Saturday, November 17, 2012

A baseball peace offering

Since the debate started over whether Miguel Cabrera or Mike Trout would be the American League MVP, I've always felt it was more than the standard debate over who was the league's best player this season, but perhaps the climactic battle between old and new ways of looking at the game.

The American League Cy Young Award selections of Zack Greinke and his 16-8 record and especially Felix Hernandez and his 13-12 record in 2010 were major achievements for the sabermetrics crowd (the type I usually call "stat-geeks," but I'm not here for reasons that will be clear below), in that one of the major points is that wins are a lousy way to judge how good a pitcher is.

But if Trout had won the MVP, it would have been their crowning glory, as the guy who won the first Triple Crown in 45 years would have lost largely because of statistically based arguments that Trout's prowess in the field and on the bases outweighed Cabrera's achievements, including leading the league in two of the great scourges of the sabermetrics movement -- batting average and RBI.

However, Cabrera won, which left people like Brian Vaughn of Call to the Pen more than a little disappointed.
 First off, David Price took home the AL Cy Young narrowly over Justin Verlander without actually ever being better than him at anything. Then we have Miguel Cabrera just crushing Mike Trout in the AL MVP voting despite only topping him in one aspect of the game, and narrowly at that. Maybe we’re not quite as advanced as I was starting to think.
And then Mitch Albom got loose.

In a battle of computer analysis versus people who still watch baseball as, you know, a sport, what we saw with our Detroit vision was what most voters saw as well.

Today, every stat matters. There is no end to the appetite for categories -- from OBP to OPS to WAR. I mean, OMG! The number of triples hit while wearing a certain-colored underwear is probably being measured as we speak.

We need to slow down the shoveling of raw data into the "what can we come up with next?" machine. It is actually creating a divide between those who like to watch the game of baseball and those who want to reduce it to binary code.
As you might imagine, his comments did not meet with universal approval.

And King Kaufman was also something less than amused on the Twitter machine.
  
People are going to look back on Trout's 2012 and say, "My gosh, those idiots didn't know what they were watching!"
If you'd told me 30 years ago that one profession would come to embrace ignorance as a virtue, I NEVER would have guessed journalism.  

I bet you wouldn't be able to find me an MLB GM who would trade Trout's 2012 for Cabrera's 2012, age, etc. not considered.
I actually agreed with the main ideas of Albom's column, that Cabrera should have been MVP and that anayltics don't have all the answers, but insulting the people who disagree with you isn't going to help your argument. The same goes for Kaufman, who, to be fair, also said that Cabrera was "stupendous" this year and is headed for the "inner circle Hall of Fame," which apparently is the very best of the best.

It is in this spirit that I would like to make a peace offering to the sabermetrics crowd, including my ambassador to analytics, my guy Poopsie out in Chicago. In the interest of peace, here are the things I'd be willing to concede:

1. Batting average and RBI aren't the be-all and end-all -- If the point of the game is not making outs, on-base percentage is a more-precise way of measuring that, and RBI require having men on base in front of you, unless you hit more solo home runs than anyone else has ever hit home runs.

2. You don't have to win a lot of games to be a great pitcher -- If Felix Hernandez played for a team that had even an average offense, he might never lose.

3. Fielding percentage and errors aren't the best ways to measure defense -- We can also call this the "Yes, I know Derek Jeter doesn't have great range" concession.

4. No, I don't know exactly what "clutch" is, or even if it exists.

5. Just because I don't know how to calculate VORP, WAR and endless other statistics doesn't necessarily mean they're worthless.

And these are the concessions I want in exchange:

1 Adam Dunn is no better than a decent hitter, and often a terrible one -- It's great that he has a great eye, and when he hits balls, they stay hit, but with a lifetime batting average of .240, if a pitcher can manage to throw the ball over the plate, Dunn will make outs more than 75 percent of the time. He hit .204 this year, meaning he made outs close to 80 percent of the time in those circumstances. Dunn is an extreme example, but it shows that batting average has some value, in that it does measure what someone does when a pitcher throws strikes.

1a. If RBI are a team stat, so are runs -- If you're going to say that driving in runs is a function of men getting on base in front of a hitter, you also have to say that scoring runs is a function of the hitters behind a baserunner, unless he steals home a lot.

2. Good pitchers tend to win a lot -- There have always been Greinkes and Hernandezes, pitchers who pitch better than their records on terrible teams. Take a look at Nolan Ryan's 1987, for instance. In general, however, if a pitcher wins a lot of games, he had a good year.

