The Baseball Economist: The Real Game Exposed (46 page)

Endnotes

1. Thomas Carlyle, “Occasional Discourse on the Negro Question,”
Fraser’s Magazine for Town and Country,
December 1849, Vol. 40: 670–679. For a history on the role of economists in the nineteenth-century debates on race and slavery, see
How the Dismal Science Got Its Name: Classical Economics and the Ur-Text of Racial Politics
by David M. Levy.

2. See Stephen Jay Gould’s section “The Model Batter: The Extinction of .400 Hitting and the Improvement of Baseball” in
Full House,
p. 78.

3. Quote from Ken Rosenthal, “Mets Get Shot with Mighty Clemens at the Bat,” in
The Sporting News,
June 13, 2002.

4. Quote from Ken Rosenthal, “Mets Get Shot with Mighty Clemens at the Bat,” in
The Sporting News
, June 13, 2002.

5. Baker’s comments came after an interleague matchup between the Giants and Yankees, in which Roger Clemens plunked Barry Bonds after hinting he would do so before the game. Baker referred to Clemens as “Roger the Dodger.” Though this idea has long been a part of traditional baseball wisdom, three economists—Brian Goff, Bill Shughart, and Bob Tollison—first proposed and tested the theory in a 1997 article in the economics journal
Economic Inquiry
.

6. Economists Akihiko Kawaura and Sumner La Croix (2002) identified this similarity in the Japanese game.

7. This hypothesis was offered in two separate papers: one by Greg Trandel, Larry White, and Peter Klein (1998); the other by Steven Levitt (1998). Both papers were published as responses to the Goff, Shughart, and Tollison (1997) paper discussed in footnote 4.

8. “Teams Touchy After Incident Sunday,”
ESPN.com
, August 27, 2004, and Retroseet game-logs (
http//www.retrosheet.org/boxesetc/VHOU02004.htm
).

9. This data, as well as the game-by-game data discussed in the chapters, are available at Retrosheet (
www.retrosheet.org
).

10.
The Armchair Economist
, p. 3.

11. Quote from Steve Henson, “Win Is Big in Grand Scheme,” in
The Los Angeles Times,
April 25, 2006.

12. Jason Stark, “The Book on Hooks,”
www.espn.com
, May 10, 2004.

13. This calculation is made from the averages of all of the statistics. One hundred points of OPS is about a one standard deviation difference in OPS.

14.
The Economist as Preacher and Other Essays,
p. 10

15. All data on handedness comes from
The Lahman Baseball Database 5.2
(
www.baseball1.com
).

16. This list actually overcounts the lefties, because the old rules of the game only required players to be listed at a position in the lineup, not to play in the game. So some of these lefties never saw any playing time, though official baseball rules credit them with a game played at that position.

17. While other tables exist with some minor differences, the Lindsey tables will be fine for this example. See
The Numbers Game
by Alan Schwarz (2004) for a description of this data-gathering project. For an excellent description of how these tables can be used, see
Curve Ball
by James Albert and Jay Bennett (2001).

18. Why are steals of third base rare? Table 37 shows the “breakeven” points, where the expected gains equal the expected costs, for stealing third base at each base/out configuration. They are expressed as the percentage of stealing success needed to offset the lost runner and out if the runner is caught. As you can see, a runner needs to have a pretty high success rate to justify swiping the bag. For example, with no outs and a runner on second base, a runner needs to be successful more than 83 percent of the time to justify the risk of stealing. Why is such a high success rate necessary? According to Table 4, the runner risks an expected 1.194 runs (if he just stays on second base) to increase the team’s expected runs by 0.196 (1.39 − 1.194 = 0.196). If he fails, he loses 0.96 expected runs (0.234 − 1.194 = −0.96). Wagering nearly one run to gain 0.2 runs is a perilous strategy. Because the potential gains are so low and the losses are quite high, it’s only worth taking the risk to steal third if a player can make it safely more than 83 percent of the time in this circumstance. The average breakeven point for all of the base/out states is 67 percent, so it’s no surprise that teams are reluctant to steal third in almost every situation.

19. Right-handers may throw out some runners lefties will miss, but there will be plenty that lefties will get too. Plus, runners will not attempt to steal with every
opportunity. Furthermore, some lefties have stronger arms than right-handers, and it is possible for the strong arm of a lefty to compensate for the throwing-angle disadvantage. Therefore, if the benefits from being right-handed at catcher are real, they have to be small.

20.
The New Bill James Historical Abstract,
2003, p. 41.

21. Quote from Travis Haney, “La Russa Badmouths Braves,”
Augusta Chronicle
, May 2, 2005.

22.
Atlanta Journal-Constitution,
May 2, 2005. 23 Mazzone and Freeman, p. 196.

24. “Spahn & Sain,” Gerald V. Hern,
Boston Post
, September 14, 1948.

25. Mazzone and Freeman,
Tales from the Braves Mound,
pp.13–14.

26. Merron, “Rock of Atlanta.”
ESPN.com
E-Ticket,
September, 17–18, 2005.

27. Merron, “Rock of Atlanta.”
ESPN.com
E-Ticket,
September, 17–18, 2005.

28. I use the three-year pitcher park factor from The Lahman Baseball Database, which is based on home and away batting performances by teams (
www.baseball1.com
).

