Expansion Effect Revisited Essay

Last year, in an article entitled “Expansion Effect,” I showed that winning teams win more in expansion years. Now I am going to analyze whether Babe Ruth (1927) or Mark McGwire (1998) hit more “home runs” when the Expansion Effect is considered.

I first started by looking in The Home Run Encyclopedia (a great SABR volume) and discovered that the authors had conveniently listed the top five home run hitters annually by league. However, the book ended at the 1995 season. An e-mail to one of the authors quickly yielded information for 1997 and 1998. Thank you, David Vincent.

Since Ruth (1927) and McGwire (1998) each led his league in homers, I felt an analysis of the top five in each league during expansion years as compared to the year before would be fair. In other words, no matter the players’ names, the first spot’s number of home runs produced would be compared to the year before expansion. The second spot, the same analysis, and similarly through fifth spot. See Table 1.

Table 1. Year Before vs. Year of Expansion.1960 AL 1961 AL

40 (42.1) 61 Roger +21 (+18.9)
Mickey Mantle Maris

39 (41.0) 54 Mickey +15 (+13.0)
Roger Marls Mantle

38 (40.0) 46 Jim +8 (+6.0)
Jim Lemon Gentile

35 (36.8) 46 Harmon +11 (+9.2)
Rocky Colavito Killebrew

31 (32.6) 45 Rocky +14 (+12.4)
Harmon Killebrew Colavito
+69 (+59.5)

1961 NL 1962 NL

46 Orlando Cepeda 49 Willie Mays +3
40 Willie Mays 45 Hank Aaron +5
37 Frank Robinson 39 Frank Robinson +2
35 Joe Adcock 37 Ernie Banks +2
35 Dick Stuart 35 Orlando Cepeda 0
+12

1968 AL 1969 AL

44 Frank Howard 49 Harmon Killebrew +5
36 Willie Horton 48 Frank Howard +12
35 Ken Harrelson 47 Reggie Jackson +12
29 Reggie Jackson 40 Rico Petrocelli +11
25 many tied 40 Carl Yastrzemski +15
+55

1968 NL 1969 NL

36 Willie McCovey 45 Willie McCovey +9
33 Dick Allen 44 Hank Aaron +11
32 Ernie Banks 38 Lee May +6
30 Billy Williams 37 Tony Perez +7
29 Hank Aaron 33 Jim Wynn +4
+37

1976 AL 1977 AL

32 Graig Nettles 39 Jim Rice +7
27 Sal Bando 37 Bobby Bonds +10
27 Reggie Jackson 37 Graig Nettles +10
25 many tied 33 George Scott +8
25 many tied 32 Reggie Jackson +7
+42

1992 NL 1993 NL

35 Fred McGriff 46 Barry Bonds +11
34 Barry Bonds 40 Dave Justice +6
33 Gary Sheffield 38 Matt Williams +5
27 Darren Daulton 37 Fred McGriff +10
27 Dave Hollins 36 Ron Gant +9
+41

1997 AL 1998 AL

56 Ken Griffey 56 Ken Griffey 0
44 Tino Martinez 49 Albert Belle +5
42 Juan Gonzalez 46 Jose Canseco +4
40 Jay Buhner 45 Juan Gonzalez +5
40 Jim Thome 45 Manny Ramirez +5
+19

1997 NL 1998 NL

58 Mark McGwire(*) 70 Mark McGwire +12
49 Larry Walker 66 Sammy Sosa +17
43 Jeff Bagwell 50 Greg Vaughn +7
41 Andres Galarraga 46 Vinny Castilla +5
40 many tied 44 Andres Galarraga +4
+45

 

(*) McGwire actually hit 34 in the AL and 24 in the NL. To avoid really complicated calculations, his season total of 58 appears in the NL column. Statistically, AL or NL placement yields no difference in final totals or averages.

Also, when “home run” appears in quotes, I am referring to home runs adjusted for statistical analysis in one way or another.

