Since the NCAA Tournament went to the
64-team format in 1985, thousands of fans have spent hours in front
of computer spreadsheets attempting to break down the secret to
success in “The Big Dance.” Most of these people have
failed. What seems like a good statistic in one tournament might
turn out to be meaningless the following year. So what is the best
predictor for determining how far a team will go in any given
year?
- Naweed Sikora
My girlfriend says it’s all in the team name, and is
picking Gonzaga to go all the way. I’d like to think
it’s a little more scientific than that, but, then again, she
could be right. Currently, I’m in the middle of reading
Moneyball, a book about the Oakland Athletics’ general
manager Billy Beane and how he has used a system known as
Sabremetrics to draft the right players and keep the A’s
competitive, while also keeping payroll down.
Sabremetrics is a system of statistical analysis that place
emphasis on a player’s stats — particularly on-base
percentage — rather than on a scout’s gut feeling or
the future. Beane feels that on-base percentage is the best
indicator for how well a player will perform offensively in the
league, and will not hesitate to pass on the physically-gifted
players if they carry low on-base percentages.
The system has worked like a charm for the A’s, a team
that consistently finds itself in the playoffs even though it has
one of the lowest payrolls in the league.
So what is the NCAA Tournament’s equivalent of on-base
percentage? Is there one statistic that Final Four teams have in
common year after year, or is success in the tournament purely
decided by immeasurable principles like heart and
determination?
Picking based on rankings or seeding is nice, but this method
probably won’t win you that big pool.
It might be worthy to take a more statistical approach to
picking your winners. In this college basketball version of
Sabremetrics, I have broken down a few statistics to determine what
tournament champions have done consistently well throughout the
years.
Although there are always a few aberrations, numbers don’t
usually lie.
I looked at 10 season-averaged statistics: field goal
percentage, free throw percentage, opponent field goal percentage,
opponent points per game, rebounds, assist-to-turnover ratio,
blocks and steals.
I strongly believe that a good defense will overcome a good
offense in the tournament, and I personally place more weight on
defensive statistics than offensive numbers. But I couldn’t
ignore the fact that the champions have an average free throw
percentage of 71.23 percent. Success at the line is crucial in
crunch time, so if your team isn’t producing from the line,
you might want to take that into account.
But, like I said earlier, offense isn’t everything.
Arizona won the tournament in 1997 shooting just 66 percent from
the line. The key was that it held its opponents’ field goal
percentage to 43.
As these next few statistics should point out, defense is a
pretty good indicator of who wins championships. No champion since
1985 has ever been out-rebounded on average throughout its
championship season. Only one team (the 1997 Arizona Wildcats)
finished with fewer blocks than its opponents, and only two teams
(the 1986 Louisville Cardinals and the 1989 Wolverines) finished
with fewer steals than their opponents. Champions have held their
opponents to just under 70 points per game on average, and held
their opponents to 42.22 percent shooting from the field.
Additionally, every single tournament champion has finished with
an assist-to-turnover ratio greater than one. Maryland, the 2002
champion, had a ratio that was close to 1.5. The ability to protect
the ball should never be underestimated. Although this seems fairly
obvious, many people choose to ignore turnovers because they go
hand in hand with an explosive offense that takes chances.
Most of the top teams in the country are solid in these
categories. The challenge is to use these categories to determine
potential upsets. For example, take a team like Murray State. This
team ranks among the top 10 in the nation in scoring, has dominated
on the boards with 37.4 per game, has averaged 19 assists per game
and holds its opponents under 70 points. Other low-seeded teams
that have performed well in these areas include Manhattan (76
percent from the line), East Tennessee State (ranked among the top
teams in the nation in blocks and steals) and Air Force (leading
the nation in opponent points per game with 50.1).
Although there are thousands of factors you can take into
account in filling out your bracket — and I have barely
scratched the surface — I hope that following these
statistical guidelines will give you (and me) a better shot at
winning that pool. And if you’re not into numbers, or think
that no statistic could ever predict the outcome of the tournament,
there’s always the “coolest name” approach. Have
fun.
Naweed Sikora actually had fun analyzing statistics for this
column. If you would like to see his spreadsheet, email him at
href=”mailto:nsikora@umich.edu”>nsikora@umich.edu