On Monday, I detailed how I went about voting for the NL Rookie of the Year. With Chris Coghlan something of a surprise winner and left off of my ballot, as I write this I’m curious to see how my Cy Young ballot comes out. Again, as I’m in an NL city (Cincinnati chapter), I vote for NL awards.
Starter GS IP WHIP K/9 WARP3 SNLVAR SNWP Chris Carpenter 28 192.2 1.01 6.7 8.5 8.0 .673 Dan Haren 33 229.1 1.00 8.8 9.8 6.7 .588 Jair Jurrjens 34 215 1.21 6.3 8.6 8.0 .608 Tim Lincecum 32 225.1 1.05 10.4 8.7 8.2 .638 Javier Vazquez 32 219.1 1.03 9.8 8.9 7.4 .608 Adam Wainwright 34 233 1.26 7.1 8.3 8.5 .630
As much as I agonized over the RotY ballot, my Cy Young ballot went quickly, mostly because I hadn’t been informed that I was going to be voting for it. Jack O’Connell, the longtime secretary-treasurer of the BBWAA checked in with me about my ballot on the due date, giving me hours, not weeks to come to a decision. It wasn’t hard for me. I wasn’t flippant, but I knew who my votes were going to be. I made a quick check of the stat pages, a couple of calls to Joe Sheehan and John Perrotto, and then turned in my ballot. While all were close in value, my three picks weren’t straight down the line. I felt that Dan Haren had been overlooked because of his team’s performance, but that his performance for them was definitely worthy of recognition. I thought that the Cards had two quality contenders for the Cy Young, but that Chris Carpenter’s time away for injury tipped the scales over to Adam Wainwright between the two of them, that despite the value arguments; consistency and availability are two traits that I don’t think get measured well, but they have clear value for a pitcher.
So, that left a decision between Tim Lincecum, last year’s winner, and Wainwright. I called three players and one scout, asking for their opinions. One player hadn’t faced Lincecum-“lucky break,” he said-but he felt that Lincecum looked more hittable. “I’m still convinced that deception is a big part of what Lincecum does,” another said, “and that unless there’s a new wrinkle, people are starting to figure him out. He’s still good, his [stuff] is still good, but comparing him to Wainwright? Wainwright was just a shutdown guy this year.”
The value of two extra starts might not seem like much, but in a year which featured only one close divisional race, perhaps that marginal value per start isn’t as great in most years, but it’s those starts that tipped my vote.
My final ballot for the NL Cy Young went Adam Wainwright, Tim Lincecum, and Dan Haren. I’ve second-guessed the Haren pick since turning it in, and maybe I didn’t give Jair Jurrjens enough consideration, but that’s a ballot I feel comfortable with. After turning in my ballot, I sat down with MLB.tv and watched each start for my three vote-getters, plus Jurrjens and Carpenter. The more I watch it, the more I feel like Wainwright and Lincecum are in a dead heat. Lincecum is more dominant, while Wainwright is the better pitcher. The subtle changes Wainwright makes from inning to inning and start to start are more visible in a burst-viewing of condensed games.
As far as a process for voting, Ring Lardner hardly imagined this way of doing it, and I’m not sure it’s ideal, but in a season where I only saw one of these players live (Lincecum, in Cincinnati), I feel that I have to make the effort to see as much as possible.
That is at the heart of the criticism I’ve heard. There are infinite numbers, stats, and data sets out there with different ways to argue them. Saying which pitcher is the best is a lot like picking your favorite pie. I like apple pie, peach pie, cherry pie… there are probably hundreds of pies, and while you might like pumpkin pie, I don’t, and that’s okay. Trying to visualize the subtle differences between 120 starts of five pitchers is a Sisyphean task when you could just as easily come up with the same result by gut, by throwing darts, or by looking at a WARP leaderboard. Criticize the picks if you want, but in a year where the BBWAA opened up its voting to members they knew might think a little differently, I have to feel like it’s a step forward. Now, again, about those Gold Gloves…
Thank you for reading
This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.
Subscribe now
http://sports.espn.go.com/mlb/news/story?id=4671110
I stand by the statement.
:-)
I had no idea that evaluating a basically objective question -- "Who was the best pitcher in the league?" -- was so thoroughly nebulous.
