September 14, 2015

Sorry, Chip, Joe, and Glavine but Statistics Progress

During Thurday night/Friday early morning’s Braves-Mets game, Chip Carey, Joe Simpson, and Tom Glavine discussed pitcher wins, RBI, statistics, and sabermetrics.

Chip Carey started things off by bringing up Shelby Miller and pitcher wins, and bringing up the argument that pitcher wins essentially don’t matter.  He asked his broadcast partners what they thought of this view.

Tom Glavine argued that while he sees the point, you can’t convince him that wins don’t matter.  They are an imperfect stat but all stats are imperfect, and we should look at pitcher wins in conjunction with other stats.  He said that if we are going to do away with wins, we need to replace them with with something, and nothing else is perfect.  This is called the nirvana fallacy: because there are no perfect stats, we should just stick with wins.  Of course this ignores the fact that there are different degrees on the spectrum of perfection and imperfection, and there are stats that are more perfect that wins, even if they aren’t perfect.

Joe Simpson didn’t really make any kind of coherent argument.  His argument was essentially a strawman attack on sabermetrics and people who use sabermetrics.  He even went so far as agreeing with the caricature (that Carey brought up) that sabermetrics people sit in front of computers in their parents basement.

Glavine said that sabermetrics has a place in the game but both Glavine and Simpson agreed on the view that sabermetrics people are just trying to come up with something new for the purposes of bringing attention to themselves.  They argued from tradition (another logical fallacy), saying wins have been around forever, so why do we need anything else.

The point of sabermetrics is not perfection or to reinvent the wheel for the sake of reinventing the wheel.  The whole premise of sabermetrics is this: if we are going to use statistics in baseball, we should use the best, most telling ones.  In the early days of baseball, stats like pitcher wins and RBI were nice, simple stats that gave us insight.  As time has gone on, we’ve learned the problems and weaknesses with those stats, in terms of what they do and don’t tell us about players, and statistically-inclined baseball fans developed better, more telling statistics.  Statistically-inclined baseball fans will likely continued to develop better, more telling stats.  Sabermetrics is acknowledging this process instead of throwing your hands up and saying what’s always worked, statistically, will continue to work.

By now, most of you reading this understand all the issues with pitcher wins.  Anyone who has watched Shelby Miller pitch and know what he’s done understands the problems with pitcher wins as much as anyone.  Even Carey, Glavine, and Simpson understand the issues with pitcher wins.  They acknowledge these issues, but they still want to keep pitcher wins around, but use pitcher wins in conjunction with other stats.  Now, I agree that we need to look at more than just one statistic to evaluate players, and go beyond statistics when we can.  But we should pull from a variety of stats and information that actually tells us something.  We should pull from multiple stats to get the big picture.  We shouldn’t look at one stat, then pull from others, because the one stat isn’t good enough to be a piece of the puzzle.  If you have a group of stats and you can take out one of them without missing it, that one stat is useless, so why bother with it?

Joe Simpson brought up RBI, and the argument that a hitter’s RBI total is largely a product of opportunity and not the hitter’s skills.  He and Glavine did not believe that a team could put anyone in a certain spot in the order and that player would drive in lots of runs.  Well, they should because it’s happened.  It happened with the Braves with Jeff Francoeur.  For a time, Francoeur played everyday and hit around the middle of the order, behind good on-base guys, and did not take a lot of walks, and therefore was able to drive in 100 runs in a couple of season.  In 2006 and 2007 Francoeur drove in 100 runs each season, yet he made lots of outs (as evidenced by his .315 OBP).  He racked up RBI because he had power, he didn’t draw walks, he hit in the middle of the order, and behind some good on-base guys.

Perhaps the most egregious example of the problems with RBI is Ruben Sierra’s 1993 season.  Sierra drove in 101 runs but hit just .233 with a .288 OBP and .390 slugging.  If outside factors (like playing time, skill set, and other hitters in the order) don’t greatly influence RBI total, how did a guy who made outs at a very high rate and didn’t slug well drive in 100 runs?

