1
00:00:09,840 --> 00:00:12,480
Today's lecture is a
story of Moneyball,

2
00:00:12,480 --> 00:00:17,470
a book by Michael Lewis in
2003 and a movie in 2011

3
00:00:17,470 --> 00:00:20,030
starring Brad Pitt.

4
00:00:20,030 --> 00:00:25,280
Moneyball discusses how sports
analytics changed baseball.

5
00:00:25,280 --> 00:00:28,130
Moneyball tells the
story of the Oakland A's.

6
00:00:28,130 --> 00:00:32,820
The A's is a team near
San Francisco, California.

7
00:00:32,820 --> 00:00:35,550
They were once a rich
team, but the team

8
00:00:35,550 --> 00:00:39,660
was purchased in
1995 by owners who

9
00:00:39,660 --> 00:00:42,450
enforced strict budget cuts.

10
00:00:42,450 --> 00:00:44,720
Despite this, they
were improving

11
00:00:44,720 --> 00:00:48,620
over the years 1997 to 2001.

12
00:00:48,620 --> 00:00:52,710
In the table, you can see
that the percentage of wins

13
00:00:52,710 --> 00:00:55,430
was increasing in
each and every year.

14
00:00:58,420 --> 00:01:01,700
This was puzzling, and
many baseball experts

15
00:01:01,700 --> 00:01:03,980
thought it was just luck.

16
00:01:03,980 --> 00:01:07,510
In 2002, the A's lost
three key players.

17
00:01:07,510 --> 00:01:11,320
The key question: could they
continue winning without them?

18
00:01:11,320 --> 00:01:13,980
So what is the key problem?

19
00:01:13,980 --> 00:01:18,260
Let us discuss the graph
on the left of the screen.

20
00:01:18,260 --> 00:01:26,330
The horizontal axis
shows the average payroll

21
00:01:26,330 --> 00:01:29,970
during the years 1998 to 2001.

22
00:01:29,970 --> 00:01:35,450
The vertical axis shows
the average yearly wins

23
00:01:35,450 --> 00:01:37,590
over the same years.

24
00:01:37,590 --> 00:01:40,440
So let's look at some of
the teams in this graph.

25
00:01:40,440 --> 00:01:43,580
So which one is this team?

26
00:01:43,580 --> 00:01:48,320
This is a team that
won about 100 games

27
00:01:48,320 --> 00:01:52,310
and spent roughly $90
million during this period.

28
00:01:52,310 --> 00:01:53,640
So this is the New York Yankees.

29
00:01:56,660 --> 00:01:59,070
Let's look at this team.

30
00:01:59,070 --> 00:02:05,160
This team spent about $80
million and won about 90 games.

31
00:02:05,160 --> 00:02:06,200
This is the Red Sox.

32
00:02:11,200 --> 00:02:13,850
Where are the Oakland A's?

33
00:02:13,850 --> 00:02:15,470
The A's are here.

34
00:02:18,500 --> 00:02:24,640
They won about 90 games, and
they spent under $30 million.

35
00:02:24,640 --> 00:02:28,290
If you compare it
with the Red Sox,

36
00:02:28,290 --> 00:02:32,700
they won about the same number
of games during this period

37
00:02:32,700 --> 00:02:39,000
but the Red Sox spent about
$50 million more per year than

38
00:02:39,000 --> 00:02:39,500
the A's.

39
00:02:42,710 --> 00:02:46,630
Clearly, rich teams like
the Yankees and the Red Sox

40
00:02:46,630 --> 00:02:50,440
can afford the all-star players.

41
00:02:50,440 --> 00:02:53,670
But please observe how
efficient the A's are.

42
00:02:53,670 --> 00:02:56,530
As I mentioned,
they won 90 games,

43
00:02:56,530 --> 00:02:59,280
and their payroll
was under $30 million

44
00:02:59,280 --> 00:03:03,520
compared to the Yankees,
who spent almost three

45
00:03:03,520 --> 00:03:07,540
to four times as much,
and they won, of course,

46
00:03:07,540 --> 00:03:10,800
more games, but
not that much more.

47
00:03:10,800 --> 00:03:12,830
So rich teams, as
I mentioned, have

48
00:03:12,830 --> 00:03:18,300
three to four times the
payroll of poor teams.

49
00:03:18,300 --> 00:03:25,180
Yet the A's made the
playoffs every year.

50
00:03:25,180 --> 00:03:27,290
How do they do this?

51
00:03:27,290 --> 00:03:29,970
And what we will see
in this lecture is

52
00:03:29,970 --> 00:03:31,940
that by taking a
quantitative approach,

53
00:03:31,940 --> 00:03:37,420
an analytics approach, they
were able to find undervalued

54
00:03:37,420 --> 00:03:43,040
players and form teams
that were very efficient.

55
00:03:43,040 --> 00:03:45,880
So the A's started
using a different method

56
00:03:45,880 --> 00:03:48,630
to select players.

57
00:03:48,630 --> 00:03:53,010
The traditional way of selecting
players was through scouting.

