Could a Pitcher Win the NL MVP?
What would it take for a pitcher to win the National League MVP award?
Really there are two questions here: What would it take for voters to vote a starter into the award? And what would it take for a pitcher to be worth more than a position player?
First, a bit of history. Twenty-one pitchers have won the award since 1911, meaning it happens about once every 10 years. The last time it happened for a starter was 1986, when Roger Clemens won the award with a 24-4 record and 238 strike outs. Those benchmarks won’t be hit this year, but is it possible that some of the conditions will be met? Could Roy Halladay or Clayton Kershaw find themselves with two pieces of mega-hardware after the season?
Back in 1986, Clemens finished with 7.9 WAR. Among his aforementioned achievements that season, he also had a 2.48 ERA (169 ERA+) and a 0.969 WHIP. The second-place finisher, Don Mattingly, had 6.9 WAR. Even if that was enough to get Mattingly in the team picture — and his .352 batting average, 31 home runs and 113 RBI were impressive — there was one major flaw on his resume: He played for the second-place Yankees. Third-place Jim Rice managed 5.9 WAR and though he had a .324 batting average, he only hit 20 home runs. Jesse Barfield — fifth in the voting that season — had twice as many.
It’s hard to make one instance a road map, but starters have won so rarely in the modern era that one season 25 years ago gains importance, in retrospect. If there is a prescription to this, it’s clear. The commonly accepted best position player played for a team that didn’t sniff the postseason but the pitcher had a dominant year for a contender.
And that could certainly describe this year. But is there a National League pitcher this season who could be worth more than the best position player?
For the purposes of discussion, we’ll pretend that there’s a consensus that Roy Halladay is slightly ahead of Clayton Kershaw in the National League pitching pecking order.
Halladay has affected as many plate appearances as the busiest position player. He’ll face more than 900 batters this year — while someone like Matt Kemp will come to the plate about 650 times. Even if you add in Kemp’s 280+ chances in the field and give no credit for Halladay’s glove, it’s clear that a pitcher can affect enough game changes that he’d be qualified to enter the MVP race.
The Phillies’ ace has a 2.18 FIP, a figure that’s easily the best in the league. He’s saved 66 runs more than a replacement pitcher. Remarkably, because his FIP is 1.75 runs better than average, he’s probably saved around 45 runs more than an average pitcher. Even excellent pitchers are no match in a head-to-head competition with Halladay. He has also saved .24 context-neutral runs per nine innings over Kershaw.
Kemp obviously has had a great year. Even without giving him credit for anything beyond his work at the dish, his bat alone puts him in the conversation with Halladay. He’s added 50.5 runs of raw value by showing his best walk rate, upgrading his power and improving his strikeout rate.
“But,” you might say, “Kemp plays every day.” That’s reflected within the WAR framework. Because Kemp plays every day, and therefore replaces the replacement player, he gets replacement runs added to his WAR total. The replacement-player level is 20 runs per 600 plate appearances, so Kemp’s durability has added 21.6 runs to his work this year. And by improving his efficiency on the basepaths in 2011 — he’s been successful 80% of the time this year compared with 56% of the time last year — Kemp has added another couple of runs to his total.
If only Kemp were a meister with the glove, this might not even be a discussion. But in a full season, Kemp has only once been a positive in center field. This year he turned in one of his better defensive performances and was worth 4.6 runs less than the average center fielder. That’s not terrible, but it’s not great. The positional adjustment for center field is +2.5 runs, and that shaves much of his middling defense away. He still doesn’t have great range — the range component of his UZR put him as the third-worst center fielder in baseball — but he doesn’t make a ton of errors and he has a decent arm.
Add it all up and Kemp has a 7.6 WAR. Compare that total to Halladay’s 8.0 WAR and it seems like it could be possible to pick the pitcher. Kemp plays for a team that won’t make the postseason while Halladay not only has an impeccable record for a championship contender, he’s the staff ace among a staff packed with aces.
On the other hand, it’s asking a lot. Halladay only has 18 wins, and with Kershaw’s chance at 21, Halladay doesn’t own the same advantage over his pitching competitors that Clemens had in 1986. That year, the Rocket beat out Milwaukee’s Teddy Higuera for the Cy Young, and the Brewer went 20-11 with 31 fewer strikeouts.
So while it’s possible that a pitcher might actually be worth more than the best position player in baseball this year, it’s not probable that the pitcher takes home two pieces of hardware. After all, he has an award designed to reward his contributions, right?
