st Half 1914 149.3 1.33
Bob Gibson 1st Half 1968 160.7 1.34
Bob Knepper 1st Half 1981 86.3 1.36
Johan Santana 2nd Half 2004 104.3 1.38
Jose Fernandez 2nd Half 2013 68.0 1.46
SOURCE: Baseball-Reference

Just by runs, now we get Arrieta in first. And many would argue runs are a better measure than limiting to unearned runs. You’ll see some recent years, here, because offense has trended down, and these numbers are unadjusted for context. But just consider the message here: by runs per nine innings, Jake Arrieta’s second half has been the best season half — that we know of — all-time. Emphasis on “all-time.” It’s not something that can’t be debated, but there’s no debating the significance.

Let’s leave these metrics behind, though. Let’s throw away a little bit of sequencing and look simply at how the pitchers have been hit by their opponents. I went into the numbers and manually calculated wOBA allowed, with help from ours Guts page. Though Baseball-Reference makes OPS figures available, I thought I might as well go to the next step. Another top 10:

Top 10 Season-Half wOBA Allowed
Pitcher Half Year wOBA
Jake Arrieta 2nd Half 2015 0.192
Ferdie Schupp 2nd Half 1916 0.192
Reb Russell 1st Half 1916 0.198
Clayton Kershaw 2nd Half 2015 0.202
Johan Santana 2nd Half 2004 0.202
Pedro Martinez 2nd Half 2000 0.203
Burt Hooton 2nd Half 1981 0.205
Greg Maddux 1st Half 1995 0.208
Sandy Koufax 2nd Half 1965 0.208
Joe Horlen 2nd Half 1964 0.211
SOURCE: Baseball-Reference

Once more, we find Arrieta in first. Now, he’s in first by the smallest of possible margins, but he does hold the tiebreaker over Schupp if you keep following the decimals further to the right. And Schupp had his second half literally a century ago. It’s true that, since the All-Star break, Arrieta has allowed baseball’s second-lowest BABIP. That’s partially fueling this, but then, there’s a difference between talking about true talent and talking about results. Since the break, Arrieta leads baseball in groundball rate. He’s among the leaders in soft-hit rate and hard-hit rate. He’s given up just two home runs. He hasn’t been hit hard, so why try to pretend otherwise? About that .192 wOBA — Giants pitchers this year have a .208 wOBA. So, there’s that.

Of course, I have to note that, in that same table, 2015 second-half Clayton Kershaw is fourth. This is a good award race.

Now one last table, introducing an adjustment of sorts. For every pitcher, I calculated wOBA allowed. We also have league wOBA, so I created a wOBA- statistic, dividing wOBA allowed by the league mark and then multiplying by 100. Our last top 10:

Top 10 Season-Half wOBA- Allowed
Pitcher Half Year wOBA-
Pedro Martinez 2nd Half 2000 59
Jake Arrieta 2nd Half 2015 61
Johan Santana 2nd Half 2004 61
Ferdie Schupp 2nd Half 1916 62
Greg Maddux 1st Half 1995 62
Reb Russell 1st Half 1916 63
Clayton Kershaw 2nd Half 2015 64
Pedro Martinez 1st Half 2000 65
Burt Hooton 2nd Half 1981 65
Nolan Ryan 2nd Half 1986 66
SOURCE: Baseball-Reference

Pedro basically had to take the lead. Though his wOBA allowed that half-season was 11 points worse than Arrieta’s, the league wOBA now is 27 points lower than it was when Pedro was going to work. So the adjustment allows Pedro to vault out in front. I wouldn’t consider it an insult to argue that Arrieta might be having a slightly worse half-season than Pedro Martinez had at his peak. Arrieta, of course, is second here. Further adjustments would rearrange the table — I could split up the AL and NL wOBAs. Pedro was in the AL; Arrieta is in the NL. I’m also not considering park effects. You can’t adjust for everything, and sometimes it’s okay for a stat to be imperfect. All the stats are imperfect. By just about any imperfect stat you look at, Arrieta is having an all-time season half. If not the very best, then one of them. And there has been an awful lot of baseball.

