Making an MVP out of Dallas Keuchel

If I were writing this piece five years ago, the entire premise of the article would seem contrived. Only two pitchers won MVP awards between 1986 and 2010, and no pitcher had won since 1992. Pedro Martinez led the AL by four wins in 1999 (although only 2.5 using RA9-WAR) and didn’t win the MVP. But thanks to Justin Verlander and Clayton Kershaw, voters have begun to acknowledge that pitchers are baseball players who are eligible for baseball player awards, so they’re worth considering in our MVP discussions.

Attempt, if you can, to remember April 2014. Entering last season, Dallas Keuchel had thrown just 239 major league innings over two years. He was entering his age-26 season, so he had youth on his side, but he had pitched to a 130 ERA- and 120 FIP-. He gave up home runs, didn’t strike batters out at an impressive rate, and allowed an average-ish number of free passes. Perhaps you may have seen some potential in the lefty who managed a 90 xFIP- in 2013, but the odds of him turning the corner and becoming an ace were long.

Yet here we are. The emergence of Keuchel as an ace isn’t a new story. He spun a 3.8 WAR season in 2014 (4.9 RA9-WAR) and while most people expected him to be respectable but not great in 2015, he responded by elevating his strikeout rate and leading the Astros into their real first pennant race in a decade.

The Rangers found a way to jump all over Keuchel on Wednesday, scoring nine runs on 11 hits and three home runs. However, even after that devastating outing, Keuchel owns the third highest FIP-based WAR (5.5) in the AL and the highest RA9-WAR (7.1). Despite this week’s clunker, Keuchel is right in the thick of the Cy Young race. Maybe he won’t win, but he’s absolutely a finalist.

There’s also the matter of the MVP award. At first, Keuchel doesn’t have the most obvious case. In fact, when Dave was asked about this prior to the nine run start, he said this:

My initial reaction was the same. Perhaps I’d quibble and say “getting mentioned” lacks a clear definition. Keuchel is almost certainly one of the 10 best players in the AL this year, so he should be in and around the debate even if you’re a person who doesn’t buy into his candidacy for the top spot.

But in an era in which Verlander and Kershaw have won MVP awards without having a runaway Pedro-in-1999 season, it seems necessary to give Keuchel’s candidacy a fair shake.

A few ground rules. First, I interpret the MVP award as the Best Player in the League Award. I would vote for a player whose team went 40-122 with the same gusto as a player whose team won 105. I generally don’t care to assign credit based on situational performance either. If you have a different value system, you and I are honoring different things. That’s okay, awards don’t really matter!

Donaldson is the target. By our measures, he’s been a 7.9 WAR player through Thursday. For good measure, his DRS is about two runs better than his UZR, so let’s just say he’s an eight-win player and call it even. Of course there is measurement error in WAR, but to the best of our understanding, Donaldson’s WAR total at this point lines up with (a) our perception of his performance and (b) the general performance of an MVP winner. In other words, to make a case for Keuchel as the MVP, we have to sell Keuchel as an eight-win player. Can we do that?

It’s actually surprisingly easy to make a case for Keuchel as an eight-win player. Keuchel has an RA9-WAR of 7.1. Keuchel has also recorded 12 DRS. Generally, 12 runs saved are worth about 1.3 wins, meaning you could wrap Keuchel’s defense up with his RA9 and he’s right in Donaldson’s kitchen.

That might not be an airtight argument, but it’s certainly close enough to merit further conversation. If you do the same basic math with Keuchel’s FIP-WAR, you end up with 6.8 WAR. It’s excellent, but it’s not really close enough to challenge Donaldson. But it’s not that far away either. A half-win more and it’s essentially too close to call.

If you’re using RA9 exclusively, Keuchel is in the game. If you’re using FIP exclusively, he’s probably not. Let’s try to dig a bit deeper.

Staying with his defense for a moment, Baseball Prospectus’ TRAA tells us that Keuchel prevents stolen base attempts — or, in other words, holds runners quite well. TRAA is a sophisticated model that tries to sort out a lot of confounding variables, but even if you don’t believe the model, the fact that base runners have stolen just four bases while he’s been on the mound this year backs up the more complex findings.

On top of that, his actual glovework rates out quite well. This is his second consecutive great year, according to DRS, leading to this description of his defense from ESPN’s midseason defensive All-Stars ($):

What makes Keuchel so good is his unique ability to get into good position after delivering a pitch to field balls to his left and right as well as slow dribblers straight at him. Playing to his strength as well as the Astros’ shifty defense, he has helped himself with the highest ground ball per fly ball rate in the majors, creating more fielding opportunities for himself than any other pitcher.

We probably don’t have a great grasp on pitcher defense, but Keuchel holds runners well, which should allow him to prevent more runs than his FIP suggests, and he helps prevent runs after the ball is put in play, meaning he’s adding some value on the non-pitching side of the ledger.