The 1960-61 American League expansion required an extra adjustment because eight games were added in 1961. I multiplied the old 154-game results by 1.052 to produce the statistical comparison for a 162-game schedule. The results are in parentheses. Likewise, the statistically adjusted differences are in parentheses. For example, Mickey Mantle led the American League in 1960 with 40 home runs. Roger Maris led the American League in 1961 with 61 home runs. Mantle’s 40 was adjusted to 42.1 for a 162-game schedule. Therefore, the American League’s top home run hitter hit 18.9 more “home runs” in the expansion year than the year before.

One truly significant surprise is that not a single home run hitter in the top five hit fewer home runs in an expansion year than the corresponding place the year before. Only twice, 1961-62 National League fifth place and 1997-98 American League first place, did the home run totals equal. The Expansion Effect seems to be powerful.

I totaled the data and summarized it in Table 2. The final analysis showed that the average top five home run hitters hit 7.76 more home runs in an expansion year. Surely that is not just significant–it is mind-boggling.Table 2. Summaries of Increase/Decrease.

Total Average

60-61 AL +59.5 +11.90
61-62 NL +12.0 +2.40
68-69 AL +55.0 +11.00
68-69 NL +37.0 +7.40
76-77 AL +42.0 +8.40
92-93 NL +41.0 +8.20
97-98 AL +19.0 +3.80
97-98 NL +45.0 +9.00
+310.5 +7.76

 

As sort of a “control group,” I analyzed the similar data for the league that did not expand during an expansion year. See Table 3.

Table 3 Control group–no expansion1960 NL 1961 NL

41 (43.1) Ernie Banks 46 Orlando Cepeda +5 (+2.9)
40 (42.1) Hank Aaron 40 Willie Mays 0 (-2.1)
39 (41.0) Eddie Mathews 37 Frank Robinson -2 (-4.0)
32 (33.7) Ken Boyer 35 Joe Adcock +3 (+1.3)
31 (32.6) Frank Robinson 35 Dick Stuart +4 (+2.4)
+10 (+0.5)

1961 AL 1962 AL

61 Roger Maris 48 Harmon Killebrew -13
54 Mickey Mantle 39 Norm Cash -15
47 Jim Gentile 37 Rocky Colavito -9
46 Harmon Killebrew 37 Leon Wagner -9
45 Rocky Colavito 33 many tied -8
-54

1976 NL 1977 NL

38 Mike Schmidt 52 George Foster +14
37 Dave Kingman 41 Jeff Burroughs +4
32 Rick Monday 39 Greg Luzinski +7
29 George Foster 38 Mike Schmidt +9
27 Joe Morgan 33 Steve Garvey +6
+40

1992 AL 1993 AL

43 Juan Gonzalez 46 Juan Gonzalez +3
42 Mark McGwire 45 Ken Griffey +3
35 Cecil Fielder 41 Frank Thomas +6
34 Albert Belle 38 Albert Belle +4
34 Joe Carter 37 Rafael Palmeiro +3
+19

 

Table 4 summarizes the non-expanding league results. The final average turned out to be a tiny and insignificant +0.225 per home run hitter. I think this is a second proof that the Expansion Effect is powerful. There is no doubt that good players and good teams post significantly better statistical numbers in expansion years.

Table 4 Corresponding League Summaries Total Average

60-61 NL +0.5 +0.14 +0.1
61-62 AL -54.0 -10.80 -10.8
76-77 NL +40.0 +8.00 +8.0
92-93 AL +19.0 +3.80 +3.8
+00.225

 

What’s the result when we apply the simple schedule and Expansion Effect calculations to Babe Ruth’s 1927 and Mark McGwire’s 1998?

Ruth hit 60 home runs in 1927. McGwire hit 70 home runs in 1998. Adjusting Ruth’s 60 by the multiplier 1.052 for the longer schedule yields 63.12 “home runs.” Adjusting McGwire’s 70 for the Expansion Effect by subtracting 7.76 homers yields 62.24 “home runs.” See Table 5. The contest is a virtual photo finish.

Table 5 Adjustment table for McGwire and RuthRuth (1927)

60 (in 154-game schedule)
x 1.052 (eight-game adjustment)
63.12 adjusted “home runs”

McGwire (1998)

70 (in 162-game schedule)
-7.76 (Expansion Effect)
62.24 adjusted “home runs”