What are sabermetrics?

A few years ago, I painted the interior walls of an apartment with a friend. Neither of us had ever endeavored a paint job of that magnitude before, but we figured it wasn’t exactly rocket science — tape the moldings, paint the walls.

The actual painting part wasn’t terrible, but taping all the edges turned out to be a huge pain in the ass. We spent at least as much time taping as we did painting, and the project took us about twice as long as we expected.

Just before we finished, the cable guy came. He complimented our paint job, and asked if we had taped up all the moldings. We said that we had, and he informed us to the existence of paint edgers, an inexpensive tool that paints the edges of walls without the need for all that tape.

We cursed ourselves for not doing more research and cursed fate (and probably Cablevision) for sending the cable guy so late in our process, but at no point did we curse the paint edger.

That’s why it’s a bit weird to me, as I sort through all the reactions to Sandy Alderson’s introductory press conference at Citi Field on Friday, that so many people seemed to get so riled up about sabermetrics.

For one thing, I don’t even know what “sabermetrics” means. I know it involves baseball and statistics, and I know that lots of people seem willing to speak or write on behalf of all so-called sabermetricians. But which stats define sabermetrics? It’s not batting average; we know that. Is it on-base percentage, or is that still too basic? It strikes me as strange that we should need a fancy term for those who recognize the merits of hitters that get on base often.

My understanding has always been that the numbers we throw under the umbrella of sabermetrics are those that aim to give us a more precise understanding of a player’s value than the so-called traditional ones on the back of a baseball card, and that “sabermetrics” itself refers to the pursuit of those more precise metrics.

The book Moneyball, contrary to widespread belief, was not just about sabermetrics. It was about a cash-strapped baseball team identifying an inefficiency in the market and taking advantage of it. Running a successful business.

So I get a bit confused when I see debate over when Alderson first started using sabermetrics, like he at some point flipped on a light switch to enact sabermetrics, and from there his team was a sabermetric team. I’m pretty sure it doesn’t work like that. All stats are just tools, and every team uses stats, among other tools, to evaluate players.

That’s all. No real point in getting frustrated about it. Some teams use the tape and some teams use the edger, and probably most teams use both depending on the circumstances, and everyone’s got an opinion on which option works better. The point is there’s no real good reason to get upset and say, “f@#$ you, it’s tape!” or to be all, “yield to the dominance of the edger!” because it’s really silly to get so worked up over tools.

If you hope Sandy Alderson uses sabermetrics and Moneyball to run the Mets, then great. If you hope he doesn’t, that’s fine too. Both of those words are just big sweeping labels assigned to reasonably simple concepts, and if you want to use them or not use them to describe what Alderson does as Mets GM, you know, whatever.

All I care is that he seems dedicated to running the team the right way, and appears apt to do so.

10 thoughts on “What are sabermetrics?

  1. First of, the paint. I bought a house a few years ago that needed a major cosmetic makeover, in other words I repained every room of the house over the course of the next year.

    Agree on your first point, the taping is the biggest F-in pain. Once you get the thing taped, the painting feels like the home stretch.

    Second though, dont beat yourself up about the painters edge. Not sure if you have subsequently tried any of these tools used for painting edges, but I’ll tell you, they all are for the most part useless. You can paint edges with them, and save time, but it wont come out looking as good as it did when you taped it. REst assured knowing that you did the right thing by taping, dont beat yourself up. (side note, if you are painting, there is this green painters tape out there now called frog tape, best sh*t on the market, razor sharp edges)

    And to the casual fan like myself, not into all the hardcore number crunching, I think “sabermetrics” really has just become a generic term to most, to describe the use of advanced stats to evaluate a player as opposed to watching the guy play.

  2. Sabremetrics are tools. To use your painting analogy, you can do a terrible job with a paint edger, or a terrible job with tape. Or a good job with either. It all depends on how you use the tools.

    Advanced stats are useful, but they are not the only way to evaluate, and the new forms of evaluations have as many flaws as the old ones (just different ones). In addition, they often purport to show what the do not actually show (there’s a surprising amount of poor mathematics in sabermetrics).

    In any case, it’s as dumb to think they are the only way to evaluate as it is to think older methods are the only way to evaluate.

  3. For a lot of folk– whether it’s bullheaded, I’m-out-of-school-so-I-don’t-want-to-learn-anything-new sorts, or certain priggish/condescending Sons of McCracken– it’s not even a question of opinion anymore; it’s a question of what two people see when they look at the same thing, and one’s incredulity that the other party will not acknowledge that which seems self-evident to the first.

    Like, if you’ve got two random people in a room to tackle the paint-edge problem, and one says they prefer the edger, and the other simply sees a round space. (“It’s round, and there have never been edges to this room; I don’t see what the issue is with just going ahead and painting it. Do you ever actually paint rooms, or do you just sit in your mother’s basement coming up with contingency plans?”)