Does the same go for answering the question "Who is the best player in the league?" When some braindead writer from Philadelphia is shown to have voted for Ryan Howard next week, is the proper reaction just to shrug and say, "Well, there's no way to really know who was better, Howard or Pujols. Picking between those players is like choosing between apple and cherry pie. Any answer is basically right."
Is the proper reaction really just to say, "Sure, Pujols leads Howard in all the major objective measures, but there are infinite numbers, stats, and data sets out there with different ways to argue them, so who's really to say which ones are right -- the modern statistics that attempt to adjust for context and integrate a relatively sensible understanding of the game, or the statistics created 100 years ago, that mainly rely on countin' stuff up and seeing who has more?"
I know Baseball Prospectus has gone through some major changes in the last nine years, but this wink-and-a-shrug approach to evaluating a pretty basic, and essentially objective, problem -- "Who was the best?" -- strikes me as just a little odd.
I know the popular P.O.V. to emerge in the baseball analysis world over the last five years has been "Neither stats NOR scouts! Use all the available information! Don't put too much trust in any one number!" And that's a basically admirable perspective. Skepticism is almost always a good thing. But I'd be lying if this "apple pie" approach to voting on the NL Cy Young didn't strike me as an attempt to SEEM non-partisan, but in reality is just, frankly, a little intellectually soft.
Tim Lincecum lead both Carpenter and Wainwright in all metrics that demonstrate an understanding of what pitchers can and can't control. BP's own QERA reflects this (Lincecum: 2.83; Wainwright: 3.51; Carpenter: 3.68), as do the major metrics hosted at FanGraphs. Personally, I don't really see how watching every start from these pitchers on MLB.tv supports or refutes what that data pretty strongly suggests, and I don't really see a compelling argument for why anyone should pay attention to statistical information that doesn't even attempt to separate pitching from defense, even if there are "infinite ways" to argue about those -- outdated and unadjusted -- numbers.
I sincerely love Baseball Prospectus. I'm just baffled by this perspective on what "value" means, what's the best way to measure it, and the general validity of modern baseball analysis versus a more "traditional" approach.
Those who would apply only statistical measures in making their decision still need to choose and weight the stats. It isn't true to say that Lincecum swept the board among the newer measures, either; see the posted WARP3 and SNLVAR results.
http://msn.foxsports.com/mlb/story/10398050/Giants'-Lincecum-wins-second-straight-Cy-Young
it will be interesting to see how the beat writers of america will handle you and Keith Law 'jobbing' Carpenter out of his Cy Young by voting for 'one offs'.
Will they hold off or dig up Law's BP roots and say 'I told you so'?
Not saying you made bad decisions, just going to be interesting how they get dealt with in the media.
% of pitches that result in balls in play:
Lincecum - 16.6% (8th of 120 among pitchers with 100 xIP)
Wainwright - 19.1% (48th)
% of pitches that result in a swinging strike:
Lincecum - 11.0% (7th)
Wainwright - 9.2% (40th)
% of contact made when swinging at all pitches:
Lincecum - 75%(3rd out of 75 qualified)
Wainwright - 79.1% (25th)
Lincecum is much less hittable.
Also, hopefully it's just an oversight, but Lincecum throws 3 pitches regularly and a 4th about 5% of the time.
Tim IS trying to strike everyone out, AW is trying to get balls hit weakly into play.
I'm willing to believe your player source can probably tell who was making better pitches, based on film and conversations. There's no way he has a clue who was getting better results (much less defense-adjusted results), not without looking at the numbers. And there's absolutely no way he can account for the different number of starts and innings.
But I think 10.4 K/9 vs. 7.1 tells you everything you need to know about hittability. I think guys only felt like they were doing better because they expected so little from themselves. Maybe BP has done a good job convincing people that strikeouts aren't such a bad thing and that these hitters thought they were helping their teams as they were walking back to the dugout after strike 3.
BP lobbied for this, and the premise was we'd get a more thoughtful and quantitative vote from their representatives. This one seems more like an intuitive, haphazard vote.
And my favorite pie is any. Great race this year, all 3 top guys would be deserving winners.
I'm also not sure that asking a player's perception should factor much since I'm sure Greg Maddux looked quite hittable too.
