How important are pre-season Top 25 polls? How successful are they in predicting deep March runs?

[Travelman6]

Saw this floating around, here is a condescended version:

Q: How often did the poll rank the eventual champion number 1?

A: 6 times. That means, out of 33 tournaments, the AP poll has nailed the champ 18% of the time.

• The last team to be ranked preseason #1 and win: UNC, 2008-09.
• The preseason #1 team has been the runner-up 7 times.
• 3 of those 6 preseason #1 teams to win the title were pre-2000: 1990 UNLV, 1992 Duke, and 1996 UK.

Q: How often did the poll rank the eventual champion in the top 5?

A: 20 times. That gives the AP Preseason Top 5 a record of 20-13. If you want to know how many times the AP poll has ranked the eventual national champ period, that would be 29 out of 33. The only unranked national champions were 2003 Syracuse, 2006 Florida, and UConn twice (2011 and 2023).

A couple of other notes:

• 26 national champions were ranked in the Preseason AP top 10.
• The AP top 5 missed 4 times in the 90s, twice in the 2000s, and 7 times since the 2009-10 season.
• The Preseason #6 team has won the title 3 times.

Q: How many Final Four teams were present in the poll’s top 5?

A: Out of 132 possible Final Four teams, 56 were ranked in the AP Preseason top 5. That’s good for a 42% success rate. That’s good for an average of 1.68 final four teams correct per season.

Some other tidbits relating to this question:

• The entire Final Four has only been present in the preseason top 5 twice: 2001 and 2008.
• In 2008, the Final Four teams were identical to the preseason top 4 teams.
• 2006 was the first time the preseason top 5 went 0-fer. It also happened in 2011 and 2023 (both UConn championship seasons)

 

[AZHuskiePop]

For me pre- preseason polls are pretty worthless .
The first real meaningful poll that I pay attention too is around the end of the year .
How your conference or more importantly you did in your out of conference schedule pretty much determines if you or how many teams from your conference will likely make the tournment . Conference play will. determine the order of seeding . There is always the possibility of a collapse of a team or a team with a so so OC record to get hot but that’s abnormal .
Just look at where we were picked last year or in years we won NC . A lot of the Preseason voting is based on the previous year which is totally meaningless UNC is at or near the top coming off a NC but fell apart.
Wake me up at Christmas

 

[auror]

About the only thing that preseason polls are good at is identifying "this team is absolutely loaded". It is extremely rare for a top 5 team to be bad and as pointed out the top 5 includes nearly half of all final four teams. UNC from last year was so noteworthy because it's such an outlier.

Once you get closer to 10 and beyond, the poll is mediocre. But fun to talk about.

 

[Palatine]

The real value of being in the preseason Top 25 is keeping the team name in the media. That reinforces the status of the school as an elite brand.

On a related note. North Carolina Tar Heels Make Dubious History As First Preseason #1 To Miss NCAA Tournament

 

[BGesus4]

Saw this floating around, here is a condescended version:

Q: How often did the poll rank the eventual champion number 1?

A: 6 times. That means, out of 33 tournaments, the AP poll has nailed the champ 18% of the time.

• The last team to be ranked preseason #1 and win: UNC, 2008-09.
• The preseason #1 team has been the runner-up 7 times.
• 3 of those 6 preseason #1 teams to win the title were pre-2000: 1990 UNLV, 1992 Duke, and 1996 UK.

Q: How often did the poll rank the eventual champion in the top 5?

A: 20 times. That gives the AP Preseason Top 5 a record of 20-13. If you want to know how many times the AP poll has ranked the eventual national champ period, that would be 29 out of 33. The only unranked national champions were 2003 Syracuse, 2006 Florida, and UConn twice (2011 and 2023).

A couple of other notes:

• 26 national champions were ranked in the Preseason AP top 10.
• The AP top 5 missed 4 times in the 90s, twice in the 2000s, and 7 times since the 2009-10 season.
• The Preseason #6 team has won the title 3 times.

Q: How many Final Four teams were present in the poll’s top 5?

A: Out of 132 possible Final Four teams, 56 were ranked in the AP Preseason top 5. That’s good for a 42% success rate. That’s good for an average of 1.68 final four teams correct per season.

Some other tidbits relating to this question:

• The entire Final Four has only been present in the preseason top 5 twice: 2001 and 2008.
• In 2008, the Final Four teams were identical to the preseason top 4 teams.
• 2006 was the first time the preseason top 5 went 0-fer. It also happened in 2011 and 2023 (both UConn championship seasons)

Would be interested to see the same for the poll right before the tourney each year. I remember hearing the preseason poll is a better predictor of tourney success.

 

[auror]

Would be interested to see the same for the poll right before the tourney each year. I remember hearing the preseason poll is a better predictor of tourney success.

The pre-season AP poll is great | The kenpom.com blog

 

[QuickDraw]

last years champ wasn't even ranked in the preseason

 

[MTBingNinja]

We were unranked last year, I forget how we ended up on the season.

It’s good for recruiting purposes.

 

[TheHorde]

I think in the era of the transfer portal, the polls are worth less than ever. And they weren’t a great predictor to begin with. It’s tough to guage team chemistry and portal fits in the preseason polls.

 

[navery12]

I think in the era of the transfer portal, the polls are worth less than ever. And they weren’t a great predictor to begin with. It’s tough to guage team chemistry and portal fits in the preseason polls.

This is a hilarious response to a post with data showing that the AP Poll is a very good predictor of March performance

 

[TheHorde]

This is a hilarious response to a post with data showing that the AP Poll is a very good predictor of March performance

The initial AP poll? The one that had how many 2023 Final Four Teams ranked? And how many national champions were ranked #1. Oh you are hilarious! Lollol

 

[superjohn]

This is a hilarious response to a post with data showing that the AP Poll is a very good predictor of March performance

I think his first sentence is probably true. Polls are going to have more trouble capturing transfer impact. We've probably only had two or three tournaments with the transfer portal and one of them we won it all while being unranked to start the season.

 

[TheHorde]

I think his first sentence is probably true. Polls are going to have more trouble capturing transfer impact. We've probably only had two or three tournaments with the transfer portal and one of them we won it all while being unranked to start the season.

Examining the Accuracy of "Way-Too-Early" Preseason Top 25 Rankings - Stadium

The "way-too-early" preseason Top 25 rankings have value even though the 2019-20 college basketball season won't tip off for five more months.

watchstadium.com

It’s very true. The preseason polls are good at identifying the top teams that will be national contenders and likely in the final polls. It’s not good at identifying the bottom half of the top 25 or predicting March success. It’s like predicting the Bills, 49ers, and Chiefs will be top NFL teams in August. They’ll make the playoffs but some (many) fall flat on their faces in the playoffs.

 

[navery12]

The initial AP poll? The one that had how many 2023 Final Four Teams ranked? And how many national champions were ranked #1. Oh you are hilarious! Lollol

Yes, try reading the initial post and the KenPom article linked. Using a sample size of 1 year is stupid, even for someone like you who I have zero respect for