There Simply Isn’t an AL Cy Young Frontrunner
When I started researching a post about the American League Cy Young Award, I was prepared to make a case in favor of Chris Sale. I know you’re not always supposed to go into these things with an outcome in mind, but, look! The rest of this post proves I wasn’t too biased. When I got a little into the work, I started imagining a somewhat contrarian argument in favor of Danny Duffy. That turned into my pursuit, until I became more convinced to support Corey Kluber. I was just about ready to begin a draft. Then I told myself, no, look at the numbers. The favorite should be Aaron Sanchez. I’ve assembled cases for all these guys. A few more, too. Start to finish, this wasn’t supposed to take more than a couple hours.
I wish I could give you something better. I wish I could give you a reason to lean toward one name. Truth be told, there are plenty of those reasons, but many of them point toward different names. It’s the middle of August right now, and there’s roughly a quarter of the season left. That’s going to settle the Cy Young race, because at least as far as I can see it, right now there’s just a multi-way tie.
The Cy Young conversation has picked up over the past week or two, with the pitcher in the middle being Zach Britton. The other day I laid out why I don’t think Britton makes for a great candidate. I could be persuaded otherwise, I’m sure, but that’s where I am today. In all the Britton talk, a big reason for the support is the perceived lackluster group of other AL candidates. There is no Clayton Kershaw. There is no 2015 Zack Greinke, or 2015 Jake Arrieta. There’s no freak. There’s no small assortment of freaks.
I do think the various candidates have been better than the credit they’ve gotten. Standards might be a little too high, and they might not yet have adjusted for the new home-run era. Low ERAs aren’t so easy to come by. Sanchez has a 66 ERA-. Jose Quintana is at 67. Duffy is at 64. Dallas Keuchel last year wound up at 62. The year before, Kluber was at 64. The year before that, Max Scherzer finished at 72. No, there isn’t a Kershaw, but there are aces. There is just a bunch of…somewhat indistinguishable aces. I think that can lead to Britton as almost an “easy” choice. Instead of putting in the work to try to separate the starters, you can just point to the one crazy reliever. I do get it. And relative to their own occupied roles, Britton has been literally the most outstanding pitcher in the population. If you’re open to a reliever winning, Britton is right there.
I’ve already told you I’m not that open. This post is about my own perspective, and not how I think the voters are going to see things. I’ve put in the work. I’ve tried to identify a starter whose case I like the most. I can’t get there. It does feel a little like trying to decide between Kershaw, Greinke, and Arrieta. The pitchers haven’t been on quite that level, but there’s that little separating them. It feels silly that only one pitcher can win.
I want to explain, but I want to do so treading lightly on the statistics. Right now I’ve got three different spreadsheets open, so I could proceed with an avalanche of numbers if that’s something I wanted to do, but I’m not sure it would be that helpful, especially with so many games still left. So we’ll try to stick mainly with words! We could use a starting point. Might as well be WAR. Kluber leads the group in WAR. There are five pitchers at the top within one win. Based just on this starting point, Kluber should be the current favorite.
And Kluber, also, is the leader in wOBA allowed. That’s a pretty pure statistic, that treats all plate appearances the same. Helping to support that number, Kluber grades well in terms of contact quality against. Outside of context, Kluber has been terrific. Even given context, Kluber has been terrific, but what gives me pause is that he’s performed worse with runners on base. Even more so with runners in scoring position. Those are more important opportunities, and it’s not like Kluber is out there unaware of his in-game scenarios. The whole job is to prevent runs from scoring. Kluber has allowed a few extra runs to score. It’s a certain point against him.
And when you open up those situational splits, contenders emerge. Sanchez has allowed a higher wOBA than Kluber. Quintana’s is higher still. Yet they’ve both been phenomenal with runners on. You can look up the splits yourself, if you want, but those two pitchers have killed run-scoring opportunities. Sanchez, under pressure, has eliminated walks and hard contact. Quintana, under pressure, has eliminated homers and boosted strikeouts. Again, the job is run prevention. It doesn’t matter if an opponent dies at third or in the batter’s box, so long as he doesn’t get home.