Basically, Carpenter increased his team's chances of winning by quite a bit in the 28 games he started. Wainwright increased his team's chances, by not quite as much per game, in 34 games. Given how close they were in performace, that means Wainwright added more expected wins -- which is pretty much what the SNLVAR column is looking at.
I'm dying to see the analysis of how Haren could be so far ahead in WARP (and PRAR), but trail in SNLVAR. That seems to be the crux of the case for Haren.
I guess I don't quite understand the intricacies of WARP/SNLVAR enough to comprehend the difference or know which metric is better for determining the best pitcher...
Will explained how he came to his results in a short article. A person has to be allowed to have said and done more than he actually reports. This is a thorough community but BP staff aren't always writing white papers.
As media become ever more interactive, it's reasonable to expect that articles themselves will probably eventually die out in favor of Google Wave-type media and Twitter-like communication - information on demand. People should grow increasingly reticent to commit to any thought in a format that could be seen as permanent, since it's so easy for one or 750 words to disproportionately affect your livelihood.
Modern pitching stats provide a greater level of analysis than anything we've ever had, but by universal admission there are gaps in the data. It is no wiser to vote in lockstep with a set of numbers than it is to ignore them altogether. This is not the AL MVP race where the numbers are so overwhelming. Margins of error are such that excoriating Will for contemplating a few outliers in this race is the same sort of bullying groupthink that the Plaschkes of the world engage in. I for one prefer a person who is willing to admit that he thought a little different. It's not like Will went with wins or "makeup." Ultimately, the problem we are dealing with here is a voter sample size issue. The number of voters should be increased to render out the furor.
Second, his ballot is not unreasonable. It is not as though he put Carlos Zambrano on his ballot. If I had a vote, my ballot would have been a little different, but so be it. Reasonable people can have reasonable differences.
There are, however, two things in the article which I find disturbing.
First: "Lincecum is more dominant, while Wainwright is the better pitcher." Huh? You can be better by being less dominant? Maybe its just sloppy writing, but as written it is so utterly illogical that he shouldn't be allowed outside, to say nothing of allowed to vote.
Second, Will seems to suggest that he was swayed by one player -- that's right, a sample size of one (who didn't even face Lincecum!!) -- who said "people are starting to figure him out". Really? Other than hearsay from a passive observer, what is the evidence of this? The league-leading VORP? The 10.4 SO/9? Yeah, they're really drawing a bead on the guy.
Again, I don't think Will turned in a silly ballot, and the above arguments are a bit harsh in light of that. But I do think it is fair to suggest that some of the new voices in the BBWAA aren't that different from the old.
I would guess as well that Will probably means something similar to what you suggested. What I (over)reacted to was not an interpretation, but what he actually wrote. I did try to leave an out for "sloppy writing".
Speaking of sloppy writing, in my own comment I labeled my points of fault-finding as "disturbing", which is far too strong of a word. "Concerning", perhaps. Mass shootings are disturbing; nothing Will wrote is going to give me nightmares. Also, I referred to myself as "reasonable", which is an obvious falsehood.
That being said, it's the first time he voted as a member of the BBWAA (and most likely, wrote an article about voting). I didn't find it sloppy, though I had an odd impression that he would only let himself vote for one Cardinal and had to reread it to clear that impression.
I guess another part of my wishful thinking is if BP was to try something innovative, they'd use how we vote on their Internet Baseball Awards to determine the BBWAA voting.
Ironically, the NL Internet Awards had the ranking _exactly_ the same as the real Cy Young voting... Linecum, Carpenter, Wainwright, Vazquez then Haren.
http://baseballprospectus.com/article.php?articleid=9758
"I made a quick check of the stat pages, a couple of calls to Joe Sheehan and John Perrotto, and then turned in my ballot." On the other hand, you later mention that you asked 3 players and 1 scout for their opinions. Was that after you turned in your ballot, a kind of effort to seek confirmation, or was it prior to handing in the ballot as a way to distinguish between Lincecum and Wainwright?
So I don't have any problems at all with how Will or Keith voted.
Lincecum is still getting better each year
Each system seems to yield a different NL ranking and not a clear consensus on even a top one or top three in the NL.
Some systems have them close, others have gaps between the first and second rankings.
Nor do we really even have a consensus on the most accurate fielding or offensive metrics with wOBA VORPing around WARP.