With a phone full of pictures of pitchers' fingers, strange beers, and his two toddler sons, Eno Sarris can be found at the ballpark or a brewery most days. Read him here, writing about the A's or Giants at The Athletic, or about beer at October. Follow him on Twitter @enosarris if you can handle the sandwiches and inanity.
Halladay wins the MVP; Kershaw takes the CYA.
Now, everybody’s happy … or no one is.
Exactly. We know this won’t happen and I’ll be shocked if Kershaw doesn’t win the Cy. Actually, I’m not even convinced that he shouldn’t win the Cy.
Someone explain this to me please. Pitching WAR is calculated largely using the principles of DIPS theory, right? So is it not a better predictive measure than a measure of results? And is it not fair to say that end-of-season awards should be handed down to the guys with the best results (even if they’ve been “lucky” or whatever?
Pitching WAR has always seems suspect to me when it comes to crediting a pitcher for results (and even to some extent when it comes measuring his performance). I honestly don’t know how to form an opinion of this.
Well, when you consider that Kershaw’s BABIP might be under his control more than most pitchers, normalizing it because of ‘defense’ does seem odd.
By the way, Kershaw and Halladay are now almost dead even in rWAR– and if you take out the leverage component (why, why would you have leverage in there?!), they’d actually be tied.
FIP, and therefore, fWAR for pitchers is still results based. It’s simply the results of ONLY the pitcher, not his defense. It is a good predictor as well, but it is absolutely results based (Ks, BBs, HRs, IP). It ignores balls in play since they are affected by defense and luck.
Good point. You’re right, they’re results-based, but to me those results are secondary (although admittedly strongly correlated) to the ultimate result of preventing runs.
I agree but I don’t think there is a good way isolate a pitcher’s contribution to preventing runs. As it stands, using runs allowed is a combination of factors attributing to runs prevented. Even with a blanket defensive adjustment (as rWAR uses) it’s still very murky, and not truly results based anymore. I think FIP is the best available tool right now, until there is some tool that can more accurately discern the contributions of the pitcher from that of his defense and luck.
Very well put (and for some reason more effective on me than the massive articles concerning WAR posted lately — maybe just because I’m actively involved in this), but I’m having an internal struggle over whether to appreciate a pitcher with an ERA and identical FIP of say 2.5 versus a pitcher with an ERA of 2.0 and a FIP of 3.0.
I tend to prefer the second guy, even though, as you say, the underlying reasons for those numbers are very murky.
I don’t like fWar because it is DIPS based. No one looks at batted ball profiles in evaluating hitters’ WAR. Every single is a “line drive in the books” so to speak, even if it is a seeing eye groundball through a porous defense. I see no reason why pitchers, for retroactive WAR calculations, should not get the credit for inducing an out to their defense.
To me, this ties in with the Plate Appearances argument posited above. Considering that fielding plays a part in a significant percentage of a pitcher’s Batters Faced, should we be inclined to only count BB, K, and HR in a pitcher’s BF count? Or maybe a percentage of the Ball In Play outcomes?
I think a pitcher MVP is much more likely in the AL.
In the NL, Halladay doesn’t even have the best ERA on his team (Lee does), nor is he the Cy Young frontrunner (Kershaw leads in ERA, W-L, and K’s). I think Braun is the NL frontrunner.
Meanwhile, in the AL, the would-be consensus (Bautista) is ineligible because his teammates suck. And saberists might love Ellsbury, but a lot of his value is from defense/position (i.e., things that go unrecognized by voters), and he’s competing against two teammates for the same award (Pedroia and Gonzo). Verlander has the Cy Young wrapped up with a stat line that looks similar to Clemens’ 1986, and is the only great player on a playoff team. Oh, and he could win 25 games.
You take that back!
Yeah!
A new article could be what would it look like for a pitcher to be MVP without being CY as generally the pitchers winning it (Eck, Fingers) also got the Cy. Short of being Babe Ruth and playing a lot besides pitching, that is.
Maybe something like 1965 NL with no Willie Mays, where Koufax gets the Cy and teammate Drysdale’s 7 HR and .300 BA (team leading 140 OPS+) in a pitcher’s park in a pitcher’s era influences enough MVP votes?