If the schedule keeps up, Arrieta gets one more go. According to the Cubs website, he’s to be the Friday starter against the Brewers. Arrieta, naturally, is going to start the wild-card game against the Pirates, but that isn’t scheduled until October 7, which would put Arrieta on regular rest. So there’s one more chance for him to boost his numbers. The outing will probably be abbreviated, but it’ll bring an end to his second half. It’ll set the numbers in stone, and then we’ll have an even better idea of how this season half compares to all the others.

But it’s enough to say: it compares really well. It’s been a historic half season. And we split the numbers at the All-Star break. Arrieta pitched against the White Sox the Sunday before. He allowed a run and two hits in nine innings.

Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.

Newest Most Voted
Inline Feedbacks
View all comments
8 years ago

I know this is not the take away from the article, but what was up with the Washington Senator’s defense in 1918? From the looks of it, at least HALF of the runs Walter Johnson allowed in the 1st half were unearned (he goes from 2nd on the ERA list with 0.76 to not even in the top ten of the RA list).

I mean, part of it was the era (not ERA), but Schupp doesn’t have the drastic shift that Walter Johnson did.

Nathaniel Dawson
8 years ago
Reply to  Jaack

Gloves that were more like mittens. And fields that were more like your neighborhood vacant lot.

8 years ago
Reply to  Jaack

Well, the Wikipedia entry on Griffith Field (quoting a book about the man it was named for) mentions “…the unkempt field that can be seen in photographs of Griffith prior to 1923.” (The quote in the book specifically refers to the infield.) So that could be part of it.

And part of it could be the official scorer, which in those days was typically a newspaper reporter with fairly wide latitude in how the scored plays. Maybe he (it was usually a he, but not always) was hard-nosed about what constituted normal fielding plays and awarded a lot of errors; maybe he just liked the pitchers. Looking at the pitching staff on that 1918 team, all of them had a difference between raw runs and earned runs of 20 or more. And that wasn’t a one-year thing: Walter Johnson’s B-R page shows he had a wide gulf between runs and unearned runs for several years in the late teens.

And it definitely seems to be a home field thing: just looking at 1918, and summing up the difference between runs and unearned runs, Johnson had a delta of 16 at home and just 9 on the road, so clearly it was something to do with playing in Washington, though that doesn’t tell us if it was the park or the scorer.

8 years ago
Reply to  Jaack

People keep talking about the NL Cy Young award as if it is a three horse race between Arrieta, Kershaw, and Greinke.

But I think the winner will be Arrieta in a landslide because he has the following:

(1) the second half narrative (a compelling story, so to speak);
(2) 20 wins;
(3) a no hitter; and
(4) various advanced statistics that support his case as the best pitcher.

Sure, on this site 20 wins might not be what authors and readers of Fangraphs use as voting metrics. But that still matters to most of the voters quite a bit. Not as much as in the past – King Felix won with a 13 win vote total in 2010 – but if the voters see a guy who has a strong narrative, good traditional stats (i.e., 20 wins, no hitter), and is at the top near the advanced statistics, I don’t think Kershaw and Greinke come close to winning.

The two of them will get a few first place votes, but they’ll be also rans at the end of the voting, I think.

8 years ago
Reply to  Bat

You are correct. It will be Arrieta first, Greinke second and Kershaw third I imagine. The two Dodgers should probably be the other way round though.

8 years ago
Reply to  Bat

Not just that, but there’s …

[1] The Cubs good team record.
[2] The Cubs exciting playoff story.
[3] Greinke and Kershaw splitting votes because of being on the same team/region.

These aren’t primary reasons … but the primary reasons are so close that other reasons start to matter more than usual.

8 years ago
Reply to  Bat

I mostly agree with you, with the caveat that Kershaw could pull off an upset if he pitches again Sunday and hits 300 K. Even if an extra 6 Ks are tiny compared to the ridiculous things he, Arrieta, and Greinke have been doing all season, voters seem to like big round numbers, and putting up the first 300 K season in 13 years might knock them over.