Consider this a freebie. Oh wait, what? They're all free? We don't make you pay for this? Well, that's dumb. Anyyyyway ...
Over the past month, we've been delving into the differences between everyday statistics and nuanced statistics, and this is the time of the year where the battle between the two really rages. It's also the time of year where both sides really blast the rhetoric to volume 11. What'd I like to do today is focus on how to properly use the nuanced statistics we've learned when evaluating award winners.
There's a great line from the show House in which an administrator grants that rules don't work 100% of the time, but he asserts, "While the rules only apply 95% of the time to 95% of people, everyone thinks they're part of the 5%." What's great about the discoveries that have been made in regard to baseball research and nuanced statistics is that they apply to so many players. They apply to so many players as a result of the great research that has been done, and that research was done using the general baseball population. That led to many of the theories we use today, and they work a vast majority of the time. The issue, of course, is the extreme.
On the ends of the performance spectrum, these theories that apply to most of the baseball population begin to bend - not break but bend. For instance, most pitchers will tend to have a BABIP near .300, but when you're dominating hitters to the extent that you're one of the best pitchers in the game, we should probably allow for some wiggle room on that front. Maybe not .230-.240 wiggle room, but .270 seems more plausible. And when we're talking about the best players in the game - who are definitely at the extremes - and trying to differentiate between them, we should probably keep that in mind. With that in mind, let's take a look at some of the awards "discussions" from the past week.
Might as well get to the fun one, right? It was down to either Mike Trout or Miguel Cabrera. According to FanGraphs, it wasn't particularly close - Trout at 10.4 vs. Miggy at 7.6 - and Baseball-Reference didn't have them much closer - 9.2 to 7.2. What's really funny is that Josh Donaldson, the breakout 3B of 2013, ranks higher on both lists than Cabrera. This is one of the best things about baseball research and nuanced statistics - they force us to consider things we previously hadn't considered. Though Cabrera and his .348/.442/.636 line is clearly better than Donaldson's .301/.384/.499, Donaldson ranked very highly on defense in each metric while being capable on the basepaths. Was Donaldson clearly better than Cabrera? Maybe, but this is an instance of the WARs taking everything into account with many of the writers picking-and-choosing what they care about.
But with the discussion being what it is, let's focus on Trout and Cabrera. Trout's .323/.432/.557 is clearly inferior to Cabrera's line, but defense and baserunning count. Cabrera's never been good at defense, no matter the position, and he's certainly not fast enough to think the baserunning stats don't give him enough credit. Trout, on the other hand, has the speed to back up the love he gets on the bases, and while his defensive prowess "took a hit" this season, I don't think too many people consider him a bad defender. Do the run values absolutely make sense? I won't promise you anything, but having established that the components appear reasonable, 2-3 wins is a lot for Cabrera to have to make up.
And this is where the "value" argument comes in. I have a few thoughts on that. One, when you think back on who won the MVP in 1972, do you think that guy was the best player or the best player on a playoff team? Two, why does the value argument only seem to matter for the first spot and not any further down the ballot? Three, if this is an individual award, why do his teammates matter? When it comes to voting, all I really ask is for consistency, and when it comes to people who manipulate the definition of "valuable", I tend to find that they're less consistent with how they apply that definition.
This one was more interesting, though the voting didn't indicate that. Andrew McCutchen didn't even lead in WAR in both metrics as Carlos Gomez topped him - barely - in bWAR. In fact, the worst part of the voting might have been the relative lack of enthusiasm for Gomez's season. There's a clear difference in batting lines as Carlos Gomez's .284/.334/.506 line is worse than Cutch's .317/.404/.508. The difference comes from defense. Shocker, right? Even with Gomez's sterling reputation in center, DRS's +38 run value seems a tad excessive and not in line with his career marks. Does that mean it's "wrong"? Not exactly. Gomez is a fantastic defender who was healthy and on the field for the most playing time of his career, but when it takes such an otherworldly number that's out of line with his career to top Cutch, I'll side with Cutch being better.
The argument really shouldn't stop there, however. Paul Goldschmidt, Matt Carpenter, and Joey Votto appear fairly close to Gomez and McCutchen in both lists, and they certainly deserve consideration. When it comes to differentiating between guys within a win or two of production, claiming one guy is better than another becomes a precarious proposition. Cutch beats these guys in each list, and he does so by a pretty good margin each time. But if you can find components with inconsistencies or other arguments, then make your argument. By all means. But like I said in a previous post, do so with evidence.