3. Range is great, but so is making the plays you should make -- Call it the "At least Derek Jeter isn't giving teams four outs an inning" concession. If you can't stomach saying that, watch Eduardo Nunez play shortstop; you'll be crying for Jeter in no time.

4. If "clutch" doesn't exist, explain your first thought when David Ortiz or Alex Rodriguez come up in a big spot.

5. People are always going to be skeptical of any statistic they don't understand.

But even beyond the way we look at baseball, there's one more trade-off I'm willing to make.

I won't call sabermetric types geeks, if they don't call skeptics stupid.

Deal?

3 comments:

  1. I don't know if you're talking about me, but I don't call skeptics stupid. That is, unless they happen to be stupid, or do stupid things. I called Mitch Albom's column stupid. That's because it was stupid, not because it was skeptical. You can be skeptical and smart. You can also be skeptical and stupid. Like Albom in that column.

    It's not skepticism to mischaracterize or willfully misunderstand something, then argue against the phony thing you just made up. It's just really bad journalism. Or stupid.

    I actually don't think Albom is stupid. I think he pretended to be stupid in that column to pander to his hometown readers.

    I don't speak on behalf of any sabermetric crowd, but here are my responses to the five concessions you want "in exchange" for admitting certain basic truths about baseball, such as "Batting average and RBI aren't the be-all and end-all."

    1. Adam Dunn is better than a decent hitter. You write, "with a lifetime batting average of .240, if a pitcher can manage to throw the ball over the plate, Dunn will make outs more than 75 percent of the time." Dunn has walked in 16.2 percent of his career plate appearances. During his career the overall MLB walk rate has been between 7 and 9 percent. If pitchers can get Dunn out 76 percent of the time just by throwing strikes, why do you think the same pitchers who walk everybody else 8 percent of the time walk Dunn twice as often? Has he just been spectacularly lucky year in and year out?

    This past season was the second worst of Dunn's career, after his disastrous 2011. He only hit .204, but he created 5.5 runs for every 27 outs he created. Another way to look at that is: A lineup of nine 2012 Adam Dunn's would have scored about 5.5 runs per game. The average AL team scored 4.5 runs per game. So Dunn, in one of his worst years, was more than 20 percent better than league average, looking at just that one very general stat.

    1a. I don't know of anyone in the sabermetrics crowd who wouldn't agree with this. Runs are a lineup-dependent stat, just like RBIs.

    2. There is some correlation between wins and quality pitching. This is especially true over a long career. Nobody gets anywhere near 300 wins without being a hell of a pitcher. But in a given year, win totals can be hugely misleading. Just one example each way: Look at Matt Cain's 2008 and Randy Johnson's 2006. Cain won eight games. Johnson won 17. By any measure, Cain in 2008 was a vastly better pitcher than Johnson was in 2006.

    You have to be a good pitcher to win a lot of games. But one season's worth of wins is not "a lot," in the scheme of things.

    3. Every time Jeter does not get to a ball that another shortstop would get to, he is, in effect, giving the other team four outs just as surely as if he'd gotten to the ball and made an error. Asdrubal Cabrera and Alcides Escobar led AL shortstops in errors this year with 19. So 19 times they gave the other team four outs. Jeter's range factor per 9 innings was 3.76 this year, meaning he averaged 3.76 (putouts plus assists) per nine innings played at short. The league average for shortstops was 4.43. By that stat, over the 135 games Jeter played at short, there were 90 plays he didn't make at shortstop that the league-average shortstop would have made. Jeter also made 10 errors. You wouldn't trade those nine balls Jeter got to and didn't make an error on for the 90 balls he didn't get to?

    4. My first thought when David Ortiz or Alex Rodriguez comes up in a big spot is: "Confirmation bias is a powerful thing."

    5. If you don't understand something, your opinion about it is worthless. You have every right to be skeptical. But if you can't be bothered to try to understand something, don't pontificate about it. You'll just end up sounding stupid.

    ReplyDelete
  2. King --

    Thanks for reading. When I saw I had a comment, I wasn't expecting it to be from you.

    You didn't call skeptics stupid, but you said people who didn't vote for Mike Trout were idiots, that journalists are embracing ignorance and that I end up sounding stupid if I'm not being bothered to try to understand something.

    Regarding Albom's column, which I learned about due to your Twitter feed, we disagree on the conclusion, but we both agree that it was a bad column. I thought a lot of it was childish, and it made his skepticism look petty.