Do we really know anything or just making the best guesses off of imperfect models... and in a similar vein, can the "mainstream" be faulted for using their own measuring sticks?
Welcome to science; that's all it ever is. There isn't any absolute knowledge that isn't trivial.
And yes, using the best-available (i.e. least-falsified, most consistent with the data, most predictive so far) models is qualitatively different from the way the mainstream evaluate things. The biggest difference is that the models don't care who was the most heartwarming story, who's a total jackass, or how excited the fans got at any particular point in time.
I like to use OPS to evaluate a pitcher, must as it is useful with a hitter. There is a significant correlation between OPS and runs scored, to the point where a simple back-of-the-envelope formula such as 450 + (2000)(OPS - .600)gives a reasonable estimate of the runs a team scores in a 162-game season. Thus, an increase of one OPS point is worth about two runs to a team over the course of its season. Five more OPS points is worth about 10 runs, or one win.
Lincecum not only has the lowest batting average against (.206 to Clayton Kershaw's .200), he also had the second-lowest OPS (.271 to Dan Haren's .260), the second-lowest SLG (.290 to Kershaw's .282), and the lowest OPS (.561).
Carpenter's OPS was close at .581, but Wainwright's .646 OPS wasn't even close. Yes, Wainwright had two more starts than Lincecum, but he completed only one of them compared to Lincecum's four and Carpenter's three. Lincecum had two shutouts, Carpenter one and Wainwright none. Lincecum tied for the league lead in both categories.
Even though Wainwright had two more starts than Lincecum, he pitched only 7.2 more innings. And in those few extra innings he yielded 48 more hits and 7 more home runs.
Wainwright harder to hit than Lincecum? If so, Adam had HORRIBLE luck.
According to ESPN, Wainwright received nearly a run and a quarter more run support than either Lincecum or Carpenter. Why WOULDN'T he have won more games?
Aside from wins, it is hard say that Wainwright -- despite his allegedly being harder to hit -- came CLOSE to Lincecum, or to Carpenter (with the exception of a big innings gap over Chris).
Lincecum also had a higher quality start percentage than either of the other two, although Carpenter was very close (79% to Lincecum's 81%) and Wainwright wasn't too far behind at 74%. So it appears that even though he pitched 13 games in which he yieldeed one run or less -- winning only 10 of them -- Lincecum was the most consistent of the three.
And of course Tim showed how hard he was to hit by blowing both Cardinals away in the strikeout department. Despite a .297 BABIP to Carpenter's .272, Lincecum beat Chris by about 0.6 hits per nine innings. And with Wainwright coming in at a .309 BABIP, Lincecum blew Adam away in the hits yielded department.
That balls weren't hit very hard against Lincecum is demonstrated by the low SLG he yielded. He had the third-lowest home run percentage (although Carpenter was slightly ahead of him).
Now, let me give full credit to Will for recognizing Lincecum's full potential before anyone else did so publicly. Before Lincecum had thrown a single pitch above Single A ball, Will picked Tim as his franchise pitcher for the next 10 years -- and indeed in the first three, Tim has been the best starting pitcher, with Roy Halladay likely being second.
And Will's opinion of Tim this season may have been slightly skewed (although I know Will tries hard for objectivity, which is why I say MAY have been skewed) by seeing Tim in one of his rare bad starts.
Lincecum had temporarily lost the best pitch in baseball according to Fan Graphs, his change up, had lousy control and generally pitched awful that night. I suspect he had Will squirming in his seat.
But from that game on, Tim yielded only 5.55 hits per nine in his seven remaining starts. In his next start after Cincinnati he began throwing more curves and sliders, and a couple of starts later he had the change up back.
Yes, despite what the batter who hadn't faced Lincecum said, Tim isn't a two-pitch pitcher. He had his fastball, curve and change up when he arrived in San Francisco on Willie Mays' 76th birthday, and the following winter he further honed the change up and added his slider.
According to Fan Graphs, last season 21% of Tim's pitches were change ups, 18% were curves and 5% were sliders. That doesn't sound like a two-pitch pitcher to me.
