Among the many things he brought us, Bill James came up with the concept of Pythagorean won-lost records. In case you don't know, Bill found that a team's winning percentage could be very accurately predicted simply using their runs scored and runs allowed totals. The reason he called it the Pythagorean method is that, in its simplest form, the relationship between runs scored, runs allowed, and winning percentage could be stated using the squares of runs scored and runs allowed:
Win Pct.= runs scored2/runs scored2 + runs allowed2
This equation accurately predicts a team's winning percentage. However, not everyone believes it. When I worked for the Orioles, I was explaining this to the baseball staff in an organizational meeting and one scout stood up and said, "you can't do that. Some teams score runs when they need them and others score them when it doesn't matter." I told him that this relationship was one of the fundamental laws of the baseball universe and saying it doesn't work was like saying you can't measure temperature with a thermometer.
Pythagoras had a pretty good year in 2000, especially in the American League. Take a look:
|
The Pythagorean Method
|
|
Team
|
LG
|
W
|
L
|
Pct
|
PyW
|
PyL
|
Py%
|
RS
|
RA
|
WDiff
|
|
Tor.
|
AL
|
83
|
79
|
.512
|
77
|
85
|
.476
|
861
|
908
|
6
|
|
Bal.
|
AL
|
74
|
88
|
.457
|
71
|
91
|
.436
|
794
|
913
|
3
|
|
ChW
|
AL
|
95
|
67
|
.586
|
92
|
70
|
.570
|
978
|
839
|
3
|
|
NYY
|
AL
|
87
|
74
|
.540
|
85
|
76
|
.531
|
871
|
814
|
2
|
|
Ana.
|
AL
|
82
|
80
|
.506
|
81
|
81
|
.497
|
864
|
869
|
1
|
|
Tex.
|
AL
|
71
|
91
|
.438
|
71
|
91
|
.437
|
848
|
974
|
0
|
|
K.C.
|
AL
|
77
|
85
|
.475
|
77
|
85
|
.474
|
879
|
930
|
0
|
|
Min.
|
AL
|
69
|
93
|
.426
|
69
|
93
|
.426
|
748
|
880
|
0
|
|
Bos.
|
AL
|
85
|
77
|
.525
|
86
|
76
|
.528
|
792
|
745
|
-1
|
|
Oak.
|
AL
|
91
|
70
|
.565
|
92
|
69
|
.569
|
947
|
813
|
-1
|
|
Sea.
|
AL
|
91
|
71
|
.562
|
92
|
70
|
.569
|
907
|
780
|
-1
|
|
Tam.
|
AL
|
69
|
92
|
.429
|
70
|
91
|
.437
|
733
|
842
|
-1
|
|
Det.
|
AL
|
79
|
83
|
.488
|
81
|
81
|
.498
|
823
|
827
|
-2
|
|
Cle.
|
AL
|
90
|
72
|
.556
|
92
|
70
|
.569
|
950
|
816
|
-2
|
|
NYM
|
NL
|
94
|
68
|
.580
|
88
|
74
|
.541
|
807
|
738
|
6
|
|
Fla.
|
NL
|
79
|
82
|
.491
|
74
|
87
|
.461
|
731
|
797
|
5
|
|
Atl.
|
NL
|
95
|
67
|
.586
|
90
|
72
|
.557
|
810
|
714
|
5
|
|
St.L.
|
NL
|
95
|
67
|
.586
|
91
|
71
|
.564
|
887
|
771
|
4
|
|
S.D.
|
NL
|
76
|
86
|
.469
|
75
|
87
|
.463
|
752
|
815
|
1
|
|
Mon.
|
NL
|
67
|
95
|
.414
|
66
|
96
|
.409
|
738
|
902
|
1
|
|
S.F.
|
NL
|
97
|
65
|
.599
|
97
|
65
|
.597
|
925
|
747
|
0
|
|
Ari.
|
NL
|
85
|
77
|
.525
|
85
|
77
|
.522
|
792
|
754
|
0
|
|
Mil.
|
NL
|
73
|
89
|
.451
|
73
|
89
|
.450
|
740
|
826
|
0
|
|
Cin.
|
NL
|
85
|
77
|
.525
|
87
|
75
|
.534
|
825
|
765
|
-2
|
|
L.A.
|
NL
|
86
|
76
|
.531
|
88
|
74
|
.541
|
798
|
729
|
-2
|
|
ChC
|
NL
|
65
|
97
|
.401
|
69
|
93
|
.424
|
764
|
904
|
-4
|
|
Pit.
|
NL
|
69
|
93
|
.426
|
73
|
89
|
.448
|
793
|
888
|
-4
|
|
Phi.
|
NL
|
65
|
97
|
.401
|
69
|
93
|
.428
|
708
|
830
|
-4
|
|
Col.
|
NL
|
82
|
80
|
.506
|
87
|
75
|
.535
|
968
|
897
|
-5
|
|
Hou.
|
NL
|
72
|
90
|
.444
|
81
|
81
|
.497
|
938
|
944
|
-9
|
In the American League, only one team finished more than three wins away from their Pythagorean record. (Just so you know, I used the slightly more accurate method which raises runs scored and runs allowed to the power of 1.83 instead of two. With a computer, changing the exponent is no big deal.) Overall, 18 of 30 teams finished within two wins of their Pythagorean record. It looks like the thermometer still works.
Eddie Epstein, a former front-office executive with the Orioles and Padres, now works as a senior sports analyst for PricewaterhouseCoopers. He can be reached at eddie.epstein@us.pwcglobal.com. | |
|