The list doesn’t stop there. By WAR, Sale is within a half-win of Kluber. He’s faced ever-so-slightly more challenging opponents. He’s pitched in front of an inferior team defense. And Sale has worked with a pretty unfriendly strike zone, on account of the White Sox’ catchers. Using my own homespun framing numbers, pulled from the FanGraphs leaderboards, pitch-receiving has cost Sale about five runs. There’s no way to know how he’s actually been affected, but he’s worked with a worse zone than he did a year ago. Presumably it hasn’t been helping. Sale would probably have better numbers if he had better catchers, and that isn’t his fault.
Quintana’s also pitched to those catchers! Interestingly, I don’t have him being negatively affected. His command is awfully good.
You can keep on going. Justin Verlander has allowed a lower wOBA than Quintana. If you’re a big fan of playing time, only David Price has pitched to more batters on the year. I constructed a statistic similar to K-BB%, only also folding in pop-ups and hit-by-pitches. It measures those elements mostly under a pitcher’s control. In that category in the AL, Verlander ranks second. The only name in front of him: Danny Duffy. Duffy is your candidate if you don’t care that much about playing time. It’s just about finding the best pitcher, right? I can see this going both ways. Since re-entering the Royals rotation, Duffy has been almost untouchable. He’s thrown just 20 fewer innings than Sanchez. There’s nothing he’s not doing right now. I guess he could allow softer contact, but mostly, he just allows no contact.
And, oh, what the hell, Michael Fulmer is second out of everyone in wOBA against. He’s appeared in just 19 games, so he’s pitched even less than Duffy, but he’s the ERA leader. You can’t not mention the ERA leader. I’m pretty comfortable not supporting Fulmer as the AL Cy Young, because there just aren’t enough other supporting numbers, but he’s not completely off the radar. Yet there are enough other names to try to sort through.
You could very reasonably go with any of these names. Kluber, Sale, Sanchez, Quintana, Verlander, Duffy…precedent says Duffy won’t get there because of limited innings, but I don’t know how much that should truly matter. He’s proven himself as a quality starter. Even if you think innings are critical, that still leaves you with four or five candidates at the moment. And I just can’t tell them apart. I mean, I can tell them apart, but I can’t tell who’s pitched the best. I’ve thought I could. I’ve put hours into thinking. I’ve gotten both somewhere and nowhere.
Thankfully, there’s more time left for the pitchers to pitch. And, thankfully, I won’t be officially voting. I can’t imagine the potential headache.
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.
I feel like the Cy Young should go to the pitcher that lowers the league’s FIP the most. I’d love to see a year-by-year break down of that and see how it compares to the actual Cy Young winners. Anyone looking for a project?
Wouldn’t that just be the pitcher with the lowest FIP to IP ratio? Wouldn’t tell you much more than finding the qualified pitcher with the lowest FIP season to season I’m pretty sure.
I’m not quite as well versed as a lot of people in the way that some statistical principles work, but I don’t think that is exactly what it would be… I’m going to go ahead and do some work on this though, because I’m curious. The idea would be have the league FIP and then take each pitcher’s stats out and of it and then see what the league FIP is after that. If two pitchers have an equal FIP and one pitched 30 more innings, then I think he’s the guy you go with. But what if the guy with 30 more innings has a slightly higher FIP? Is there a chance that he might still be considered better because he had a greater effect on more actual game time? What about a relief pitcher with a miniscule FIP – just how miniscule would it have to be for him to get the nod, if you agree with this sort of methodology? We shall see.
If anyone is curious, I ran the AL numbers…
Here were the AL pitchers with the greatest effect on the league FIP, through tonight, meaning that the league FIP of 4.2603 would be this much higher without their numbers included.
1. Corey Kluber 0.0128
2. Dellin Betances 0.0108
3. Aaron Sanchez 0.0093
Over notables from the discussion.
14. Zach Britton
20. Justin Verlander
23. Michael Fulmer
24. Rick Porcello
48. J.A. Happ
The bottom?
350. Clay Buchholz
351. R.A. Dickey
352. Jered Weaver
353. James Shields
354. Chris Young -0.0130
If you’d like a link to the Google doc, let me know.
Not even a mention for Rick Porcello?