I suspect Will has good respect for the hitter he asked about Lincecum, but I think the guy gave him some bad goods. He said Lincecum was a two-pitch pitcher, which just isn't true at all. Lincecum's change up is considered by Fan Graphs to be the best pitch in the game, and when he came up, his curve was considered to be his toughest pitch to hit. His slider is in that same ballpark, although he doesn't throw it as often. He threw it more late in the season.
And for the hitter to say Lincecum was more hittable than Wainwright certainly flies in the face of Tim's clear advantage in strikeouts, hit rate and home run rate.
It is possible that when the hitter faces Lincecum next season, he will change his opinion. And if batters were truly figuring out that Tim was a two-pitch pitcher (which he wasn't) late in the season, why did Tim's hit rate keep dropping?
The facts don't seem to support the hitters' opinion, no matter how respected he may be.
A vote for Lincecum was great IMO. A vote for Carpenter wsa good. A vote for Wainwright came either from misinformation or an over-reliance on wins.
I believe that Wainwright finished only third in the voting despite having the most first-place wins than either Lincecum or Carpenter illustrates this fact.
It's as easy as one, two three.
Wainwright had one more first-place vote than Lincecum.
He had two more votes total than Carpenter.
Yet the spot he finished in was indeed three.
Rightfully so IMO.
Wainwright was the SNLVAR leader. You may choose to ignore that, or give it less weight than other measures, but it's simply not true that the only case for Wainwright is a stupid one based on W's.
For the record, I think Lincecum was the right choice, but I don't think Will made any egregious errors here. He made a concerted effort to get as much information as possible from as many angles as he could on short notice and weighed it as he thought appropriate. If only all the voters put that much effort into it and shared their process with the public.
What an idiot.
Of course, that won't stop talk radio chuckleheads from burying the "Stat heads" for their decisions.
I can't think of another reason for you to ignore the many points made in the above discussion. To conclude that Will is "being different for the sake of being different" while completely bypassing his clear statement that Carpenter's DL time was a factor, is disingenuous. To skip Haren's better WARP3 and K/9 scores so you can call Will's judgement "embarassing" smacks of willful ignorance.
And to ignore the acknowledgement that there's not a lot to separate the top few pitchers doesn't reflect well on you.
Saying it was absurd might be, though. :)
1) People talk about me "leaving Carpenter off." There's only three slots, so I left a lot of people off.
2) Carpenter would have been my fourth, by a slim margin to Haren. Is calling someone the fourth best pitcher in baseball really that much of an insult?
Anyway, thanks Will and Christina for having the courage of your conviction and explaining your official votes to your readers. I'm sure you expected us to dissect your processes. Thanks again.
This was the the theory of the red sox signing Smoltz.
OTOH, if they lack the flexibility or have too many pitchers who are unreliably available, it can be a disaster.
That said, I would not have voted for Carpenter either. I find it interesting that Joe Sheehan has such a vehement statement in defense of Carpenter, when two of the primary BP pitching value stats, WARP3 and SNVAR, have Carpenter 5th and tied for 3rd respectively, which makes as a marginal choice by the BP stats.
it could be that i'm reading this incorrectly, and the reason that Carroll and Law had their ballot contents included in the original AP story is because both were very quick to write articles including those contents, which the AP then picked up. just curious.
You make a good point that the OPS of the batters Wainwright faced was nine points higher than that of the batters Lincecum faced. But Wainwright's OPS was 85 points higher than Lincecum's.
Adam's OPS was almost 10 times higher than the difference in the OPS of the batters he faced and those that Tim faced.
Yes, that nine point difference adds up over 905 plate appearances. So does the 85 point difference.
Wainwright got 12 first-place votes, so obviously at least a dozen people think he has something special going for him. But aside from wins I see the evidence tilting heavily in the favor of Lincecum and Carpenter.
By the way, I think Haren is underrated and at one time during the season I thought he was the best pitcher in the league. But unfortuately for Dan, he really fell apart toward the end.
He posted ERA's of 4.95 and 4.79 in August and September. Dan was great in April, June and July. He was good in May. But the last two months of the season he wasn't very good.
I thought overall he was very good. But I didn't think he was close to Lincecum or Carpenter.
Dan gave up 27 homers, for crying out loud. That is 10 more than Lincecum and Carpenter combined. It is within seven of "the big three."
At the end of July I might have voted for Dan to win the Cy. By the end of September he seemed to me to have pitched himself out of the top three or four.