With the advent of faster and faster computers, we can process statistical information easier and we can learn things we didn’t know in the past.  This is true in many things, including in baseball.  I understand the skepticism of folks coming along and refuting what’s been ingrained and rooted in the game for over 100 years.  But most sabermetrics people are simply intellectually curious people who want answers and want better statistics.  Most aren’t seeking attention.  Most respect and may be fascinated by scouting, and aren’t against it.  Most don’t look at baseball as an us versus them but as “let’s immerse ourselves in the game and learn and get as much information as we can.”

Most front office people don’t view the game like Joe Simpson, Chip Carey, or even Tom Glavine.  They are fully committed to finding, possibly coming up with, the best statistics possible, and they are aware of the problems with stats like pitcher wins and RBI.  They are fully committed to the most progressive stats just like they are fully committed to scouting.  They realize it’s not either/or but an all-of-the-above approach, if they want to field the best team possible.  Fortunately, in John Coppolella, the Braves have one of the best in advancing an all-of-the-above approach.



8 Responses to “Sorry, Chip, Joe, and Glavine but Statistics Progress”

  1. 1
    Tyler Says:

    Saber metrics enthusiasts are so annoying. People like Brian Kenny are the worst.

  2. 2
    Brent Says:

    Great article. Your points are spot on.

  3. 3
    Walker Says:

    Hmmm ignoring facts. Railing against math. No wonder Carey,Simpson and Glavine are all republicans.

  4. 4
    Shaun Says:

    Tyler, I think folks who get paid to know a lot about baseball should be baseball information enthusiasts. It doesn’t really matter the terminology or how one classifies himself or categories of information (“sabermetric,” “scouting”), they should just crave baseball information. I think too many get caught up in wanting to classify themselves or others as a certain type or in with a certain group of baseball fans. I get that some are going to process the game differently. But don’t ignore or discount or refuse information from other ways of processing the game.

    I don’t think Brian Kenny discounts scouts or scouting information. I think he just utilizes the best stats he can find. What he discounts is misinformation and stats that don’t tell us anything and are less useful.

  5. 5
    Tyler Says:

    Don’t get me wrong, I like some of the saber stuff. Not all of it, I just can’t stand to hear some of these guys talk about.

  6. 6
    Shaun Says:

    I think it’s odd that some look at “saber stuff” as a preference thing, liking it or disliking it. I think too many people view “sabermetrics” as some sort of club or group or cult or whatever that you either want to be in or you want to avoid like the plague, instead of this neutral way of analyzing and learning and processing the game that is just there.

    It’s sort of like scouting. It would be odd for someone to say, “I like some of the scouting stuff, not all of it. I just can’t stand to hear scouts talk.” Scouting is what it is. It’s this neutral way to process, analyze, learn and filter the game. The idea that someone or some group came up with scouting or sabermetrics for the primary purposes of making themselves relevant and not as a way to process the game is absurd.

  7. 7
    Tyler Says:

    I never said they use it as a way to sound smart or be relevant. I just like the game as it has been the last 100 years. Is it going to get to the point where someone is kept out of the Hall of Fame because some of their saber stats aren’t up to par? If

  8. 8
    Shaun Says:

    Tyler, people try to use statistics to help them determine whether a player did enough to meet the standards for Hall of Fame induction, whether it’s “sabermetric” stats or other stats. But isn’t it silly to separate stats into “sabermetric” or “other”? Why not just use the best stats that help you determine whether a player is worthy of the Hall of Fame or that help you determine whether a pitcher is worthy of the Cy Young or whether a player is worthy of MVP?

    To determine which stats are the best, we should look into how stats relate and correlate to winning and go behind the stats to see what they are telling us.

    Look at pitcher wins, for example. Pitcher wins, for starting pitchers, simply tell us whether a pitcher went at least 5 innings, left the game with his team in the lead, and his team never lost the lead. Just from sheer reasoning we understand that that tells us little about how well a pitcher has pitched. But there are better stats out there to get at how well a pitcher has pitched that we should weigh much more heavily.

Leave a Reply