Given that the spread of Offense Runs Above Average (Off) for AL pitchers ranges from +1.4 to -2.3, it’s probably safe to ignore that part of the game for AL pitchers. If you want to get technical, however, Chris Sale and David Price are a couple of the pitchers with the lowest offensive value, so Keuchel could presumably buy a run or two back from them if needed.

Naturally, his pitching is going to get him into the conversation. Let’s call the defense a one-win boost, including his ability to hold runners. That seems like a fair assessment, but I recognize we’re working with imperfect estimates all around.

If you’re a fan of Baseball Prospectus’ Deserved Run Average, his DRA- (68) is right between his ERA- and FIP-, and BP pegs him as a 5.4 WARP player this year (this includes his base-running impact). That puts him on par with Price, Sonny Gray, and Chris Archer for the top spot in the AL, but it does leave him well shy of the Donaldson threshold. For the record, Baseball-Reference has him at 6.4 WAR without considering his defense.

So that’s a range of options. Keuchel is somewhere between a 5.3- and 7.1-win pitcher just looking at the statistics that try to measure his pitching. It’s safe to bump those numbers up thanks to his quality defense, but we need to decide how much of Keuchel’s run prevention (RA9) is really within his control. We already stipulated to his ability to hold runners, so that’s a point in his favor.

But on the other side, he throws to quality framers who deserve some credit for the quality of his strike zone this year. According to BP’s run values, he’s also pitched in a favorable set of stadiums and faced a slightly less impressive group of hitters than average. Separately, he’s also left nine men on base for his bullpen and none have scored, meaning his RA9 is a tad better that it would be if he yielded the mound to average relievers. These aren’t huge factors, but they do tip the scales away from Keuchel being the MVP.

There’s one big question sitting before us, and it’s the question that nags us repeatedly in analyzing pitchers. What role does the pitcher play in limiting damage on balls in play. FIP, which doesn’t help Keuchel’s case, treats every ball in play equally. On average, I’m fond of that assumption, but we also know that it’s not a flawless approach. There is some variation in pitcher ability when it comes to limiting quality contact.

And Keuchel leads qualified starters in Baseball Info Solutions’ Soft Hit Rate with 24% of his batted balls allowed falling into the Soft% bucket this year. On the flip side, he also has the lowest rate of hard-hit balls allowed. If you look at the incomplete, but useful, Statcast data, it supports the notion that Keuchel allows weak contact. So that’s a point in his favor.

It comes down to this: there is a certain list of things you have to believe in order to consider Keuchel to be worth roughly eight wins. Keep in mind, the Donaldson WAR value is also an estimate, so anything solidly above 7.0 WAR certainly makes it a conversation.

To believe Keuchel is a 7-8 WAR player, you have to believe he has pitched better than his FIP while also controlling for the fact that he’s had some good contextual factors like good catchers, defense, and slightly weaker opponents. You have to give him credit for limiting the running game, which seems fair, and for taking away runs with his glove, which also seems completely reasonable. The final piece centers on how much you believe Keuchel is able to limit the danger of balls in play against him.

You have to be willing to believe that Keuchel can save 0.3 or 0.4 runs per nine innings by limiting the damage on balls in play to believe in Keuchel as the MVP. I’m not sure we know if that’s a reasonable thing to believe. It strikes me as plausible, especially given that Keuchel seems to be one of the better weak-contact-inducers in the game, but I’m not sure we have strong evidence to defend the position as more than a hunch.

In other words, you can make a case for Keuchel as the MVP. It’s not an outlandish claim. If you believe pitchers have a lot of control over what happens behind them, Keuchel looks an awful lot like an eight-win player. If you think pitchers have very little control over what happens behind them, it’s challenging to make a case for the bearded lefty. The tricky part, of course, is that we think the truth is somewhere between the extremes and exactly where the truth falls makes the entire difference. Keuchel, I feel confident saying, is at least a 6.5 WAR player this year. But a 6.5 win player isn’t the MVP.

Yet there’s a completely rational case to be made that Keuchel is better than his FIP due to his runner thwarting and his contact quality suppression. The real issue is that we can’t put a good number on that last part. It’s something, but it is two runs or 12 runs? We can’t yet say.

All told, Keuchel is clearly a Cy Young contender and deserves placement on a good number of MVP ballots. Has he been the best player in the American League in 2015? I would bet against it, but I’m not extremely confident in that distinction. He’s close enough that a couple of good starts down the stretch could easily push him over the top.

Keuchel’s pitched very well according to every measure we have, but his defense gives him a chance and his ball-deadening tendencies leave us some room to debate. I expect Donaldson will win comfortably, but I’m not totally ready to dismiss Houston’s ace from consideration.