I have a ton of respect for Will, and we (and certainly I) have focused on disagreeing with his decision, whereas most of all we should be thanking him for sharing his rationale and the work he put into it.
A point of interest here. Neither of the Bay Area writers placed Lincecum first on their ballots. One of them had Carpenter first and Lincecum second, and the other one had either Carpenter or Wainwright first and Lincecum second.
When the season ended, I thought Carpenter would win and might well have voted for him myself. But the more deeply I dug, the more I came around to Lincecum, who aside from walks was at or near the top in just about any type of stat one could come up with.
Don't know how much of it was because he pitched better or how much of it was due to luck, but Carpenter benefited from a very low .272 BABIP. That is 31 points lower than his career average and 12 points lower even than his Cy Young year of 2005. Even so, his batting average against was 15 points higher than Lincecum's with a .297 BABIP.
Still, as I mentioned before, had Carpenter won, I wouldn't have been upset. The more I think about it, Carpenter's ERA was nearly a full run lower than Dan Haren's and he gave up only a bit more than a quarter as many home runs.
But almost everyone has a different way of looking at things. I respect Will's about as much as anyone's, even if I don't agree with him in this particular case.
But, clearly the hitters are still being deceived (or it has nothing to do with deception and just has to do with stuff) as the K rate and other stats in this thread demonstrate. Given the Cy Young is a reward for past performance and not for future performance, the player's comments about deception and underlying hittability are as irrelevant as if I said I was leaving Carpenter off my ballot because he's so much older than the other players and isn't likely to maintain this performance in the future.
In other words, Will should have ignored this player's comment when deciding his vote, but maybe factored it in when projecting future performance.
Overall in his career, his home ERA is 2.84, while away it is 2.97. That's actually a rather small home/away split. His homers went from 14/9 prior to 2009 to 14/19 now.
Interestingly enough, Tim's WHIP is 1.151 both at home and away.
Tim has a lot of splits that are eerily similar. For instance, his OPS with RISP, with no one on and with runners on are .602, .609 and .606 respectively.\
His OPS with 2 outs, RISP, with the game tied, one run, two runs and three runs are .594, .596, .591,.584, .601 and .602.
His OPS at home is .608; away it is .608. His first-half OPS is .615. His second-half OPS is .600. His OPS against right-handed hitters is .599; against lefty hitters it is .617.
Will was quite diligent in gathering information for his vote, but let's look at the factors that tipped the scale toward Wainwright for him.
He liked Wainwright's two more starts than Lincecum. Certainly those have value, yet Lincecum pitched longer per start and thus finished with only 7.2 innings fewer than Wainwright. Wainwright had 7.2 more innings, 48 more hits and 7 more homers. Hmm.
He checked with three players and a scout. Will didn't say what the other two players and the one scout said, but he cited one player who said Lincecum's stuff was more hittable. When Will check back, the player said the difference was very slim. Virtally all the objective evidence indicates that Lincecum is much harder to hit.
If Wainwright truly was harder to hit, why did he give up so many more hits and homers despite pitching just 7.2 more innings than Lincecum? If he was harder to hit, why did he yield a .244 batting average compared to Lincecum's .206? If was harder to hit, why did he give up a .349 slugging percentage compared to Lincecum's .290?
If he was harder to hit, why did he not only give up close to two more hits per nine innings than Lincecum and 1.43 bases per hit compared to 1.40?
Adam Wainwright had better control than Tim Lincecum. He enticed more batters to ground into double plays. He held runners on base better. What else did he do better -- or even as well -- as Lincecum?
So? Not all metrics are equally relevant. If we throw in shoe size, IQ, number of wisdom teeth, and hair length, do they favor Carpenter or Wainwright? Who cares? The point of advanced metrics is to get us beyond having to decide by eyeball whether more WHIP here is more important than fewer innings there.
"When one looks at all the metrics, it is apparent that aside from won/loss stats, Lincecum leads, with Carpenter not too far off in second, and Wainwright a distant third."
As I already pointed out to you above, Wainwright was the SNLVAR leader. Until you address why that isn't a good argument for saying Wainwright deserves the Cy Young, you're doing the logical equivalent of putting your fingers in your ears and saying "La la la I can't hear you."