The College Basketball season kicks off in just three days, headlined by the State Farm Champions Classic. For the first time since the tournament introduction, the four participating teams are the top four ranked schools in the preseason in the Associated Press Preseason Poll, with matchups between #3 Kansas and #4 Duke and, of course, #1 Michigan State and #2 Kentucky.
The beginning of the 2019-20 season also marks the thirtieth anniversary of the expansion of the AP’s poll to 25 teams, with the Preseason Rankings for the 1989-90 being the first instance of the expanded list. At the time, the University of Nevada Las Vegas Rebels were the #1 seed, captained by Larry Johnson, the eventual first pick of the 1991 by the Charlotte Hornets. And what a way it was to kickoff a new era in the Associated Press Poll – the Rebels would end up winning the 1990 National Championship, defeating the #9 seed from the East, the Duke Blue Devils.
But, if the accuracy of the Preseason poll is impressive, the same cannot be said by its evolution during the season, culminating with Oklahoma, the first ranked team at the end of the year, falling in the second round of the tournament against #8 North Carolina, 79-77.
Even though the Rankings compiled by the Associated Press aren’t made with the intention of predicting who will win the Championship, they are a Poll of which teams are seen as being/playing the best. Therefore we should be able to, during the years, analyze exactly how accurate has the AP Poll been in predicting the success of teams in March Madness? We analyzed the Preseason Polls and the End of the Year Polls and found some fascinating trends:
First of all, credit where credit is due – in the thirty season since the expansion to twenty five teams, not once has the Associated Press given the Preseason number one rank to a team that didn’t make the tournament, and only once has the number one team not won a single game in March Madness – the 2004-05 Kansas Jayhawks, who fell to #14 Bucknell, in a game that won the 2005 ESPY for Upset of the Year. That number is exactly the same achieved by the End of the Year Polls, with the Virginia Cavaliers being the one team eliminated in the first round, in the 2018-19 season, against UMBC, the first time that a number one seed was eliminated by a sixteenth seed in March Madness history.
But perhaps the most unexpected fact is that, in both predicting the Champion and the Runner-Up, the Preseason Rankings, based more on the results of the previous season, the players who left and the incoming recruiting classes, are significantly more reliable than the End of the Year Poll, which has the advantage of knowing which teams have performed better or worse, which injuries happened and how exactly is the field supposed to perform in the Tournament. As you can see in the chart above, the Associated Press Preseason Poll, correctly predicted a combined 12 out of 30 March Madness Finalists, for a total of 40%, against only 7 correct results, 23% or almost half, in the End of the Year Polls.
More so, in the four times (1992, 1996, 2001 and 2012) in which the number one ranked team at the End of the Year Poll went on to win the Championship, the ranking of said team in the Preseason Rankings was either #1 (in 1992-93, when the Duke Blue Devils swept the rankings all season) or #2 in three occasions, with the only exception coming in 1995 when the AP Preseason Poll had the Arkansas Razorbacks as number one and the UCLA Bruins, the eventual champions, were only the sixth ranked team.
If we look at the same set of years, the four years specified earlier, but from another perspective, we can also see that the team that was number one in the Preseason (and excluding the 1992-93 Blue Devils that we have already talked about), was the Runner-Up in both 1995 (#1 Arkansas lost to UCLA) and 2001 (#1 Arizona lost to Duke), with 2012 being the only exception, seeing as the Preseason number one North Carolina Tar Heels only managed to reach the Elite Eight, falling to the eventual runner-ups Kansas Jayhawks.
Overall, in the last thirty seasons, the Preseason Pool was a better predictor of March Madness success than the Final Rankings in a third of the seasons, while, in thirteen other seasons both the Preseason leader and the End of the Year leader fell on the same round of the tournament.
All and all, Tom Izzo and the Michigan State Spartans, hoping to be Nationals Champions twenty years later, should be fairly optimistic for the 2019-20 season, the first in their history where they start as the number one ranked team in the country, seeing as, according to the historical performances reviewed in this article, they have a 50% chance of reaching the Final Four in March (more like April).
A Moment In Sports History 