Thursday, September 25, 2008

Holy Cow, Does Jon Heyman Hate VORP

I just crunched some numbers and data about various things that exist, using a scientific process verified by several mathematicians at top universities, and I came up with some interesting results.

Here are the five most boring things in the world, in order:

5. Slowed-down time-lapse photography of a small puddle of room-temperature water evaporating.
4. Two people you have never met, wearing identical colorless shirts and pants, talking about the dreams they had last night.
3. Debating what "Valuable" means w/r/t "Most Valuable Player."
2. Lying in a sensory deprivation tank and staring straight ahead at a blank wall while you listen to white noise.
1. Ann Coulter

So right away, we're on risky ground with articles like this one from's Jon Heyman -- it's the third most boring thing in the world, according to science. And beyond that, it's borderline hysterical in its boring and righteous anger:

Once again, VORP has nothing to do with MVP

Zero. There's a number the stat people will understand.

That's the relationship between VORP, the stat that the stat people love, and MVP.

Well, that's just not true.

If you hate the stat, you hate the stat. I'm not sure why you should hate a stat that uses a relatively sophisticated model to calculate not just how good a guy's stats are, but also what position he plays, and essentially evaluates how hard he is to replace (the true measure of "value," to me...oh God...I'm the third-most boring guy in the world right now). But if you hate it, you hate it. Not much I can do but keep posting on this blog.

You cannot, however, say that there is "zero" relationship between VORP and MVP. Because even if you choose to ignore it, it exists. Last year ARod won the MVP, and was 1st in baseball in VORP. Rollins, kind of a crummy pick, was at least top-10 (9th, actually, in the NL, behind several other more deserving candidates). People were generally happy with the choice of Ryan Howard in 2006...and it just so happens that he was 2nd in the NL in VORP, right behind Pujols. Morneau was a terrible choice, much-reviled and controversial...and he was 13th in the AL. I don't really remember anyone complaining about Pujols or ARod in 2005...and they were 2-3 in VORP in all of baseball. Only DLee was above Pujols in the NL, and if he had won, nobody would've been angry.

Keep looking at the list. The MVPs of the league are generally very high VORPulators, year-to-year. So it makes some sense that in order to predict who will win, or who should win, we can look at VORP. Right?

Baseball Prospectus, as of a few days ago, had Alex Rodriguez leading the AL in VORP (which stands for (Value Over Replacement Player) the stat its enthusiasts think is the best stat in the world to determine player value, and also the best to help determine who's the Most Valuable Player.

Maybe not "the best," but, you know, pretty effing good, I think. Better than batting average.

But as you can see, while VORP may tell you something, it shouldn't determine who wins the MVP award. Beyond containing two of the letters in MVP, there appears to be almost no relationship whatsoever here.


I happened [sic] to love A-Rod. He's turned himself into a very good third baseman (he's probably the best defender on the Yankees), he's a three-time MVP (though I don't believe he deserved it the year his Rangers finished last), he's the best all-around player in the game and he's not among the prime list of reasons for the Yankees' demise this year (though, there are plenty of Yankees officials who'd have him on that list).

Yet, A-Rod shouldn't sniff the MVP award this year.

I'm with you on this, for the record. Like every other bored American who is bored at the yearly debate over what boring ways we should boringly parse the boring term "Valuable," when there is no 100% obvious winner, like Barry Bonds the year he steroided .370/.582/.799 with 46 contes for a team that made the playoffs, I take the famous approach espoused by Supreme Court Justice Stewart in reference to pornography, who said, "I can't define exactly what pornography is, but oh lordy, this FMF pictorial has me hard as a diamond." In other words, given a number of players with roughly equal stats, there is a kind of gut-level instinct one uses to cast the tie-breaking vote. That could be: did the guy's team make the playoffs, and was he an important part of the stretch run? Did the guy happen to have a lot of hits in crucial situations? Did other players on his team go down with injuries, making his production even more important to his team? And perhaps most importantly, is this guy a SS or CF or C or something, meaning that his production from that position is even more valuable, given the paucity of high-production players at that position?

(In other words, in addition to whatever kind of gut-checking you want to do, you can look at VORP and WPA and stuff like that.)

For the record, again, ARod's WPA is barely above 0 this year -- 0.28. Look at Mauer's. Or Pedroia's. Or a bunch of other people's. He is not the MVP this year.

If devotees of VORP (I'm already on their bad side after calling them VORPies last year) think their stat is key to determining the MVP, they should think again. It's worth a glance, at best.

It's worth a glance at least. It's a measure of how valuable a player is, compared to other people at his position. What is the downside of looking at it very seriously?

But VORP is supposed to be an all-encompassing stat,

No it's not. Doesn't account for defense, and doesn't account for "clutch" the way WPA does. No one is arguing it is all-encompassing. No one. What people do argue, occasionally, is that if a guy isn't even in the like top 10 for VORP or something, like Justin Morneau that year, maybe he shouldn't be the MVP.

and it led some numbers people to determine that Hanley Ramirez was a viable NL MVP candidate last year. And led many to say that David Wright was the NL MVP in a year in which Wright's Mets choked (Wright himself says no way was he MVP).

1. HRam was, indeed, a viable candidate.

2. What is Wright supposed to say? "I know my team choked harder than any team in the history of sports, but: Me for MVP!"

3. I don't understand why people debate about whether a guy's team has to make the playoffs to win the MVP, and some say "yes" and some say "no," but when a guy's team just barely misses the playoffs at the last possible second, meaning that they were in the race the whole year, and the guy in question hit .352/.432/.602 in September with 6 HR and 9 2B, it's like, "No fucking way that guy is teh MVP!!1!!!!!111!!!"

VORP, like other stats, doesn't come close to telling you everything. It doesn't take into account how a hitter hits in the clutch (oddly enough, some stat people think that's just luck, anyway),

See above. Then see WPA page. Then remember that no one in the world with a brain thinks that the MVP award should be blindly handed out to the guy with the best VORP.

As for "clutch" "just being luck," what we actually think is that it's very hard to be "clutch" year in and year out. (For example, ARod's WPA last year was 6.85. This year it's 0.28. Two excellent offensive years, two wildly different "clutch" results.)

or how many meaningful games he played in (at last count Grady Sizemore was high up on the VORP list, as well). VORP has some value. But like all other stats, it doesn't replace watching the games or following the season.

I have never watched a baseball game, so I can't speak to this. I'm not even sure what it is. What I can tell you is: watch live baseball all you want. I'll be in my grandmother's attic (following a legal dispute over squatter's rights with my mom w/r/t her basement), staring at my computer, looking at a little thing I like to call "data." That's all I care about. Data. Raw data. Baseball is good for one thing only: the production of data. That's what I believe. If I and my friends had it my way, the games wouldn't even be "played," but rather "simulated" by 1000 PCs, and the results would be downloaded directly into my brain through Optical Quanta Resonance (OQR), and instead of "discussing" the games the next day, my friends and I would just await the Retinal Scans and then text each other brief congratulations, depending on whose favorite "team" won, and then we would all go on with our lives, grateful that the annoyance of actual "baseball" had been removed from our lives, allowing us to spend more time writing code for our start-up social network site, which we are I think going to call "Together-ing!"

A-Rod may have the best VORP. But he shouldn't be on anyone's MVP ballot, much less at the top of the ballot.

I now want ARod to win.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.