Moneyball and “the batting average illusion”

For decades the baseball world obsessed about batting averages.  Then in the late 1970s an obscure night guard at a pork and beans cannery named Bill James published a book pointing out that the emperor had no clothes.  That batting averages and many other baseball statistics were essentially worthless, not measuring what everyone assumed they measured (such as offensive prowess.)

The batting average treats singles and home runs equally, and doesn’t even include walks, which can be a good way of getting on base.  I recall reading Bill James in the early 1980s and wondering how anyone could disagree with him.  It was all so obvious.  The sum of on base percentage and slugging percentage picks up the dual mandate; get on base, and drive in runs.

For years, the economics profession has been obsessed with inflation and inflation targeting.   Yet inflation treats supply shocks and demand shocks as if they were the same thing, and doesn’t even include the prices of new houses (the main data point for the housing bubble.) Wouldn’t it be neat if some obscure academic from a small school came along and pointed out that the emperor had no clothes?

We need a single statistic, like OBP plus SP in baseball.  Something that incorporates the Fed’s dual mandate, not just one side of the mandate.  Something that adds inflation and growth to focus on what the Fed can actually control.  Any ideas?

PS.  The publishers thought no one would be interested in Bill James’ ideas, so he was forced to self-publish his early books, and would’ve done well with something like Amazon KDP, had that existed in his time.  Perhaps our monetary policy reformer will be forced to self-publish his ideas in a blog.

PPS.  Do you think Brad Pitt looks more like Beckworth or me?  And who would play the villain (i.e. Bob Murphy?)

PPPS.  Some statistics nerds tried to go beyond Bill James, and argue that OBP and SP shouldn’t have equal weights.  People; you need to realize that there is beauty in simplicity.

PPPPS.  For those who don’t know, in 2006 Time magazine named Bill James one of the 100 most influential people are Earth.  And that was for changing the way we think about baseball.


Tags:

 
 
 

32 Responses to “Moneyball and “the batting average illusion””

  1. Gravatar of Cthorm Cthorm
    3. November 2011 at 06:05

    “People; you need to realize that there is beauty in simplicity.”

    Right – there is beauty in having a single statistic that tells the average Joe what the Fed’s policy is going to be (not to mention telling the Fed what the Fed’s policy should be). If you complicate things with moving averages of second derivatives (straw man) then you trade technical correctness for increased obscurity, which is never good when it comes to policy. It’s striking to me how significantly things have changed since Greenspan was Chairman.

  2. Gravatar of John John
    3. November 2011 at 06:07

    Calling Bob Murphy the villain is like Goliath calling David the villain. Scott’s views are much more mainstream and “progressive” and appealing to morons like Krugman.

  3. Gravatar of Becky Hargrove Becky Hargrove
    3. November 2011 at 06:36

    Per your P.S.: People no longer write letters as they once did, for a while thoughts were shared on telephone but now much is expressed on formats such as this. It really seems that blogs will become more a part of history than we presently imagine, honest “warts” and all. Publishing? Already there. I like to think that what is happening electronically in the present can one day, finally, return to local levels of society.

  4. Gravatar of ssumner ssumner
    3. November 2011 at 06:54

    Cthorm, Good point.

    John. It benefits both of us–pumps up the interest in our January Pay per view debate. People want to see good guys and villains.

    Becky. I think you are right.

  5. Gravatar of Cthorm Cthorm
    3. November 2011 at 07:34

    Thanks Scott, great post by the way. I’ll use the batting average analogy next time I try to explain the idea to my dad.

    If you find your way in my neck of the woods, I’d like to say hello. If you haven’t already that is – based on my boss’ comments on CNBC lately it sounds like you’ve already briefed him.

  6. Gravatar of Gregor Bush Gregor Bush
    3. November 2011 at 07:48

    My nominees for the villain in the upcoming smash hit NGDPball are:

    Narayana Kocherlakota: “Low rates are consistent with deflation so we need to hike rate aggressively to prevent deflation”

    Jean-Claude Trichet: “What a great job I’ve done! Inflation has been even lower than the official target!”