Because you’d have to start putting Chris Tillman in the conversation right after him
I don’t see why not. If Porcello ends up 23-3, or Happ, and he gets his era down to about 3.20 (Kluber is at 3.10), that would I think influence voters. Tillman also. Although his FIP is a bit worse than Porcello’s, but with all these choices so close I think counting stats will weigh more.
Fulmer actually leads by baseball reference war. He has an innings limit. Duffy has been great.But of course he has baseball’s best D behind him. Kluber has the 4th best (by UZR) and Sanchez 5th best. Tigers are 20th.
I don’t like to necessarily go by FIP unless clear cut, but what actually happened, even though we know defense matters, etc. Kluber’s era and FIP about match up. Porcello’s don’t, but he has excelled nevertheless. I might not vote for him myself but I wouldn’t castigate those that do. I’d probably go with Duffy. But if Porcello or Happ gets to low 20s in wins with 3-4 losses, and Kluber only gets to 16 or so and doesn’t lower his era, they will have quite the case.
I don’t believe that Rick Porcello should win the Cy Young, but I agree that it’s a significant possibility. Voters value wins less than they used to, but they don’t discount them entirely.
One of the great examples of (over)reliance on counting stats:
1990 AL Cy Young
Bob Welch (winner) 27-6, 1.8 fWAR
Roger Clemens 21-6, 8.2 fWAR
1990 AL Cy Young
Bob Welch (winner) 27-6, 1.8 fWAR, not an asshole
Roger Clemens 21-6, 8.2 fWAR
Fixed it.
Worse yet, my favorite example: 1980 AL CY Young:
Steve Stone (winner) 27-7, 3.23 ERA, 3.99 FIP, 2.9 WAR
Mike Norris 22-9, 2.53 ERA, 3.25 FIP, 6.0 WAR
I was barely a teenager back then, but I recall thinking Norris was jobbed. And that was way before FIP and WAR stats were out there to show how bad that voting was.
Oh, I guess I didn’t see the 1.8fWAR, so maybe the Welch election was even worst than Stone’s
There was no such thing as fWAR in 1990. The man went 27-6 with a 2.95 ERA. By the criteria in use 26 years ago, there was no way he wasn’t going to win.
Welch had the same thing going for him that Porcello has so far this season. He was having a good year and his team was lighting up the scoreboard every time he pitched. But there’s no way that selection was as bad as choosing Steve Stone over Mike Norris in 1980. If Porcello won this year, it wouldn’t be as bad as that one.
And nothing was as bad as when they gave it to Pete Vuckovich in 1982 or LaMarr Hoyt in 1983.
I guess Welch’s election was a make-up vote for another A’s pitcher being jobbed 10 years earlier.
And I don’t know why I can’t edit my post where I said “worst” instead of “worse”
Stone only pitched one more year after that and not well.
Stone said later on that he threw an enormous number of curveballs in 1980 and it ruined his arm. But he said he’d do it all over again if he could, because that season was worth it to him.
I went into this searching for a No. 1. I could never make Rick Porcello my top candidate. He is very good! But I wasn’t writing about who I think could win. I was writing about who I would support right now. And I don’t care that Porcello is 16-3.
I get that. By bref WAR it’s a bit closer (4.4-3.7) than fip-based. He pitches in a tougher park somewhat, and maybe tougher road parks. AL East isn’t what it used to be and he isn’t facing the Red Sox. Half a walk less per game. Slightly worse D.
My choice is leaning Duffy if he keeps it up.
Maybe if this was 2006 and not 2016. Porcello isn’t even the best pitcher on his team.
Price is still pretty damn good, right. Still, that 4 era even in a more sabr friendly era.
Back in June when Tony did contact management update, Porcello was slightly ahead of Kluber. Kluber’s tru era- (based on launch angle etc) was slightly better, but only 82 to 83. Price was at 93 and his contact score much worse. Sale’s tru was much better, at 65. But he has slumped since then I think. Quintana also very good.
http://www.fangraphs.com/blogs/american-league-contact-management-update/
This is why I hope fWAR one day includes contact management.
As a Red Sox fan who has watched most of their games this season, I would disagree. In terms of true talent, Price is still better than he is. In terms of how well they have pitched this season, Porcello has been the best pitcher on the team. Steven Wright has been better than Price as well. Not all of Price’s problems are bad luck. Bad luck hasn’t made David Price hang an ungodly number of pitches with men in scoring position; he’s done that on his own. And the first person who would tell you that is David Price.
