Kentucky Sports Radio

University of Kentucky Basketball, Football, and Recruiting news brought to you in the most ridiculous manner possible.

Is the pre-season AP Poll better than the final AP Poll?


We are just hours away from the kickoff of college basketball season. The Wildcats will soon begin their quest for the ninth banner in school history. Big Blue Madness will be as extravagant as ever, and a host of elite high school stars will be in the house. Media members and coaches alike have decided on Kentucky as the No. 1 team in the land, and while we still wait for the AP Poll it is more than likely the Cats have earned the top spot; they certainly have the talent to back up the lofty expectations.

We often hear how pre-season polls mean little in the long run, and to an extent it is true. With teams regularly facing player turnover from year to year and nothing more than a handful of close closed practices and recruiting rankings to base it on, pre-season rankings have a tendency to be a little off sometimes.

But does a lack of information really alter the perception of the top tier teams? If one squad is such an easy pick to be a title contender that the majority of people tuned into the sport pick them as top dog, can they all be wrong? What exactly are the end-game implications for pre-season top teams? I set out to find the answers, and the results might surprise you.

I took the Final Four teams from each year for the last 20 years and compared their pre-season AP Poll ranking to their final AP Poll ranking. Outside of a strangely bizarre 2006 season when Florida, George Mason, and LSU all made the Final Four, there was never a season where more than one team didn’t appear in the top-25 somewhere along the way.

What’s even more surprising is that the eventual national champion will more than likely come from the very top. In the last 20 years, 65 percent of the title winners were pre-season ranked among the top four teams in America. Only three came from outside the top-25, with the most recent example of Connecticut in 2011.

The graph shows the relationship between the pre-season and final AP Polls based on how correct they are at predicting the Final Four teams. The higher the peak the less accurate the rankings were. What you’ll notice is back in the late 90s and early 2000s the pre-season poll was by far a less accurate predictor versus the final poll, outside of a couple of years.

But watch what happens once you get to 2007. Those lines get really close. That suggests the pre-season poll is no more accurate than the final poll. What does that mean? It means Kentucky is ranked No. 1 and that’s a good place to be. It turns out intuition and gut instincts can be just as good as watching teams play for 18 weeks.

Next I took the final AP Poll top-four teams and recorded the end result in the NCAA Tournament. Now this is tricky because you can only have one winner and upsets happen all the time. However, we can still see just how many upsets occur with the perceived top teams at the end of the year.

What I found was that 14 title winners came out of the top-four teams in the final AP Poll. Compare that to 13 titles winners that came out of the top-four teams in the pre-season AP Poll. Again, this signals that the pre-season poll is a very good predictor of the Final Four and national champion.

The second graph is a plot of the final AP poll teams against how early they were bounced in March. A numerical value was assigned to each round of the tournament and the higher the peak, the bigger or more upsets happened. The results are across the board, which means the final AP Poll alone isn’t an accurate gauge for Final Four teams – however it’s still more than likely a title winner comes out of the group. Obviously March is an untamable beast, so this comparison is a little unfair. For what it’s worth, the final AP poll top four teams average an Elite Eight appearance.

So what does all of this mean? Well, it means the pre-season poll isn’t quite the joke that many make it out to be. It turns out media members, coaches, and fans are pretty good at assigning the best teams, and even with 18 weeks to watch teams play it still doesn’t mean you’re the best team. Sometimes what we know from the get-go is best. Now last season Kentucky was a major outlier and it would be really easy to use the same argument against the Cats this season. But everyone knows this year is much different. And there isn’t a place I would rather be right now than among the pre-season top four teams.

Article written by Stuart Hammer

B.S. Broadcast Journalism from the University of Kentucky. @StuartHammerKSR

9 responses to “Is the pre-season AP Poll better than the final AP Poll?”

  1. Sayeth

    Are you graphing in Hebrew or something? Who puts a timeline right to left on a horizontal axis and puts more accurate below less accurate? Do you even graph, bro?

  2. Duuuuuude

    It all depends on the year. Last year, there were no teams that stood out. Any of the top 4 teams this year could have been #1 last year. With a field so weak last year, why not pick the defending national champions with the #1 recruiting class. As we now know, they were just not that good despite the number 1 ranking as a class. Any of the top 4 teams this year could lay claim to #1 for a number of different reasons. The outcome cannot be determined until after they play. This is why they play.

  3. twocoach

    What a mess. Just a reminder that Kentucky was preseason #3 last year…

    Why don’t you go figure out how many preseason Top 4 teams ended up making the Final Four that year? Odds are it’s no more than a coin flip meaning the “experts” were just as likely to be wrong as right.

    The one crazy year was 2008 when all four were #1 seeds and all four made the Final Four.

    Preseason Top 4:
    2012-13: Indiana, Louisville, Kentucky, Ohio State (one made Final Four)
    2011-12: Carolina, Kentucky, Ohio State, UConn (two made Final Four)
    2010-11: Duke, Mich. State, Kansas State, Pitt (zero made Final Four)
    2009-10: Kansas, Mich. State, Texas, Kentucky (one made Final Four)
    2008-09: Carolina, UConn, Louisville, UCLA (two made Final Four)
    2007-08: Carolina, UCLA, Memphis, Kansas (four made Final Four)
    2006-07: Florida, Carolina, Kansas, Pitt (one made Final Four)

    So if you are looking for a preseason poll to mean ANYTHING, you are more likley to be disappointed. It is truly meaningless.

  4. PhilUK18

    This doesn’t surprise me too much. The people tasked with establishing preseason rankings are given that job because they generally know what they’re talking about, both the coaches and the AP. It looks like they don’t sometimes because of a bias people have where they point out every time an expert or group of experts was wrong (Kentucky last year), rather than pointing to the numerous examples in the same season or poll where they were right.

  5. ThreeCoach

    3.

    The national champ came from the top 4 in 5/7 of the years you used as examples in your much smaller and therefore less significant sample size. You’re not too bright, are you?

  6. twocoach

    Oh, I understand. But who is to say which one of the Top 4 teams it will be?

    Any person that pays any attention to college basketball could name the eventual nation title winner if they were allowed to pick FOUR teams. Gimme Kansas, Kentucky, Michigan State and Duke this year. I likely have a better than 71% chance (your “impressive 5 of 7) that one of them will win the national title.

    The point is, you don’t HAVE to be “too bright” to accomplish that.

  7. ThreeCoach

    6.

    That’s the whole point of this post, the accuracy of the preseason polls in regards to picking the national champion…

    Can you read, twocoach?

  8. Jeff

    Who cares about pre-season polls?? You have way too much free time. You should be more excited about Phish’s Fall Tour opener tonight in Hampton, VA than a practice.

  9. Kentucky_lady

    Last time we were preseason #1 it worked out.