We've been waiting for nine months to get a look at this
year's team. The opening game will tell us a lot about what
kind of team we're going to have this year. We all know how
important it is to get off to a good start. A win will get
us on the right track and will set the stage for a successful
season. Right?
Uh, maybe not.
It turns out that a victory in the opening game is not a very
good predictor of success. In fact, recent statistics tell
us that we are better off losing our opener.
I know that seems counter-intuitive. You would expect a
good team to be more likely to win its opening game, while a bad
team would be more likely to lose. Over the long run, you
would think that there should be some correlation between the
result in the first game and the result for the season.
That once was true. I looked at the last 50 years of
Stanford football, starting with Chuck Taylor's first season as
coach in 1951. From 1951 through 1971, there was a strong
correlation between the opening game result and the success of
the season. Stanford had a record of 15-6 in its openers
during that period. In the 15 seasons that Stanford won its
opening game, Stanford had a combined winning percentage of .581,
with 11 seasons at .500 or better and only 4 losing seasons.
In three of those seasons, Stanford went to the Rose Bowl.
On the other hand, in the 6 seasons that Stanford started
off with a loss, Stanford had a combined winning percentage of
only .358, with only one season above .500, no seasons with more
than 6 wins, and no bowl appearances.
Winning the first game also was a good predictor of success in
the Big Game. During the 1951-71 time period, when Stanford
won its opener, Stanford had a 10-5 record in the Big Game.
When Stanford lost its opening game, its Big Game record
was only 1-4-1.
But things have changed. From 1972 (Jack Christiansen's
first season as coach) through the present, the result of the
first game has had no relationship to Stanford's final record.
From 1972 through 2000, Stanford has had a opening game
record of 8-20-1. In the seasons Stanford opened with a
win, Stanford's combined winning percentage was .517, with 4
winning seasons and 4 losing seasons. In the years Stanford
started off with a loss, Stanford's combined winning percentage
was .507, with 10 winning seasons, 9 losing seasons, and 1 season
at .500. Those records are virtually indistinguishable.
In fact, from 1972 to the present, Stanford actually has
done better in the Big Game in seasons that started with a loss:
Stanford is 4-4 in the Big Game in years that started with
a win, and 15-4-1 in seasons that started with a loss.
Looking at the more recent past, the overall trend runs exactly
opposite to expectations. In the last 10 seasons (1991
through 2000), Stanford has done better when it loses its first
game than when it wins the opener. In the three seasons
Stanford won its first game, its combined winning percentage was
.515, with 1 winning season and 2 losing seasons. But in
the seasons Stanford opened with a loss, Stanford's combined
winning percentage was .563, with 4 winning seasons and 2 losing
seasons. Since 1991, 4 of Stanford's 5 bowl games
(including the Rose Bowl) have come in seasons where Stanford
lost its opening game. In fact, in seasons where Stanford
lost its first game, it has won 2 more games during the remainder
of the season than in seasons where it won its first game.
In the Big Game, Stanford has dominated since 1991 regardless of
whether it wins or loses its first game: Stanford is 3-0 in
the Big Game in seasons that opened with a win, and 5-1 in
seasons that opened with a loss. (Stanford opened with a
tie once, losing the Big Game that year.)
I'm not saying we should hope Stanford loses this Saturday.
But the stats tell us that a loss is not a disaster.
In fact, over the last decade, a loss has been more likely
than a win to signal a good season. We all know Stanford's
Rose Bowl two years ago came in a season that started with a
loss. It turns out that this is more the rule than the
exception, at least in recent history. . . .