My vote would be Kluber right now. His updated ZiPS projection has him finishing the year with 6 WAR, 220 IP and an FIP and ERA right around 3. Probably end up with somewhere around 18 wins to go with plenty of Ks, and the Indians are one of the best teams in baseball which voters like.
Now a related question I have – can someone explain to me how Fulmer can have 5.1 WAR and be #1 in baseball over at BBREF, but be #29 at Fangraphs with 2.6 WAR? That is an enormous discrepancy. Is it because BBREF uses something closer to RA-9 WAR? I understand that these are two different orgs using different methods, but it being 2x as large over at BBREF just rubs me the wrong way. That’s not a marginal difference.
fWAR is based largely on FIP, bWAR uses runs allowed. fWAR is what should have happened in a just world; bWAR is what actually happened.
Basically, Fulmer has been insanely lucky. ERA 2.25 vs xFIP 3.75.He’s pitched well, with the great random number generator in the sky giving him great results.
But fip is still hypothetical and it isn’t even what should happen in a just world. It is the average result of balls in play Ks walks and homers in front of an average D. It looks like Kluber pitches in front of a better D than Fulmer. I take fip, xfip into consideration too but contextualize, and I give some weight to actual results. I agree Kluber has been a bit better. More Ks, fewer walks. Fulmer more grounders counts a bit. I guess contact management counts a bit. Have to ask Tony B.
no, they both are what “should have happened” if you are talking about runs.
the only one that measures what “actually” happened is ra9/war.
i don’t think so. It doesn’t say what “should have happened” as far as runs scored. FIP doesn’t know anything about what kind of contact a pitcher gave up. Take Hendricks and Roark. They are 1st and 2nd in most soft contact and 2nd and first in least hard contact. Pomeranz is 2nd lowest line drive rate, 16th lowest in hard contact and 17th in soft contact. They all have low babips and ERAs well under their FIPs. And they have pitched pretty well even by FIP. Now, these may not be repeatable skills, but that’s the type of contact they allowed.
FIP only looks at average results, not what type of results occur given the type of contact actually allowed. I don’t think you can say, well, Pomeranz “should” have allowed 20% line drives, not 15%. He’s just lucky. Given the actual batted balls he gave up, I’d say his babip, and hence his ERA, more correctly reflects what “should” have happen in front of an average defense, given his K rate, walk rate, and homer rate. Ditto with hendricks and Roark, although to somewhat lesser extent.
I do look at xFIP and siera more than era, but there are times when era may reflect something else. And just because a skill isn’t necessarily repeatable doesn’t mean that the results were all just luck. Many times yes, sometimes not. I like to dig a little deeper, especially when weighing who has been better thus far.
These links might help explain it:
http://www.fangraphs.com/library/war/differences-fwar-rwar/
http://www.baseball-reference.com/about/war_explained_comparison.shtml
BR’s founder has stated that “fWAR is better for looking forward (hence it’s reliance on FIP on the pitching side), but bWAR is superior for backwards looking at what happened during the season.” I would make a different distinction, arguing that bWAR is recording essentially what did happen while fWAR is attempting to estimate what should have happened. You can see this on the leaderboards when looking at ERA minus FIP. Michael Fulmer, Jason Hammel, Kyle Hendricks, and Cole Hamels all have ERAs below 3, but are those results an accurate reflection of their performance or are they heavily influenced by luck and sequencing? I’m not a huge fan of FIP because there are elements that I feel it fails to properly weight, but I strongly dislike pretending that luck doesn’t exist or somehow isn’t relevant, so I lean much more towards fWAR.
But for the purposes of the Cy Young award, which is better? As Sullivan pointed out in his Britton piece, the award (now) says “Most Outstanding Pitcher.” Is your definition of “Outstanding” something like the “best true talent (at least this year)”? Then you should use fWAR. Is your definition of “Outstanding” something like “the guy who got the best results (at least this year, no matter how he got them)”? Then you should use bWAR. At least “outstanding” doesn’t seem to invite considerations of team postseason trajectory or relative worth on a roster the way “Valuable” always, ridiculously, does.