    Regarding skepticism, yes, I am skeptical, and I admit it's because I don't understand a lot of where sabermetrics come from. That's why I wrote, "Just because I don't know how to calculate VORP, WAR and endless other statistics doesn't necessarily mean they're worthless." People are skeptical of things they don't understand.

    It's funny you mention the lineup of nine Adam Dunns, because I was chatting with my friend Poopsie (a former co-worker, it's his nickname) about Dunn versus someone sabermetrics types don't like, who I nominated as Juan Pierre. How many runs would a lineup of nine Juan Pierres score? I said someone out there probably had figured it out.

    I also wrote that Dunn is an extreme example in that his batting average is so low and his on-base percentage is so high in relation. I have no idea why that is, other than that he obviously has a good eye.

    Can I give you specific examples where people have cited RBI as not being meaningful because it's a team stat while saying runs are meaningful? I confess I cannot, but I'm fairly certain I've seen them. I wish I could do better than that, but I can't. If it turns out my memory is mistaken, then I'll admit I'm wrong.

    When it comes to fielding stats, I'm asking you this because I honestly do not know. Is there a fielding stat that is in no way based on any judgments at all, but is strictly objective?

    Finally, if you only read the post to figure out why my opinion was worthless, do me a favor and read it again. I called it a peace offering because I'm willing to say that while I tend to be a traditionalist, people who believe otherwise might have valid points. I'm just asking for the same from the other side, unless they (or you specifically, since it's you I'm responding to) think they're right about everything and people like me have no valid points.

    ReplyDelete
  3. Well, I didn't exactly call people who didn't vote for Trout idiots, but I can see how it reads that way. What I meant by that is that I think that in the future, the arguments for Cabrera are going to be so outmoded that people won't even remember them, and the argument for Trout, looking back, will seem so obvious that people will just shake their heads wondering what the hell we were thinking.

    And I was talking about Albom when I wrote, "if you can't be bothered to try to understand something, don't pontificate about it. You'll just end up sounding stupid." You admitted you didn't understand some sabermetric concepts and stats, but you did not pontificate about them, as Albom, who also either didn't or pretended not to understand them, did.

    The answer to your question about Juan Pierre, this year, is 5.0. He had one of his best years this year. The league-average NL hitter this year created 4.5 runs per game. Pierre got on base a lot and stole a lot of bases at a high success rate.

    The fielding stat I mentioned -- range factor per nine innings -- is strictly objective. It is simply putouts plus assists divided by innings played, times nine. All of the work that sabermetricians do on fielding stats, as far as I know, is an attempt to come up with objective stats not based on judgments. What they're trying to get away from is the non-objective, judgmental assessment of fielding. That is, "Derek Jeter is a great shortstop because I've seen him make so many great plays."

    I didn't read your post to figure out why your opinion is worthless. I don't think your opinion is worthless. I merely responded to your "demands," the things you said you wanted in exchange for your "concessions" about sabermetrics. I don't think "people like [you] have no valid points." But that doesn't mean I'm going to pretend I think points are valid when I don't just to be nice or something. I think your point 1B is valid, and your point 5 -- "People are always going to be skeptical of any statistic they don't understand" -- is true, I suppose, but not in any meaningful way. If someone doesn't understand something, then by definition their opinion of that thing is not particularly valid. That's what I meant by my response to number 5.

    I don't understand the first thing about nuclear physics. Why would anyone care, then, what I have to say about nuclear physics? If a nuclear physicist told me about some interesting and controversial theory in nuclear physics, I could say I was skeptical of this theory. But so what? I'd be speaking from ignorance. (If you think my comparing baseball to nuclear physics makes baseball seem misleadingly complex, you can substitute "nuclear physics" in this paragraph with "Pokemon" or "gardening.")

    Don't understand something? Learn about it. Understand it. Then your opinion about that thing will be worth something. Until then, it won't be, which is why Mitch Albom's column is about as worthwhile as a column I would write about Pokemon or gardening.

    I get that it's a bit harsh of me to say, "There is no reasonable argument for Miguel Cabrera being a better baseball player than Mike Trout in 2012," and that when I say that some people's feelings will be hurt. I just don't believe that "you hurt my feelings" is a valid argument. There are people somewhere whose feelings will be hurt if I say there is no reasonable argument for the earth being flat. But their hurt feelings don't make their flat-earth arguments any more reasonable.

    ReplyDelete