With the release of the first AP poll last Friday, it’s time for a refresher on the historic value of those rankings. The preseason AP poll is not going to tell you exactly how the season will play out, but given the poll has a long track record, we can use history to tell us the chances of a team ranked in a specific position getting a particular seed in the NCAA tournament.
If you’re familiar with my work, you know by now my support for the preseason poll, and this will seem repetitive. But the data below is updated to include last season, so this is not a complete waste of your time.
Last year, two teams ranked in the preseason failed to make the tournament: #17 Marquette and #21 Notre Dame. That’s better than normal. Since the poll expanded to 25 teams in 1990, an average of 3.7 ranked teams per year have missed the field altogether. Kentucky’s appearance in the NCAA tournament kept a perfect streak alive for the preseason’s #1 team. Every #1 team has made the field since it was expanded to 64 teams in 1984-85.
However, the Wildcats’ eight-seed was the lowest a preseason #1 has received, eclipsing the five-seed given to UConn in the 2000 tournament. It snapped a streak of eight consecutive seasons that the preseason #1 received a one-seed. And last season marked the first time in the 64-team era that none of the top five teams received a one-seed. It’s a possible that the pollsters were collectively dumber than their predecessors, but it’s more likely that the regular season was less predictable than usual.
Anyway, let’s get to the data. The table below represents the number of occasions (or percentage of years, where indicated) each slot in the rankings has received a particular seed in the NCAA tournament (NT = no tournament bid). For example, the tenth-ranked team has received a one-seed three times since 1985, or 10 percent of the time.
P r e a s e a s o n R a n k
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
1 20 14 11 9 12 4 8 2 3 3 6 3 1 0 0 2 3 1 2 0 2 2 0 3 0
2 5 8 7 5 7 10 5 10 5 5 1 3 6 1 3 1 0 2 2 2 2 1 0 1 5
S 3 3 3 4 7 5 3 3 2 2 5 3 3 2 5 2 2 1 1 1 4 0 4 3 3 1
e >3 2 4 7 8 4 12 13 15 17 14 18 16 16 17 19 21 17 21 18 17 13 15 16 10 14
e NT 0 1 1 1 2 1 1 1 3 3 2 5 5 7 6 3 8 5 7 7 7 2 6 8 5
d %1 67 47 37 30 40 13 27 7 10 10 20 10 3 0 0 7 10 3 7 0 8 8 0 12 0
%1-3 93 83 73 70 80 57 53 47 33 43 33 30 30 20 17 17 13 13 17 20 16 28 12 28 28
%NT 0 3 3 3 7 3 3 3 10 10 7 17 17 23 20 13 30 17 23 23 32 12 24 32 20
Champ 6 6 4 1 0 2 2 2 1 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 4
The following table takes the above data and aggregates it by +/- 2 of the position indicated. For example, the percentages under the third-ranked team represent the data for teams ranked first through fifth historically.
5 - s p o t a g g r e g a t e
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
%1 44 33 29 23 19 13 15 11 11 9 7 4 4 4 5 5 6 5 4 5 6
%1-3 80 73 67 61 54 47 42 37 34 31 26 23 19 16 15 16 16 19 19 21 22
%NT 3 4 4 4 5 6 7 9 12 15 17 18 21 21 21 21 25 21 22 25 24
%Champ 12 9 6 5 5 5 4 2 1 0 0 0 0 1 2 2 2 2 1 0 0
The idea here is that the voters do a fine job distinguishing the top-ranked team from the fifth-ranked team and the fifth-ranked team from the tenth-ranked team. However, there’s very little difference between the performance of the 15th-ranked team and the 25th-ranked team. The teams at the very end of the poll tend to miss the tournament slightly more often than teams in the mid-teens, but they’ve also received a high seed slightly more often. There are a lot of unknowns at the beginning of the season and too many to reliably distinguish between teams in the bottom half of the poll.
So the voters are probably right about #1 Kentucky being better than #5 Kansas and #5 Kansas being better than #10 Texas. But the teams ranked from 15th to the end of the poll could be reordered randomly and the poll’s accuracy may not suffer at all.
Maybe we shouldn’t say the preseason AP poll is “great”, but it definitely can be useful.