I favor fWAR because I think bWAR incorporates more elements outside of the player’s control.
I know wins aren’t the most important stat, and maybe Cy Young voters are wising up to that, but on a whole, they’re still going to get behind the current and forever AL wins leader, J.A. Happ, and he at least deserved a mention.
I would have also liked Happ to be mentioned in the article, not that I think he should win the award but because he has been seen as an early front runner on some sites, so clearly he deserves a mention.
“This post is about my own perspective, and not how I think the voters are going to see things.”
“precedent says Duffy won’t get there because of limited innings, but I don’t know how much that should truly matter”
But the argument against Britton is either, a) he’s a failed starter, or b) he doesn’t pitch enough innings.
I ignore a, because I don’t give a crap what has been done in the done.
And now are we to ignore b because Duffy’s inning truly shouldn’t matter?
I know there is a big difference between say 70 and 160 – but innings don’t matter right?
Sullivan didn’t say innings don’t matter. He wondered aloud how *much* they should matter given that Duffy is only a handful of starts’ worth behind. Duffy could still finish the year with 190+ innings. Kershaw won the MVP with 198.
Duffy has thrown 114 innings as a starter. Britton has thrown zero innings as a starter. Starting is a lot more difficult.
I would like to see your justification for this statement. The increased value attributed to starting is solely due to their larger number of innings, not any contention that beginning a game is more difficult than ending one.
Starters:
1. Throw more innings (can’t rear back and throw their best stuff for around 20 pitches like relievers)
2. Often face lineup more than once, meaning that hitters have time to adjust and see more pitches
You are right that the start of a game and end of a game don’t favour the starter, but I would also argue if you gave even a middling starter the ball in the first inning and said: just get through the first inning and you are done, there would be a lot more zach brittons
There would be a lot more guys with a sub-1 ERA and a FIP of 2? You must have been a starting pitcher in high school
No graphs, no imbedded video. Bummer.
Have they really, or have they just been somewhat lucky in those situations? To answer that question it’s helpful to look at career splits in order to create a larger sample size. Kluber’s inferiority with runners on base holds true, as his career wOBA with runners on-base is .045 higher than his career wOBA with the bases empty. He is simply a better pitcher out of the wind-up than out of the stretch, and therefore slightly discounting him is perfectly fair. On the other hand, the wOBA advantages for Sanchez and Quintana with runners on completely disappear when looking at the larger career sample. Sanchez is .001 better with runners on over his career while Quintana is .011 worse with runners on.
You could make the argument that Quintana is doing something different in 2016 that merits extra credit, but I don’t find that to be the case. Jose has always decreased his percentage of pitches within the strike zone with runners on base, but this year for whatever reason hitters are chasing more and making less contact in those situations. Another huge factor is Quintana’s 3.4% HR/FB% with runners on base compared to his 10.5% HR/FB% with the bases empty. Even for those who argue that Jose has some homer suppressing ability, 3.4% is unsustainably lucky. Basically, what I’m saying is that Kluber’s results with runners on are legitimate and worth considering yet the apparent improvement for Sanchez and Quintana is in actuality just noise.
As for who should win the A.L. Cy Young, I have absolutely no idea.
Dan Szymborski posted Cy Young predictions on ESPN. #1 – Cole Hamels…..don’t see a mention here.
I know what you mean and I agree with you in terms of projecting how these pitchers would look going forward, but I think the single-year results should count for something. Sanchez has allowed a .283 wOBA with the bases empty, and .240 with RISP. He has split FIPs of 3.79 and 2.84. For Quintana, he’s at .299, .259, 4.12, and 2.42. It’s way too early to say these pitchers are actually, naturally, better in pressure spots. But they’ve *been* better in those spots, and demonstrably so, and I don’t think that shouldn’t count for anything. Timing is a part of success in hindsight.
Having a 3.4% HR/FB% with runners on base isn’t skill, though. That’s being extraordinarily lucky. No pitcher can suppress home runs to anywhere near that level, and unfortunately FIP doesn’t take luck on HR/FB% into account.
xFIP does! And there his split is 4.54 and 3.50. 18% strikeouts with the bases empty. 28% with runners in scoring position.