29. Mazzone and Freeman,
Tales from the Braves Mound,
pp. 14–15.

30. Mazzone and Freeman,
Tales from the Braves Mound,
p. 186.

31. Mazzone and Freeman,
Tales from the Braves Mound
, p. 190.

32. Mazzone and Freeman,
Tales from the Braves Mound,
p. 175.

33. Merron, “Rock of Atlanta.”
ESPN.com
E-Ticket,
September, 17–18, 2005.

34. Thomas Stinson, “Renowned Mileage of Retreads,”
Atlanta Journal-Constitution,
October 22, 2005. p. E1.

35. Merron, “Rock of Atlanta,”
ESPN.com
E-Ticket,
September, 17–18, 2005.

36. Thomas Stinson, “Renowned Mileage of Retreads,”
Atlanta Journal-Constitution,
October 22, 2005.

37. Gary Curtright, “Coach’s Style’s a Change-up for Pitchers”
Atlanta Journal-Constitution,
February 10, 2006.

38. Merron, “Rock of Atlanta,”
ESPN.com
E-Ticket,
September, 17–18, 2005.

39. Merron, “Rock of Atlanta,”
ESPN.com
E-Ticket,
September, 17–18, 2005.

40. Merron, “Rock of Atlanta,”
ESPN.com
E-Ticket,
September, 17–18, 2005.

41. Eric Mack,
CBS.sportsline.com
,
January 24, 2006.

42. Mark Feinsand, “Yankees Sign Jaret Wright,”
MLB.com
,
December 29, 2004.

43.
The Report of the Independent Members of the Commissioner’s Blue Ribbon Panel on Baseball Economics,
July 2000, p. 5.

44.
The Report of the Independent Members of the Commissioner’s Blue Ribbon Panel on Baseball Economics,
July 2000, p. 43.

45. I used this same method in an earlier study, “Quantifying the Market Size Advantage in MLB,” in
The Hardball Times,
March 27, 2004.

46. Posnanski, “Bill James Finally Gets Enough of Royals,”
Kansas City Star,
November 10, 2002.

47. Ritter,
The Glory of Their Times
.

48. Schmidt and Berri (2003) find some support for Gould’s hypothesis in major-league baseball.

49. Note that in the past, the new talent was not equally distributed across all clubs; however, in the last two rounds of expansion, MLB made an effort to encourage success among expansion teams. Of the four new teams, only one (Tampa Bay) has not made the playoffs. The other teams have been quite competitive, with Florida winning the World Series twice and Arizona winning it once.

50. Coeffcient of Variation of X = (Standard Deviation of X)/(Mean of X).

51. Additionally, the coefficient of variation is preferred to the standard variation because it eliminates any influence of a rising and falling of the mean values that can raise and lower the standard deviation even when the dispersion of performance does not change.

52. Interestingly, Gould’s study of the variation of batting averages ended just as the batting talent dispersion bottomed out.

53. Economist Art De Vany (forthcoming) argues that these great feats are expected given the non-normal distribution of home run hitting.

54. My use of decades to denote eras is arbitrary. I do this only because most baseball historians discuss baseball history in decades. For an example, see
The New Bill James Historical Abstract
. However, it can be useful to break eras into other time periods

55. Barro,
Getting It Right: Market Choices in a Free Society,
p. 154.

56.
Boston Globe,
January 14, 2005. It was not until 2006 that the owners and the MLBPA agreed to test and punish the use of amphetamines.

57. “Goodby to Some Old Baseball Ideas,”
LIFE
, 1954.

58. Michael Lewis,
Moneyball,
p. 128.

59. Hakes and Sauer, 2006, p. 184.

60. p. 36.

61. James and Neyer,
The Neyer/James Guide to Pitching,
p. 436.

62. In this model each player receives a 1 or 0 for each conference variable. This means that a pitcher pitching in the ACC with have a 1 for the ACC variable, while pitchers in other conferences will have the value of 0 for ACC. They will receive a 1 for their conference variable. Each player in a conference will receive the weight of his conference only, because 1 × weight = weight and 0 × weight = 0.

63. For an earlier discussion of this, see my article “Moneyball and Efficient Efficiency” in
The Hardball Times,
July 26, 2005.

64. Michael Lewis,
Moneyball,
pp. 37–38.

65. Run production based on OPS. Run prevention based on Fielding Independent Pitching (FIP), a predicted ERA based solely on strikeouts, walks, and home runs.

66. Frederic Bastiat,
That Which Is Seen, and That Which Is Not Seen
.

67. Statisticians Jim Albert and Jay Bennett find minimal evidence for clutch ability in these situations, and any observed effect is very weak (
Curve Ball,
2001).

68. I am not claiming to have discovered the importance of these statistics. The values I present mirror past findings of many sabermetricians. I am merely trying to demonstrate the relative importance of each statistic with recent data.

69.
The Hidden Game of Baseball,
1984.