I have a huge issue with slamming Matt Kemp’s defense. For one, baseball reference actually has him as a positive defensive WAR. I know it’s a different valuation and perhaps not as accurate, however, shouldn’t there be some sort of convergence of the two subjective stats? According to Fangraphs, Jacoby Ellsbury is among the best defensive outfielders in baseball, but he throws water balloons from the outfield…whereas with Kemp (and I know it’s not a full reflection of throwing ability) is second in NL outfield assists…he has been pretty good this year.
Additionally, Kemp is currently a relatively hot week and a couple slumps away from winning the first NL triple crown since 1937. I don’t remember anyone recently entering the last week of the season having a chance at it, and yet no one mentions it. Obviously, it’s three anti-SABR stats, but even most of the SABR stats agree he has been the most dominant offensive player in the NL.
This is a great post. Very well thought out.
additionally, cargo is currently a hot week and couple slumps away from wining the first NL triple crown since 1937.
oh, we’re not talking about 2010?
Maybe the fact Kemp is 1 HR back, he’s first in RBIs, and about 8-9 points in batting average behind first makes it a relevant statement.
again, all things that can be said about cargo/votto/pujols in 2010
Nice , but this is about 2011.
Do you have a point, or just saying something that doesn’t matter? It would make more sense to say the same could be said about Braun. I’m not sure of the point you’re trying to make since the guys you mentioned finished 1-2-3 in the voting. Maybe you’re just being captain obvious, so whatever suits you.
Speaking of triple crowns, both Kershaw and Verlander lead their leagues in ERA/Wins/K’s. It’s a bit more common with pitchers, but certainly noteworthy.
And Verlander has a good shot at the MLB-wide triple crown too. He’s clinched the win column, and is leading in K’s, but is 0.02 back in ERA.
An AL pitcher taking the full league triple crown would be amazing. Yes the three categories are somewhat arbitrary but still rare and still incredible considering league differences.
Now that you mention it, that there is real chance of 3 triple crown winners – Verlander and Kershaw both have it right now (Kershaw is tied in NL wins currently at 20), and Kemp is making a strong end-of-season push. There has never been as many as 3, and the last time there were 2 was 1933 when Jimmie Fox took the AL crown and Chuck Klein the NL.
http://www.baseball-reference.com/awards/triple_crowns.shtml
Kemp has an excellent arm, but brings almost nothing else to the table as a defensive player. Iffy range for a centerfielder and constantly mis-reads balls. His defensive metrics are a pretty accurate assessment of his defensive ability.
Then how do you reconcile that with the fact that B-R WAR has him as a 1 WAR defensive player? Is their metric completely wrong?
Even as an admittedly biased Dodgers fan, I admit that Kemp doesn’t seem to be a plus fielder. For years he hasn’t gotten good first steps and sometimes takes bad routes to the ball. That being said, he has an absolute cannon of an arm (and it’s accurate too).
Though I’m no scout, I must say that his defense has seemed better to me this year, and certainly MUCH better than last year (I buy his -25 UZR in 2010). The mental mistakes that have plagued him in past years seem less frequent this year and his arm and speed are a good as ever.
The large divide between UZR and BBRef’s Total Zone on his defense is intriguing though. I wonder if anyone has Kemp’s Fielding Bible DRS numbers for this year.
I’d love it if he won the MVP. I think I speak for all Dodgers fans when I say that we need SOMETHING to cheer about.
Could a pitcher? No. National writers picked their pitcher to champion weeks ago and are sticking by him, no matter what’s happened since.
And who is that?
Yeah, I thought I heard on ESPN that Kershaw clinched the CYA with his performance against SFG.
How many points back is Halladay?
“Add it all up and Kemp has a 7.6 WAR.”
Not a huge deal, may have been written yesterday and just posted today, Kemp is at 7.8 fWAR now. He’s over 9.0 (likely 9.2 or so) baseball reference WAR as well, though it’s clear what is preferred here.
One argument I think holds valid weight is that yes Halladay may be the best pitcher in baseball this year (I think Kershaw is right there, I know others don’t agree though), but he has a great offense behind him for run support, whereas Kemp (for all but maybe August and September) has had a horrible lineup all around him this year. It’s also another reason why I’d place Braun and Fielder below Kemp, also because they just arent as good this year.
I definitely have no problem having a pitcher win MVP, but I think it should only be if he clearly dominates more than any other player in the league, which he has not over Kemp or even Kershaw or Lee at this point.
http://i.imgur.com/UJnkl.png
Link is to the top baseball-reference players with WAR in the NL. Kemp has 9.1 and is 1st, Kershaw has 7.4 and is second, Lee and Braun have 7.2 at tied for 3rd, Doc has 6.8 for 5th.
