As the dawn of the 2013-2014 basketball season draws near, different polls and power rankings will be published forecasting the top teams in the nation. One of the most prominent polls, already released, is the USA Today Coach’s poll. As most of you already know, this poll gathers a Top 25 ranking from 31 Division I coaches and combines them into an aggregate ranking. Although the subjectivity of these non-analytic rankings can be argued, this poll annually presents an accurate representation of the public’s perception of the top teams in the nation. Or maybe put more accurately, is instrumental in shaping the public’s perception of the best teams in the nation. Teams that linger in and around this poll are normally safe bets to dance in March. The poll is generally important in predicting March Madness seeding, and a Top 25 team in theory will always be favored in their first round (or whatever its called now, second?) matchup. As the season progresses these coaches update their rankings weekly based upon the winning, losing, and happenings (strength of schedule, close loses to good teams, etc) within the league that given week. This means that the rankings at the end of the season are a better indicator of which teams were successful over the course of the season, than the preseason rankings. As more information arises (games played) the final Top 25 picture develops.
Assuming the preseason rankings are based upon the perceived strength of last season’s returners in combination with incoming freshman/transfers, the coaches initially have a decent amount of information to base their projections on. But how good are the preseason rankings? Do these coaches normally have a good grip on the season before it starts? These questions can be addressed by comparing the initial season poll to the end of season poll, and by analyzing how much these two polls differ. The following graphic shows the recent history of such an investigation throughout the past ten seasons. The red line represents the percentage of teams that fell out of the poll by season’s end, while the blue line shows the number of teams falling out. About ten teams each season or 41% of teams that enter the season ranked do not finish as ranked teams.
The above trend peaks at 13 teams falling out of the top 25 in 2006 and 2010, and falls to its lowest level of 8 teams falling out in 2005, 2009, and 2012. The inverse way of looking at this data (for initially unranked teams) is that typically only 8-13 spots in the rankings will open up by season’s end, or about 41% of the Top 25 positions are up for grabs on day one. The obvious point here is that it’s much better to be ranked, in terms of your season’s potential, in a preseason poll than to not be ranked. If a team starts off ranked (depending upon where they are ranked) historically they tend to have a 60% chance of staying ranked, while all of the other 300+ teams fight for the 40% of slots that open up. This is just the historical trend, and could change from year to year but in general spots will become available. Considering all of the factors that go into a team’s success during the season (injury, players reaching their potential, chemistry, luck with officials, etc.), the coaches forming the Top 25 should be complimented for picking 60% of the top teams before day one.
The next graphic presents the frequency of specific "ranks" falling out of the top 25 over the past ten seasons. As one would expect the higher ranked teams (1-10), from the preseason, have less of a tendency to fall out than the lower ranked (11-25) teams. In fact, over the past ten seasons no team that has started the season ranked number 1 has fallen out of the rankings by season’s end. All other ranks (2-25) have had at least one team fall out of the rankings by season’s end. The teams that started ranked within the top 10 had about a 77% chance of staying within the Top 25, as compared to a 47% chance for the teams ranked between 11 and 25. And breaking that down further the tail end teams (21-25) saw their chances plummet to 38%.
So are certain teams inherently overvalued from season to season, based on name alone? The answer I think is yes (this is just a guess, a more thorough analysis would need to be done to confirm). The following graphic shows the specific teams over the past ten seasons that fell out of the poll three or more times. Although the rate of teams falling out of the rankings would give us a better indication of who is and who isn’t overvalued, the number of classic basketball schools on this list stands out nonetheless.
Arizona, Georgetown, Kentucky, Michigan State, Syracuse, N.C. State are all teams that are generally regarded as great basketball schools. Arizona is a unique case. Over the past ten seasons they have started off in the preseason top 25 six times, but had fallen out by season’s end five of those six times. In addition, two of the times they started unranked they rose into the final top 25. The coaches had no feeling for this school. The presence of Gonzaga and Memphis on the list suggests a non poll–bias, but in recent years these teams along with other non-power conference schools have seemed to become public teams. The last graphic shows the teams on the other end of the spectrum. These teams represent all of the teams who climbed into the Top 25 three or more times, by seasons end after being initially unranked.
It’s interesting to see Georgetown and Notre Dame appear on both the falling out and rising into lists. The appearance of Butler, BYU, and Xavier helps to support a coach’s bias theory. It seems (based on a little evidence) that the coaches overrate "traditional" basketball powers and underrate non-power conference teams. As a side note, Xavier had a phenomenal run in the mid 2000’s where they seemed to collect sweet sixteen births like the Phillies used to collect NL East championships. They reached five sweet 16s and two elite eights over the ten-year span. The polling coaches were clearly missing something with the musketeers.
What does this mean for Villanova? Currently only one big east team graces the top 25 (Marquette), while Creighton, Georgetown, and Nova linger close with votes. I think the majority of us maintain high hopes for Nova’s upcoming run, but if March rolls around without a seed in the tournament than many will view their campaign as a failure. However, if they compete for the Big East regular season crown, as many expect, than a spot in or around the top 25 is for the taking. As history suggests though, the odds of making it into the final top 25 rankings are already against them (not saying they won’t, just that it’ll be difficult). The reason is because of the relatively small turnover of teams within the initial top 25 mixed with a high number of quality teams fighting for those few spots. This isn’t to say the odds are against them to make the tournament. The odds are most likely in their favor as a tough big east could earn four or five postseason berths. Hopefully Nova can work their way up the rankings, and the big east as a whole displays a strong showing in non-conference play. Go Wildcats!