Updated for the first weekend of the 2023 Tournament:
It has been 39 years since expansion to 64 teams in 1985 but there have only been 38 tournaments due to the COVID-19 cancellation in 2020. Thus, there have been 152 of each seed.
Earlier,
@ELA and I had a discussion about the dropoff between #1 and #2 as contrasted to the dropoff between #15 and #16. I submit this as more evidence for my theory that the dropoff between #1 and #2 is VERY significant:
In the first round where #1 and #2 are playing #15 and #16 the winning percentages are fairly close:
- #1 is .987 against #16, 150-2
- #2 is .928 against #15, 141-11
It is the second round when the #1's and #2's start playing real live opponents with ability where the gap between #1 and #2 really shows up:
- #1's have made it to the S16 84.21% of the time, 128 of 152.
- #2's have made it to the S16 62.50% of the time, 95 of 152.
- #1's are .853 against the 8/9 winner, 128-22.
- #2's are .674 against the 7/10 winner, 95-46.
The problem is there there just isn't enough data on the #16's to base conclusions on. We know they are somewhat worse than the #15's in the first round and that the two that did manage to knock off #1 both lost (to #9) in the second round but a sample-size of two is not large enough.
I don't disagree with ELA entirely. There IS a big dropoff between #15 and #16. My point is simply that I think the #1's are, in general, a breed apart. There are exceptions (cough - Purdue - cough) but, in general, the #1's are just INCREDIBLY good teams.
The other thing that stands out to me is the HUMONGOUS dropoff between #12 and #13. This shows up in both the first and the second rounds:
If you think about it, the path to the S16 is effectively nearly identical for a #12 (5->4) and a #13 (4->5) and yet:
- #13's have been to six S16's
- #12's have been to more than three times that many, 22
Similarly, the path to the S16 is effectively nearly identical for a #11 (6->3) and a #14 (3->6) and yet:
- #14's have been to two S16's
- #11's have been to thirteen times that many, 26
Similarly, the path to the S16 is effectively nearly identical for a #10 (7->2) and a #15 (2->7) and yet:
- #15's have been to four S16's
- #10's have been to six times that many, 24
Similarly, the path to the S16 is effectively nearly identical for a #8/9 (8/9->1) and a #16 (1->8/9) and yet:
- #16's have never been to the second weekend of the Tournament
- #8's (16) and #9's (8) have been to a combined 24 S16's.