58
00:03:53,010 --> 00:03:56,350
Scouts would watch high
school and college players,

59
00:03:56,350 --> 00:03:58,840
and they would report
back about their skills,

60
00:03:58,840 --> 00:04:03,390
especially discussing their
speed and their athletic build.

61
00:04:03,390 --> 00:04:05,880
The A's, however,
selected players

62
00:04:05,880 --> 00:04:08,550
based on their statistics,
not on their looks.

63
00:04:11,350 --> 00:04:14,490
The following are quotes
from the book Moneyball.

64
00:04:14,490 --> 00:04:17,360
"The statistics enable
you to find your way

65
00:04:17,360 --> 00:04:21,390
past all sorts of sight-based
scouting prejudices."

66
00:04:21,390 --> 00:04:23,410
And a direct quote
from Billy Beane,

67
00:04:23,410 --> 00:04:26,890
the manager of the Oakland
A's and the architect

68
00:04:26,890 --> 00:04:34,409
of this approach: "We are
not selling jeans here."

69
00:04:34,409 --> 00:04:37,700
Let us contrast how
the A's selected

70
00:04:37,700 --> 00:04:40,480
players versus the Yankees.

71
00:04:40,480 --> 00:04:46,080
On the left, you see a
catcher, Scott Hatteberg,

72
00:04:46,080 --> 00:04:51,090
that the A's selected, who would
not throw particularly well

73
00:04:51,090 --> 00:04:53,300
but got on base a lot.

74
00:04:53,300 --> 00:04:56,700
On the right, you
see Derek Jeter,

75
00:04:56,700 --> 00:05:00,250
one of the top
players in baseball,

76
00:05:00,250 --> 00:05:04,000
a consistent shortstop
and the leader

77
00:05:04,000 --> 00:05:05,340
in hits and stolen bases.

78
00:05:08,900 --> 00:05:10,930
Let us look into pitchers.

79
00:05:10,930 --> 00:05:15,030
On the left, you
see Chad Bradford,

80
00:05:15,030 --> 00:05:18,210
a pitcher for the
A's, a submariner

81
00:05:18,210 --> 00:05:23,640
who used an unconventional
delivery and slow speed.

82
00:05:23,640 --> 00:05:25,980
On the right, you
see Roger Clemens,

83
00:05:25,980 --> 00:05:28,650
one of the best pitchers
in the game that

84
00:05:28,650 --> 00:05:34,170
used conventional
delivery and fast speed.

85
00:05:34,170 --> 00:05:38,970
Billy Beane was the manager
of the Oakland A's since 1997.

86
00:05:38,970 --> 00:05:42,820
He played Major League Baseball,
but he was not a great player.

87
00:05:42,820 --> 00:05:47,420
In fact, he sees himself as
a typical scouting error.

88
00:05:47,420 --> 00:05:53,500
In the 1980s and 1990s, analysts
were hired by baseball teams,

89
00:05:53,500 --> 00:05:56,340
but none of them
had enough power

90
00:05:56,340 --> 00:06:00,410
to affect anything important.

91
00:06:00,410 --> 00:06:02,440
As I mentioned,
Billy Beane became

92
00:06:02,440 --> 00:06:05,680
the general manager of
the Oakland A's in 1997,

93
00:06:05,680 --> 00:06:08,930
and he was given a
rather small budget.

94
00:06:08,930 --> 00:06:13,160
He understood the
importance of analytics.

95
00:06:13,160 --> 00:06:16,180
But during this period,
most general managers

96
00:06:16,180 --> 00:06:20,490
knew little about statistics
and based decisions primarily

97
00:06:20,490 --> 00:06:22,980
on feelings.

98
00:06:22,980 --> 00:06:26,770
In contrast, Billy Beane hired
Paul DePodesta, a Harvard

99
00:06:26,770 --> 00:06:30,240
graduate, as his assistant.

100
00:06:30,240 --> 00:06:35,520
Furthermore, he was not afraid
to alienate scouts, managers,

101
00:06:35,520 --> 00:06:41,320
and players if the quantitative
approach suggested decisions

102
00:06:41,320 --> 00:06:45,060
that were different than
the scouts or the managers

103
00:06:45,060 --> 00:06:46,220
or the players suggested.

104
00:06:49,610 --> 00:06:55,900
An important player
in this story

105
00:06:55,900 --> 00:06:58,870
is Paul DePodesta,
a Harvard graduate.

106
00:06:58,870 --> 00:07:02,780
Paul spent a lot of time
looking at the data.

107
00:07:02,780 --> 00:07:08,600
The analysis suggested that
some skills were undervalued

108
00:07:08,600 --> 00:07:11,840
and some skills were overvalued.

109
00:07:11,840 --> 00:07:14,350
The key premise of
the Oakland A's is

110
00:07:14,350 --> 00:07:17,810
that if they could detect
the undervalued skills,

111
00:07:17,810 --> 00:07:20,980
they could find
players at a bargain.