With Kershaw out for so long you might end up with a similar problem on your hands for the NL, Jeff.
I wouldn’t vote Happ, but he is having a surprisingly effective year.
I’m pretty surprised JA Happ isn’t even mentioned here. He may not match up in FIP and therefore fWAR, but he is the front runner when looking at traditional metrics. For example, the ESPN Cy Young forecaster has him as the current favorite. That is strictly a model based on prior Cy Young voting
http://www.espn.com/mlb/features/cyyoung
I’m really hoping Duffy finishes strong, but typically CY winners make the rotation out of spring training (I believe KC hurt his overall development by shuffling him back and forth between starting and relieving). As for Zach Britton, he’s having a great year, but I couldn’t support him winning practically by default, nor due to the facts that Aroldis Chapman did not pitch the whole year, Wade Davis has battle injuries and Dellin Betances has not closed most of the year. Going with Kluber unless Duffy beats the world over the last 6 weeks.
Doesn’t it sort of seem like voters should vote on what happened, rather than what we can expect to happen going forward? Going forward one can reason Fulmer should likely not perform like he has. Whether he continues to be a guy who outperforms his peripherals remains to be seen, but he won’t perform this much better than his peripherals going forward.
With that said, I would think Cy voters should focus more on what has actually happened than what can be expected going forward. I think this is akin to someone stating that since Mookie leads the league in wind-aided lucky (really some had true distance under 300 feet and 90mph exit velocity) home runs that those don’t count and cannot factor into voting. Before his two home run game at Camden I figured out his “true” line would look more like .292/.335/.498 if we discount only half his “lucky” home runs. But the fact is those home runs did happen, and they helped to determine the outcome of real games. Isn’t that what the CY and MVP award should be based on? What happened rather than what we can expect to happen going forward?
It’s not just “what we could expect to happen going forward”; it’s also “what we could expect to happen if we re-ran this season a thousand times.” It’s an attempt (limited and flawed, granted) at getting at true-talent, rather than true-talent-plus-luck. Now, as I mentioned in a post above, you can choose your definition of “outstanding pitcher” to reflect what actually happened when those pitchers pitched, or you can choose it to reflect what we think those pitchers actually are. Either is defensible, and neither — this year — gives us a clear front-runner.
I always average fWAR and bWAR to come up with “average WAR” (aWAR) for pitchers because in a single season. It just makes too much sense to do it this way.
A batter would get extra credit for things like clutch performance of batting with RISP and things of that nature that aren’t entirely under his own control and/or a repeatable skill … as should pitchers get some credit for LOB%, BABIP, etc. It’s not like batters, all year, just hit liners right at the defense or the just so happened to have weak contact, etc whenever runners happened to be on. Ya gotta give the pitcher credit “some” for that. So, when my choices are [1] All or [2] nothing, I average it and split the difference.
Quintana and Kluber are tied with 4.3 WAR and Sale & Fulmer come up with 3.8. I go with Kluber.
NL CYA would be Kershaw at 5.1 aWAR … and then these guys tied at 4.3:
Thor, MadBum, Max.
The question might be “What’s the least amount of IP CK can pitch and win the CYA?”
I’ll use this opportunity to rehash a hobby horse of mine…..
1.going deep into games is a skill;
2.the last few outs, when your pitch count is high and batters have seen you 3-4 times, are much harder outs than earlier outs.
3.going deep effectively not only means your team only has to rely on the top 1-2 RP if the game is close and not fungible middle relievers, dramatically improving the chances of a win, but also saves a significant longterm load on the pen.
4.The difference between the top IP/GS and bottom guys is 30+%, which is massive, and again that’s the toughest 30% of the outs to get – comparing the two is like comparing a fulltime player with a platoon player.
5.the argument that there is too much manager’s discretion here doesn’t hold up anymore with most every team respecting the 100 pitch limit and using similar bullpen stalrategies. not to mention that there is manager discretion with the usage of all other players too.
6.WAR doesn’t properly account for this skill. In fact it ignores it completely. We shouls fix this.
I agree with this. WAR does of course to an extent account for the fact a guy has pitched more innings.
If you can’t ignore W and ERA entirely, why isn’t Happ in the discussion?