I'll answer your question with an article praising Hunter Pence because it's 4 PM on a Friday and I can't type anymore
The Stealth MVP Candidacy of Hunter Pence.Warning: Spoiler! (Click to show)
I’ll say it up front; the headline you just read is a little bit of a trick. This post is about Hunter Pence perhaps being more valuable this year than most of us realized, but because I have an NL MVP ballot this year, I can’t get too deep into my personal opinion of where Pence belongs on the ballot. This isn’t me explaining why Pence should rank at some particular position in the final MVP tally; it’s me using Hunter Pence to talk about one primary way where I think the stats that are normally used to determine MVPs might miss some real value, and that goes for WAR as well.
By either traditional or even normal advanced metrics, Pence doesn’t have much of a case. He’s hitting .290 with 17 HRs and 72 RBIs as a corner outfielder, which isn’t going to fly in a season where Giancarlo Stanton hit twice as many bombs and has driven in an extra 30 runs. And it’s not like the more sabermetric numbers help his case that much either; Pence has a .345 OBP and .463 SLG, so he ranks 22nd in the NL in OPS, one spot behind Seth Smith. Yeah.
Even adjusting for park factors only gets Pence’s wRC+ up to a tie for 18th in the NL, in the same range as Michael Morse, Neil Walker, and Starling Marte. You probably won’t see too many people arguing for those guys as MVP candidates, and rightfully so, and these are Pence’s peers even by our most often cited hitting metric. So, why am I writing about Hunter Pence and the MVP?
Because he lets me talk about context batting runs, or as I called it last year, “Cobra”. I really want to go on MLB Network sometime and refer to a player’s Cobra. It’s a weird life goal, but then again, I have a pretty weird life.
Anyway, as a refresher, context batting runs are officially the result of the metric called RE24, which is a really neat stat with a really terrible name. Essentially, RE24 measures the results of a player’s performance on the difference in run expectancy from the beginning to the end of his at-bat, which means that it’s instead measuring the amount of actual runs created by a player’s performance rather than substituting in average run values of events, as linear weights metrics do. In other words, if a guy comes to the plate with the bases loaded and hits a home run, he’ll get 1.4 runs of credit in our linear weights models, but he’ll get somewhere between 1.8 and 3.4 runs of credit by RE24, depending on how many outs there were.
By including this context, RE24 better measures “what actually happened” with a player’s offensive performance, so that if he managed to distribute his hits more frequently with men on base, he’ll get more credit for those hits than if he was constantly stranding runners. This isn’t a repeatable skill, but then again, we don’t really want to strip luck out of MVP voting; we want to reward a player for the value of his performance, even if he can’t repeat that performance again. And hitting better in more important situations is more valuable than getting a lot of hits when there’s no one on base.
So what does any of this have to do with Hunter Pence?
Well, as you might have guessed, Pence is one of the poster boys for adding value by hitting exceptionally well with men on base this year. With the bases empty, he’s hit .269/.324/.465, good for a 127 wRC+. With men on base, the numbers are a little better; .324/.376/.458 with a 139 wRC+. But with runners in scoring in position, Pence has hit a ridiculous .377/.457/.547, which translates to a 184 wRC+, third best mark in the National League. Pence’s overall batting line isn’t that great, but when he’s presented with opportunities to drive in runs, he’s come through at a exceptionally high rate.
This shows up in his Cobra RE24. Below is a list of players batting runs by normal linear weights, and by RE24, with the base stealing portion of our baserunning metric added to batting runs because RE24 counts SB/CS for the player as well. These players have produced far more actual runs when context is taken into account than traditional linear weights will give them credit for, due to their distribution of hits.
Name Bat+wSB RE24 Difference
Jayson Werth 25.2 42.7 17.5
Christian Yelich 15.9 32.9 17.0
Melky Cabrera 16.8 32.8 16.0
Matt Holliday 21.1 36.0 14.9
Hunter Pence 23.6 38.4 14.8
Adam Eaton 7.3 21.4 14.1
Ryan Howard -6.5 7.1 13.6
Ian Desmond 4.9 17.8 12.9
Justin Morneau 9.9 22.4 12.5
Matt Carpenter 11.5 23.9 12.4
Pence isn’t at the very top — I considered using Jayson Werth as my foil for this post, since he’s actually #1 by this metric, but I’ll explain Pence’s other advantages in a minute — but you’ll note that his RE24 is 15 runs higher than the combination of his batting runs and stolen bases. 15 runs is a lot, especially in the current run environment, where league average runs-per-win is just over 9.1. Toss in Pence’s home park, which deflates run scoring even further, and we’re looking at nearly a two win difference between what Pence’s context-neutral stats will credit him for and the benefit the Giants have actually received from Pence’s offensive performance, once you factor in the distribution of when his hits occurred.
If the offensive component of WAR was RE24 instead of linear weights, Pence as an MVP candidate wouldn’t seem quite so insane. For fun, here are the top 10 hitters in the NL by WAR, with RE24 subbed in for batting runs and the stolen base part of Baserunning.
Name RE24WAR WAR Difference
Giancarlo Stanton 7.0 6.0 1.0
Anthony Rendon 6.9 6.0 0.9
Hunter Pence 6.7 5.1 1.6
Christian Yelich 6.5 4.7 1.8
Jason Heyward 6.5 5.2 1.3
Jonathan Lucroy 6.4 6.3 0.1
Buster Posey 6.1 5.5 0.6
Jayson Werth 5.9 4.0 1.9
Paul Goldschmidt 5.5 4.3 1.2
Jhonny Peralta 5.2 5.3 -0.1
If you think situational hitting should be included in MVP discussions — and writers clearly do, given their love of RBIs — then Hunter Pence has a legitimate case as a top five MVP candidate this year. He’s played in every single game the Giants have played this year, is second in the majors in plate appearances, and has been exceptionally good at things that both traditional metrics and the most common advanced metrics don’t take into account; non-steal baserunning and insane situational hitting.
If the goal is to measure a player’s contribution to a team’s wins in the past, I think a good case can be made for giving significant credit to players who take advantage of the opportunities they are given. And few hitters have done that as well as Pence. Also, Christian Yelich! If you want to know why the Marlins were able to contend without Jose Fernandez, he’s your answer.
Of course, including this kind of context but not score/inning context opens up a bit of a can of worms, and leads to questions about why not just use full Win Probability if we’re trying to give hitters credit for clutch hitting. RE24 is going to treat a bases loaded single in the first inning of a blowout the same as it treats a bottom of the ninth walk-off, which isn’t how our emotions value those two events. There’s a slippery slope argument to using some-but-not-all clutch performance, and I understand why it’s easier to just use context-neutral numbers and not get into the gray area of trying to decide which situational runs a hitter should get credit for.
But while using win probability goes a little too far for my tastes in determining a player’s contribution to a win — the walk-off in the ninth was only possible because of the runs scored earlier — I’m personally okay with giving hitters credit for distributing their hits into on base/RISP situations. RE24 isn’t giving them sole credit for the value of those runs like RBIs does, and it takes into account the opportunities they were given. Pence has been exceptionally good with the opportunities he’s been given. To me, it seems like that’s worth accounting for on an MVP ballot.
Where he’ll actually end up on mine, I don’t know. But I know that I think he’s being underrated in the discussion, and not just by BA, HR, and RBIs. For those who believe that WAR isn’t the be-all, end-all of MVP metrics, this is where I will most heartily agree with you.