    Raghuram Rajan: “Rising equity prices are a bad sign. We needed to raise rates aggressively two years ago to prevent the buildup of dangerous asset bubbles.”

    Richard Fisher: “Monetary stimulus only lets governments and lazy private debtors off the hook and prevents them from making necessary painful adjustments.”

    Masaaki Shirakawa: “The Bank of Japan alone cannot end deflation. Also, the Bank of Japan cannot weaken the yen without the help of other central banks.”

    Charles Plosser: “How can monetary stimulus help employment? How do we know there’s any slack in the economy? Maybe the new NAIRU is 9%. Can you prove that it isn’t?”

    Bill Whyte: “There’s too much debt, it’s all structural, we were making too many cars and too much of everything. Demand will be weak for years because everyone has to pay down their debt.”

    Joeseph Stiglitz: “Monetary stimulus only helps the rich and the evil bankers”

    Richard Koo: “Because of all the debt, monetary stimulus is completely powerless to raise spending as is the case in Japan – and you’d better not try it because it will cause a hyperinflation”

    Mohamed El-Erian: “Slow growth is the new normal because we’re in a debt deflation – and monetary stimulus will only casue high inflation.”

    Of course Scott Sumner will play the hero.

  7. Gravatar of David Clayton David Clayton
    3. November 2011 at 08:12

    This is a great post. Really.

  8. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    3. November 2011 at 09:09

    Indeed, Edward Teller wrote a book with the title, ‘The Pursuit of Simplicity’ (iirc). Not to mention KISS. But what about Runs Produced (RBIs + runs scored – homeruns). After all, that is how baseball games are won, by outscoring your opponent.

  9. Gravatar of John John
    3. November 2011 at 09:30

    Scott,

    Good for you for agreeing to debate him. I’ll be sure to fork over the money for that one. I wouldn’t consider myself a marginal buyer.

  10. Gravatar of Neal Neal
    3. November 2011 at 09:36

    Since I agree with the post, let me snarkily add:

    “… inflation and inflation …”
    BAN IT

  11. Gravatar of Benny Lava Benny Lava
    3. November 2011 at 09:51

    Obviously the villain is Ron Paul. Think TV market, people. Make G Gordon Liddy his evil henchman.

  12. Gravatar of caveat bettor caveat bettor
    3. November 2011 at 10:24

    I’ve lived under the regime of the Sharpe/Information Ratios. Perhaps our nation’s economic appointees can live under them too, but using changes to NGDP instead of returns. Of course, ‘volatility’ is only a proxy for ‘risk’. And these ratios are not very intuitive–it is amazing how I do not have a CFA but need to explain the domain of Sharpe/Info to holders of that designation. Finally, what is the benchmark? (Canada?!?) But I thought it might be good to get things started.

  13. Gravatar of Benjamin Cole Benjamin Cole
    3. November 2011 at 10:27

    I like everything about this Scott Sumner post except the last sentence.

    Are you implying baseball is less important than the economy? World events?

    (Personally, I always thought base-stealing was vastly under-rated).

  14. Gravatar of mattmaison mattmaison
    3. November 2011 at 10:54

    Beyond OPS, James’ biggest contribution was arguably the Runs Created statistic which led to the field of the field sabermetrics and stats such as WAR, incorporating fielding and base running.

    http://www.fangraphs.com/library/index.php/misc/war/

  15. Gravatar of OneEyedMan OneEyedMan
    3. November 2011 at 11:31

    I What you are describing sounds like the Taylor rule. Since you have obviously heard of that I am not seeing what you are getting at. I suspect you think the ngdp gap from target would also suffice. You used to hear about a misery index of inflation plus unemployment. How about asking them to hold down the misery index?

  16. Gravatar of Paul Zrimsek Paul Zrimsek
    3. November 2011 at 15:46

    Bad: The book and movie will be called Moneymoney.

    Good: The fan organization that will spring up to apply our monetary Bill James’ insights will be called the Society for Monetary Neutrality Research, and will lend its initials to the new discipline of sumnermetrics.