Either way, it seems a bit odd that McCutchen won by such a wide margin when the separation wasn't as big. For the most part, we expect the voting to reflect how close the race was, but that's not entirely true. If we ever get to the point where we agree on how to decide "value", races may be this lop-sided all the time. If you agree a guy is better, everyone puts him #1, despite only being a few runs better. That being said, Gomez being so low is kinda sad.
AL Cy Young
Pitcher W-L records aren't the best way to look at pitchers. We discussed it yesterday, and it's not worth going over again. But a pitcher can be good, collect a lot of wins, and still be the best in the game by other metrics. Looking at both lists, no one was clearly the winner. Max Scherzer (6.4 wins) beat out Anibal Sanchez (6.2) and Felix Hernandez (6.0) in fWAR, and Hishasi Iwakuma (7.0) edged out Chris Sale (6.9), Scherzer (6.7), and Sanchez (6.3) in bWAR. You can see how this can be a bit confusing. It's also where the battle over DIPS and the ability to induce weak contact come into play.
If you go straight DIPS and peripherals, Scherzer is probably your guy, but as we talked about at the beginning, DIPS generally assumes that a pitcher can't induce noticeably weaker contact than everyone else. bWAR looks mostly at runs allowed with the adjustments for defense and park, but while BABIP can bend, there are reasonable limits. Iwakuma is the test case. His 2.66 ERA is quite different from his 3.44 FIP. Is the 3.44 FIP more appropriate because his .252 BABIP isn't a result of his talent, or is he simply so good that FIP doesn't properly evaluate him? Or is there some middle ground?
The real truth of the matter is that we have several very, very good pitchers, and none of them really separated themselves. Did Scherzer dominate the voting because no one was particularly close to his 21 win total? Probably. And that's definitely the wrong reason to vote for him. But does that fact make Scherzer less deserving of winning? Most certainly not. The production levels of these pitchers was simply too close to make a definitive statement. But a vote has to be cast, and you can't just shrug your shoulders and say, "All of them." A decision must be made. Just make sure, like I said earlier, that you're using evidence for that argument.
NL Cy Young
Here's another interesting battle that didn't end up that way in the voting. Clayton Kershaw, to avoid any drama, won both lists of WAR. fWAR had a closer battle with Adam Wainwright and Matt Harvey as essentially his equals, but Cliff Lee is the only pitcher close in bWAR. Lee's 2.82 FIP isn't as spectacular as the others, but the park and defense adjustment combined with his 222 innings to put him closer according to Baseball-Reference.
Kershaw and Harvey have an interesting matchup as well. Kershaw had a worse FIP but more innings, but his ERA was half a run better. bWAR favors ERA in this instance, and along with the 50 or so extra innings, Kershaw massacres Harvey in bWAR. As for Wainwright, his 2.94 ERA is part of the reason he doesn't do so well in bWAR. Who's the best? Eh, I'm not particularly sure. Kershaw's ERA was likely a bit better than his talent level, but it's hard to make the argument that the .251 BABIP, etc. aren't at least partly product of Kershaw's dominance. Plus, Kershaw's just really good, so I won't argue.
Now comes the pitchers as MVP argument. By definition, pitchers are allowed to be MVP, and Kershaw certainly deserved some attention as he was either 3rd or 4th in overall win production on each list. But I do understand the desire to have a position player award. Cy Youngs are only for pitchers, and while the Hank Aaron Award exists for position players, it only takes hitting into account. It would be nice if we had a position player-only award that evaluated them on the three areas of hitting, baserunning, and defense while having the MVPs be more open to pitchers. But we do not have such things, and until that moment, I again simply ask for consistency.
Most of the awards don't have obvious answers, and Trout being clearly the best player is a bit of an anomaly. Especially when we're parsing the differences between the very best players in baseball, the differences are likely to be minute. And yet, we spend a lot of time arguing over who should win. Arguing is part of the fun, but it becomes a bit too serious and an attempt to assert authority at other times.
At some point, we have to stop trying to have the "right" answer and focus more on finding "appropriate" answers.