Kershaw and Lee are higher b/c their offense contributes positively to their rWARs, unlike Halladay’s.
The shouldn’t matter for Cy Young, but when you’re talking about MVP, why shouldn’t a pitcher’s offensive contributions count? It may seem odd, but it makes sense when you consider what the awards are for. I would have Halladay ahead of Lee in the Cy Young, but not the MVP.
He Dodgers might have had a bad offense, but it would make more sense to actually look at how many runs were scored for Kershaw and Halladay in their starts. Though run support only changes the outcome of wins anyways.
Ryan Braun.
Matt Kemp walks, hits a 3 run home run, hits a single, steals second and laughs 😉
2009:
Grenke 9.4 WAR
Mauer 7.9
mauer won 27 of 28 first place votes
tough to have a pitcher MVP come from a last place team
“The replacement-player level is 20 runs per 600 plate appearances, so Kemp’s durability has added 21.6 runs to his work this year.” Somewhat tangential here to the main point, but could someone explain to me the rationale of including replacement player runs included in a player’s total WAR? It seems counter intuitive to me, since I would think that the runs ABOVE replacement player should logically discount the runs that would hypothetically be produced in his absence and just leave the delta, but I’m interested to hear the explanation. Thanks in advance.
Your intuition is basically correct, you’re just thinking about replacement runs a little backwards. Run values for players are calculated against league average. The replacement runs are added to that because replacement players are 20 runs worse than average, so what you’re left with is runs above replacement player.
Justin already explained it, but I’ll explain it a little differently.
All the value stats on fangraphs (batting, baserunning, fielding, positional) are in terms of runs above/below league average. The replacement runs are added in at the end, based solely on playing time. Without them, it would be “runs above average” rather than “runs above replacement.”
You could accomplish the same result by doing all the sub-categories according to runs-above-replacement rather than runs-above-average. But it’d be more complicated and make less intuitive sense. How do you measure positional runs above replacement? Isn’t it nice to be able to easily explain to someone that so-and-so is x runs above/below average in batting/defense? You don’t need to understand ‘replacement level’ to use those stats. And doing things this way it also allows us to easily calculate wins-above-average (a stat I prefer when discussing the Hall of Fame).
Thank you both.
Compared to Kemp, Ryan Braun has the stupid traditional numbers that determine MVP winners (AVG, RBI, HR, SB) and also WAR. In addition to comparing favorably in terms of statistics, Braun plays for a playoff bound team. Because Kemp’s numbers are only fractionally better than Braun’s, the ‘tie’ goes to the player on the contending team.
This is another thing. If Kershaw is going to get crucified or pitching against the NL West, and getting most of his games at Dodger Stadium, Petco and the phone booth, then you can’t also overlook that Kemp is playing most his games in those same parks. I hate to say it, but if Kemp played half his games at the band box that is Miller Park, he would be running away with this award.
First off, calling Miller Park a “bandbox” is a bit of hyperbole, don’t ya think? It has a 102 multi-year park factor on B-Ref, while Dodger Stadium is at 97. Yeah, 5% between them is a pretty big difference, but we have a stat (wRC+) that accounts for that, and Braun>Kemp, by 8%.
I’m not saying I’d be upset if Kemp won, as I personally see them as very close candidates. Braun has been superior to Kemp offensively this season, and they’ve probably been similar defensively. Really, the argument boils down to two things:
1. How do you value the positional difference?
2. Does it matter that Braun is on a winning team, and Kemp is not? (personally, I only use it as a tie-breaking measure)
Band box is an exaggeration. Miller Park has the 10th highest park factor for 2011, so above average, but not a lot.
Also, when he’s not playing in Petco, AT&T, and Dodger stadium, he’s playing in Coors & Chase, 2nd and 4th in park factors for 2011. The average NL West park factor is higher than the NL Central park factor.
Yes, Kemp has it harder playing half his games at Dodger Stadium, but it’s not quite to the level you seem to think.
Agreed. I think Braun wins it.
Side note: I’m having trouble identifying which comments are about what will happen vs. what should happen.
Excellent reasoning in this article.
Why is the article debating who will be second place in MVP voting behind the Hebrew Hammer?
Is there any logic behind Braun winning the MVP other than “He plays for a better team”? I sure haven’t seen it.