  17. Gravatar of Ben Ben
    3. November 2011 at 17:52

    “PPPS. Some statistics nerds tried to go beyond Bill James, and argue that OBP and SP shouldn’t have equal weights. People; you need to realize that there is beauty in simplicity.”

    Nope. I agree with the stat nerds here. The book Moneyball even admits that OBP and SLG shouldn’t have equal weights. You can have simplicity in something like wOBA which just uses a linear regression to estimate the weights and gives a simple, single number for us to use.

    /Used to arguing baseball.

  18. Gravatar of Jim Glass Jim Glass
    3. November 2011 at 18:32

    My nominees for the villain in the upcoming smash hit NGDPball…

    Do not omit Larry “Darth” Summers.

  19. Gravatar of anon/portly anon/portly
    3. November 2011 at 23:47

    Hey! You stole my idea. I’ve already been working for weeks on the screenplay for The Money Illusion. (No other title will do). It will of course be a Goodfellas quasi-documentary style, with lots of voiceover and flashbacks.

    Most scenes will begin in the classroom of “Scott,” a mild-mannered professor who delivers a lecture which dissolves into flashback, action and dream sequences, etc. Scott is constantly interupted by student (and the film’s real star) “Morgan,” who interjects colorful and hilarious points about American electoral and political realities. Outside the classroom, Scott must fend off the bombastic criticisms and witticisms of mysterious department chair (and film’s other real star) “Brad,” who apparently in the past has worked at the top levels of both the US Government and Starbucks.

    The lengthiest sequences, and advertising campaign focus, will involve helicopters, thousands and thousands of helicopters. In the film’s dramatic arc, Morgan at first leads the vigilantes (led by Brad) shooting down the helicopters but later sees the light and is selected (by Brad again) to fly the lead helicopter in the mission that turns out to save the US economy.

    Occasionally the films dissolves or “clicks” over to the adjoining classroom of fellow professor “Nick,” who has only one student and whose pronunciations on each day’s “topic” provides a zen-like counter-focus to the more grounded revelations of Scott.

    In the film’s final scene, Scott, on vacation, visits the bell-tower at Mission San Juan Bautista and discovers that monetary policy was really loose all along.

    Dream casting:

    Morgan – some big name star
    Brad – another big name star
    Scott – anybody really, but if Ricky Gervais or Steve Coogan will do it cheap enough, why not?
    Nick – Dave Thomas
    Nick’s Student – Rick Moranis
    Other faculty members – Irene Jacob, Emmanuelle Beart, Maggie Cheung, Gong Li, Julie Delpy – (showing my age, yes)
    Fed Chair – Morgan Freeman
    President of the ECB – Morgan Freeman
    BOJ Governor – Morgan Freeman

    I’m out of cheap jokes, sorry.

  20. Gravatar of anon/portly anon/portly
    4. November 2011 at 00:19

    About Bill James, of course his single stat was not OPS, but Runs Created, which in its simplest form was (TOB * TB)/PA where TOB is times on base.

    I always thought it was cool that it correlated so well with actual runs scored. Simplifying a bit, (TOB * TB) = (W + S + D + T + H)(S + 2D + 3T + 4H) = WS + 2WD + 3WT + 4WH + S^2 + 3SD + 4ST + 5SH + 2D^2 + 5DT + 6DH + 3T^2 + 7TH + 4H^2. Just weight each term with 1/PA and you’ve got a good fit with runs scored.

  21. Gravatar of anon/portly anon/portly
    4. November 2011 at 01:05

    “The sum of on base percentage and slugging percentage picks up the dual mandate; get on base, and drive in runs.”

    Er, isn’t the real insight of advanced batting statistics at least as much about the avoidance of making outs? How does that fit with Market Monetarist ideas? Where is Nick Rowe when you need him?

  22. Gravatar of Paul Zrimsek Paul Zrimsek
    4. November 2011 at 03:18

    Bill James needs to get up-to-date with the Age of Obama and change his formula to “Runs Created or Saved”.