  4. I’ve always suggested that when citing the stats jargon, to make it similar to actual scientific reporting. Tell me what an average player’s numbers are when citing a specific player. What is the standard deviation so I can gauge how good/bad a player’s stats are. Finally, there should be a way to determine statistical significance of the numbers if any.

    These are all characteristics of how stats are presented in the scientific literature. They are based on convention which helps all scientists immediately understand the stats on a single glance. The stats in baseball may be just as valid, however, they certainly aren’t presented in a proper accessible way, in my opinion, which is why there is such disconnect between the sabrmetrics crowd and many other fans. Who wants to go look at a reference just to understand and respond to a single post on a blog?

    Let me give a completely random example. If I say: “Player X is awesome, his OBP is .340!”
    What does this mean? I now have to go look at a host of references to figure out a) what OBP is b) what an average player’s OBP is, and try to figure out how much one has to be above that average OBP before it means anything.

    Now, let me present it differently: “Player X is awesome, his OBP is .340 (M=0.250, SD = 0.05)! Now, I can easily see that an average player’s (the mean [M]) OBP is almost 2 standard deviations below player X, thus player X is clearly far superior. No arguments. This is a much more accessible way of presenting the same info, and makes everyone’s lives easier.

    • I’m with you on this. I see this all the time in sportwritting, and not just with the advanced and hardcore stats, ist with everthing.

      Writers never give benchmarks as to whats normal. The reality of it is that anytime a statistic is given, unless its commonly known (something like batting average), a benchmark should be given or examples of what the norm is should be provided for context.

  5. A few years ago, before Sabermetrics became the new buzz word, I saw a nice documentary on one of the Science channels explaining the principles behind what sabermetrics.

    For what I remember; using advanced math they figured out the odds of what it takes to score runs in different situations and found that biased on averages, what type of player would best fit those odds.

    Examples;

    They found that the more runners get on base, the better the odds are of scoring (what a surprise) so the results are on base percentage is very important.

    The also found that (on average) the odds are better to score when you have a runner on first and no outs than a runner on 2nd with one out, so they say that sacrifice bunt hurts you scoring chances.

    They also figured that the odds of successfully stealing 2nd base and then scoring are less than the odds of a running from 1st, so the results say don’t steal.

    They also figured out how many runs in 162games would a team of nine Ruths would score and a team of nine Kofaxs would score. They even did different combo’s like batting five Ruths and four Kofaxs.

    They also figured that the law of averages say, your batting order doesn’t really matter. That a traditional lineup (Speed and average at the top Power in the middle, Weaker hitters at the bottom and pitcher hitting 9th) vs a random line up of the same hitters would score close to the same number of runs at the end of the year.

    With these results, they figured out new ways to measure a players value leading to all the new stats we are seeing.

    My problem with these results is, one Jose Reyes has a much better chance than the league average to steal a base, so his odds are better to score than the league average is. Castillo can’t drive in a runner from first, so the odds are better to have him sacrifice. Hitters ability to get good pitches to hit greatly depend on who hits around them in the line up. Look how much better average hitters do in the Yankees lineup loaded with great hitters than they do in weaker lineups on other teams. Look at how much better Wright made contact when he had a healthy and productive Beltran and Delgado hitting after him than what he does when he has Murphy, Frenchy or Davis for protection.

    Advanced stats still places value on what whoever created the stat feels is valuable. Someone perceives OBP has high importance, so OBP has higher value in their metrics.

    Not saying I dislike advanced stats, but they have their flaws and you can’t judge a player solely on what they say.

    • “My problem with these results is, one Jose Reyes has a much better chance than the league average to steal a base, so his odds are better to score than the league average is. Castillo can’t drive in a runner from first, so the odds are better to have him sacrifice.”

      The basic idea that seems to hang over all of this stuff re: run probability is that, in the longview, outs, as it turns out, are generally bad. So, in the overwhelming majority of cases, even “productive” outs are poor strategic choices, since more often than not, they lead to smaller-scoring innings.

      “Hitters ability to get good pitches to hit greatly depend on who hits around them in the line up. Look how much better average hitters do in the Yankees lineup loaded with great hitters than they do in weaker lineups on other teams. Look at how much better Wright made contact when he had a healthy and productive Beltran and Delgado hitting after him than what he does when he has Murphy, Frenchy or Davis for protection.”

      Seems so, doesn’t it? But it’s not so, muh fren. (Yankee hitters put up better COUNTING stats like runs or RBIs, but they’re not better situationally because they’ve got “Big Tex” staring down the pitcher in his periphery.)

      These links are a good starting point:

      http://www.baseballprospectus.com/unfiltered/?p=1042

      http://dodgerthoughts.baseballtoaster.com/archives/1125205.html

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s