Not a ton (he does have better hitting rate stats, though). But MVP voters follow that “logic,” so that’s all that really matters.
because the difference in his and kemp’s WAR essentially boils down the runs kemp receives from playing CF vs braun playing LF
Or the difference between playing at Dodger Stadium and Miller Park.
wRC+ accounts for park
Stating that “Halladay has affected as many plate appearances as the busiest position player” is a specious argument. The affect that batters and pitchers have on an at bat are quite different and not comparable. I can’t say for certain which has more influence over the outcome, but I suspect it may be the batter, as there seems to a greater difference between the run production of the best / worst batters than between the best / worst pitchers. This season, setting minimum PA to 400, the highest RC/27 is Bautista at 9.88, the lowest is Orlando Cabrera at 2.34. For ERA (min 140 innings) the spread is Verland 2.29 / Lackey 6.49.
It’s crazy to think that a few years ago guys like Morneau and Howard were winning MVPs and now there is a legitimate shot that the NL MVP and Cy Young come from the same non-playoff team. As an aside, if Kemp and Kershaw win that would have to be the first time guys from the same team won, and that the team did not make the playoffs.
Wow, that really speaks to the lack of depth the Dodgers have.
May I introduce you to 5th starter Dana Eveland (4 starts, albeit) and starting third baseman Aaron Miles (over 400 at bats).
Aaron Miles has also started a game batting 5th in the lineup 17 times. Triple slash in those PA’s: .242/.309/.274.
Just for kicks, he has also batted 3rd four times and posted a .989 OPS.
If an NL pitcher gets 1st-place MVP votes, it’s not going to be from voters who are considering WAR. It’s going to be from voters who are considering the nebulous touchy-feely meaning of “Valuable” and looking for a pitcher whose story fits that definition.
The Phillies would be great without Halladay and he might not even be their ace, so he doesn’t fit.
The Dodgers won’t make the post-season, so he won’t fit.
I’ll bet money that Ian Kennedy gets more MVP votes than Halladay and Kershaw combined.
NL MVP = Kemp, NL Cy Young = Kershaw
AL MVP = Ellsbury, AL Cy Young = Verlander
I’ve been right six years in a row on the picks.
Let’s hope it’s 7 🙂
Ellsbury will not win the AL MVP. Mainstream voters are more likely to vote Granderson.
Which would also mean that you chose Bartolo Colon as the Cy Young in 2005.
Which is not necessarily a bad thing. If Hurtlocker predicted that Colon WOULD win the CY in 2005, on the strength of his 20-win season, that was a good prediction. If, of course, he said that Colon SHOULD win the award because of his 20 wins, he deserves to be slapped around with a large trout.
You predicted Morneau in 2006?
The problem with this article is that if you’re going to going to consider a pitcher for NL MVP you should probably include his contributions with the bat and fielding. His bat has been a net minus, even for a pitcher. UZR isn’t kept track of for pitchers, but he’s been a -5 in DRS. So if you factor in these things, he’s probably been less valuable than Kemp.
In fact, Kershaw’s contributions with the bat and as a defender very likely makes him as valuable as Halladay overall.
Lee also gets a bump, at least from his bat.
The title of this article uses the word “should” when it should be using “could.” Sure, a good argument is made here that Halladay “should” win the MVP. But an argument about who will win should focus a lot more on IP, Ks, and ERA, which is primarily what the voters will use.
Basically, Halladay is in a dead heat with Kershaw over the Cy, and is not even really in the MVP discussion.
First line of the article is:
‘What would it take for a pitcher to win the National League MVP award?’
Next couple of paragraphs then proceed to talk about a pitcher who won the AL MVP?
So?
If Kemp was better in center, it would probably help Kershaw substantially as well. Kemp would see his numbers improve, but if fewer of the hits for Kershaw fell in, he would have better numbers as well. This is particularly true for ERA, which certainly factors into the awards (even if fip and such stays the same).
A pitcher winning an M.V,P, AWARD is not strange if he is used as a member of a four man rotation and has worked in relief during the season—however 1 out every 5 days doesnt really make me want it to happen .
Look at the players surrounding Matt Kemp…It’s laughable. James Loney, Aaron Miles, Jamey Carroll, Juan Rivera, Jerry Sands, Juan Uribe etc. If he had Prince Fielder hitting behind him (which we may very well see next year), his numbers would be Godly. The numbers he has put up with the lineup he’s surrounded with is far greater than Braun’s. Add in his ~6 walk-offs and his 35/35 hr/steal, he has been the most complete/valuable player in the league.
Matt Kemp says no with an emphatic thud 🙂