  23. Gravatar of Adam Adam
    4. November 2011 at 12:05

    Gloating already?

  24. Gravatar of W. Peden W. Peden
    4. November 2011 at 12:21

    Adam,

    It’s much easier to gloat before things go wrong in the application, as they always do to some degree!

  25. Gravatar of ssumner ssumner
    4. November 2011 at 17:46

    Cthorm, Who is your boss?

    Gregor, All good choices.

    Thanks David.

    Patrick, Why subtract homeruns?

    John, Glad to finally earn some money off you.

    Neal, Yes, I forgot.

    Caveat emptor, ???

    Benny, Another good villain.

    Mattmaison. I agree.

    Benjamin, Bill James says base stealing is overrated.

    OneEyedMan, No, I was thinking of NGDP.

    Ben, Don’t ruin all the fun.

    Jim Glass, Another good villain.

    Paul, Given my skills in statistics, Sumnermetrics is an oxymoron.

    anon/portly, Yes, Morgan must have a leading role. But are there any Hollywood actors today that have the chops to play that role?

    You said;

    “Other faculty members – Irene Jacob, Emmanuelle Beart, Maggie Cheung, Gong Li, Julie Delpy – (showing my age, yes)”

    Those are some of my favorite actresses.

    You are right about the formula–it’s been years since I read James, so I forgot.

    I can’t think of an analogy for avoiding outs, but perhaps Nick can.

    Paul, Yes, runs saved.

    Adam, I suppose I am gloating, but I couldn’t resist the analogy. Don’t worry, people who gloat always get their comeuppance. Mine is coming.

  26. Gravatar of Browsing Catharsis – 11.05.11 « Increasing Marginal Utility Browsing Catharsis – 11.05.11 « Increasing Marginal Utility
    5. November 2011 at 04:10

    […] Scott Sumner argues: target NGDP just like you would target On-Base Plus Slugging Percentage in base…. Wait, what? […]

  27. Gravatar of caveat bettor caveat bettor
    7. November 2011 at 12:04

    Scott, I was talking about risk-adjusted returns. William Sharpe originally proposed his eponymous ratio, then proposed the Information Ratio to not reward closet-indexing (which the Sharpe Ratio did reward). We want to measure “good” NGDP, right, which seemed to be similar to “alpha” in the CAPM. We don’t want NGDP to move around wildly, which seemed to be similar to “volatility” in Sharpe’s metrics. Finally, we can only express alpha relative to an appropriate benchmark. I through out Canadian NGDP just to get things started.

  28. Gravatar of caveat bettor caveat bettor
    7. November 2011 at 12:04

    Scott, I was talking about risk-adjusted returns. William Sharpe originally proposed his eponymous ratio, then proposed the Information Ratio to not reward closet-indexing (which the Sharpe Ratio did reward). We want to measure “good” NGDP, right, which seemed to be similar to “alpha” in the CAPM. We don’t want NGDP to move around wildly, which seemed to be similar to “volatility” in Sharpe’s metrics. Finally, we can only express alpha relative to an appropriate benchmark. I threw out Canadian NGDP just to get things started.

  29. Gravatar of ssumner ssumner
    10. November 2011 at 19:52

    Caveat bettor, I’m afraid I don’t know enough finance to respond. What is closet indexing?

  30. Gravatar of caveat bettor caveat bettor
    11. November 2011 at 07:57

    When managers correlate their portfolios to their benchmark to game the Sharpe ratio, which is fooled by correlation more than the Information ratio. As you know, once we win the victory of NGDP, there will be public choice incentives to start gaming NGDP measures. So we will need to stay vigilant even after we win.

  31. Gravatar of Scott Sumner Scott Sumner
    13. November 2011 at 06:41

    Caveat better, yes, but I’m not particularly worried. If one thinks about the flaws in inflation targeting, I’d put “gaming the CPI” pretty far down on the list.

  32. Gravatar of Scott Sumner Scott Sumner
    13. November 2011 at 06:41

    I meant caveat bettor

Leave a Reply