the standout observation, to me anyways is the following: Since we know (thanks to @medinabuckeye1 prior posts), the 9 has a winning record against that '8' seed, they are actually dreadful versus the #1 seed in second round, compared to the '8' seed on a relative and an absolute basis.
I've always found that odd because it is counterintuitive. You would think the #8 seeds would have a slight winning record against the #9's and then the two would be about the same beyond the first round but neither of those things are true. Instead the #9 seeds have a slight winning record against the #8's and then the #8 seeds are VASTLY better beyond the first round. Two other things stand out to me:
#1 seeds are just really, Really, REALLY good:This one makes sense to me. My observation over the long-term is that there tend to be about four just extremely good teams each year. Some years there might only be two or three and in that case some of the #1 seeds aren't so great and other years there might be five or six and in that case some of the #2 seeds are extremely good but roughly, the #1 seeds are extremely good and all the rest, even the #2 seeds have weaknesses. When your team plays a #2 seed, I think generally that there are some weaknesses and you hope that your team can exploit those. When your team plays a #1 seed you pretty much need a miracle, basically you are just hoping for them to have an off night.
#1's aren't just incrementally better than #2's in the same way that #2's are incrementally better than #3's and so on. It doesn't show up so much in the first round where the #1's are only incrementally better but it shows up deeper into the tournament: Making the S16:
- 136 of 160
- 102 of 160
- 84 of 160
- 77 of 160
Note that the gap between #1 and #2 is quite large (34), much larger than the gaps between #2 and #3 (18) and between #3 and #4 (7).
That continues throughout the Tournament but it hits a peak with Championships:
- 25 of 39 Championships
- 5 of 39 Championships
- 4 of 39 Championships
- 2 of 39 Championships
- 0 of 39 Championships
- 1 of 39 Championships
- 1 of 39 Championships
- 1 of 39 Championships
The #1 seeds have almost twice as many Championships as the rest of the field combined (25 vs 14).
Also, if you look at the winning percentage of each individual seed in each round, they almost all fall off a cliff in the first round in which they can encounter a #1 seed.
The #13 seeds and below are just really, Really, REALLY bad:This one also makes sense to me. The last of the at-large teams come in as #11's or #12's and then you have a few "tallest midgets" that are just about as good as those last few at-large teams but then you run out of decent teams and you just have league tournament champions from leagues that are awful.
Similar to the 1 v 2 difference, the difference here is more than just incremental. #13's aren't just incrementally worse than #12's, there is a major dropoff. It shows up somewhat in the first round where the #13's have 24 fewer upsets than the #12's (as compared to #12's only having five fewer than #11's) but it REALLY shows up deeper in the Tournament. The #10, #11, and #12 seeds that do pull off an upset in the first round actually have a higher winning percentage in the second round than the #10-12 seeds do in the first round. In theory they shouldn't because:
- A #12 that knocks off a #5 generally has to play a #4
- A #11 that knocks off a #6 generally has to play a #3
- A #10 that knocks off a #7 generally has to play a #2
Their winning percentage should drop as they face even better opposition, it doesn't:
- #12's are .356 in the first round, the survivors are .386 in the second round.
- #11's are .388 in the first round, the survivors are .435 in the second round.
- #10's are .388 in the first round, the survivors are .403 in the second round.
The few #13 and worse teams that do manage to survive the first round should do BETTER in the second round because their opposition is typically not quite as good:
- A #13 that knocks off a #4 gets either a #5 or a #12
- A #14 that knocks off a #3 gets either a #6 or a #11
- A #15 that knocks off a #2 gets either a #7 or a #10
- A #16 that knocks off a #1 gets either a #8 or a #9 (in actuality both of them got #9's)
Their winning percentage should improve as they face less challenging opposition, it doesn't:
- #16's are .013 in the first round, the two survivors lost in the second round.
- #15's are .069 in the first round, the survivors are .364 in the second round (this is the ONE exception but I see it as a lack of data issue since there have only been 11)
- #14's are .144 in the first round, the survivors are .087 in the second round.
- #13's are .206 in the first round, the survivors are .182 in the second round.
