But what is the goal of the change?
- If you want more basketball, then fine. 80 teams, let the tallest midgets stay, and let some mediocre P5 teams in. But it won't be better basketball. You'll have two issues. First, the "bad basketball" of watching tall midgets get slaughtered doesn't go away. They just get slaughtered by mediocre P5 teams. Second, you actually increase the chances that the top seeds will get knocked out... But still not by teams with legit championship hopes. So you devalue the product by increasing the chance the top seeds aren't around for the S16, the E8, the FF.
- If you want better basketball, I can see getting rid of auto-bids and just taking the top 64 teams. It still has the problem above where the top teams have a higher chance of first round upsets than today, but at least you get rid of the slaughter of the tall midgets. So it's better in the sense that those first round games are maybe a little more competitivel.
- If you want the best basketball, just get rid of autobids, cut the field to 48, and give the top 4 seeds a first round bye. It won't be as exciting for fans rooting on Cinderella, and it won't ever happen because there's no way TPTB are going to forego the money of reducing the field size. But it'll be universally better basketball compared to what we have.
I just don't understand the value of going to 80 teams and letting in more mediocre power conference teams, though. What's the point?
The point to me, is twofold:
First, as I've said, it seems ridiculously unfair to me that mediocre major conference teams get passed over by ~20 tallest midgets that are VASTLY inferior to said mediocre major conference midlings. I'm not talking about league tournament cinderellas that won a conference with at least one tournament-quality team. That, at least in theory, could happen even in a power league. Back when we were the Big11Ten there was a year in which Illinois was the 11th place team and made it all the way to the CG. They didn't win (in part because this is a quality league with multiple quality teams and each successive upset is therefore both of a quality team (unlike in midget leagues) and successively harder both because the opponent is usually tougher and because piling upsets on top of one-another gets increasingly difficult for a myriad of reasons.
Looking at 2019 (the last tournament that was actually selected and played), there were 16 leagues with at least one team in the NCAA's NET rankings top 68. They ranged from the B12 and B1G in which about 80% of the teams were top-68 (8 of 10 and 11 of 14 respectively) down to the A10 where less than 10% of the teams were top-68 (1 of 14). I'm totally fine with those 16 league champions getting in. What I object to is the other 16, they and their highest ranked team were:
- AmEast, #71
- CAA, #76
- Ivy, #86
- CUSA, #91
- Summit, #104
- Horizon, #115
- SBT, #121
- BSKY, #124
- MVC, #127
- Patriot, #132
- BSouth, #144
- Southland, #154
- MAAC, #201
- NEC, #203
- SWAC, #205
- MEAC, #260
Even here, I don't really mind the Champions of the better of these leagues getting in. They are reasonably close to "tournament-quality" so who cares if the AmEast Champion, #71 gets in over the 9th place team in the B1G that is slightly better. However, it is ridiculous that the Champions of the MEAC, SWAC, NEC, and MAAC get in. Their best teams are probably worse than the worst team on most B1G teams' schedules. Including them is just silly and it inherently deprives a VASTLY superior team of an opportunity that they are obviously more qualified for and deserving of.
At the least, the play-in games should be for all four 16-seeds such that the worst eight teams in the tournament have to play them rather than the worst four and four others that are vastly superior to those tallest midgets. In 2019 that would have at least forced the #127 and lower tallest midgets into play-in's.
Second, it would make for MUCH more competitive basketball.
In the first round the top-12 seeds each win AT LEAST 1/3 of the time. The lowest among the top-12 are the 12 seeds who win almost 36% of the time (.357, 50-90). Here is how the 13-16 seeds do:
- #13 is .207 all-time against #4, 29-111
- #14 is .150 all-time against #3, 21-119
- #15 is .057 all-time against #2, 8-132
- #16 is .007 all-time against #1, 1-139
Ie:
- A 13 seed takes out a #4 less than once a year, about four times every five years
- A 14 seed takes out a #3 about twice every three years
- A 15 seed takes out a #2 about once every four years
- A 16 seed takes out a #1 once every 35 years
That isn't frequent enough to make those games interesting to me. When we fill out our bracket most of us just automatically advance the top-4 seeds then decide which 5/12 upset we want to go with. There is a reason for that. #12's have beaten #5's almost twice as frequently as #13's have won and #12's have as many wins as #13's and #14's combined.
Back up and look at the drop-off by seed:
- #10's have won 55 times in 35 years (55-85, .393)
- #11's have won 52 times in 35 years (52-88, .371)
- #12's have won 50 times in 35 years (50-90, .357)
- #13's have won 29 times in 35 years (29-111, .207)
- #14's have won 21 times in 35 years (21-119, .150)
- #15's have won 8 times in 35 years (8-132, .057)
- #16's have won once in 35 years (1-139, .007)
From 10 though 12 the drop-off is minimal from one seed to the next but note the HUMONGOUS drop-off from the 12 seeds to the 13 seeds. The 12 seeds are reasonably competitive, the 13 seeds are not. Why? In theory there are two possible explanations: Either the #13 seeds are substantially worse than the #12 seeds or the #4 seeds are substantially better than the #5's. Upthread I said that it was because the #13's are substantially worse and that I could prove it, here goes:
First off, it is plain obvious. #4's and #5's are nearly interchangeable just like #5's and #6's, #6's and #7's, etc.
Second, the last of the at-large teams (the play-in / last four in) are generally on the 11 or 12 line. In 2019 the at-large play-in games were for #11 seeds. Thus, the committee at that point switches from picking the next best tournament-quality team to being forced to take the least short of the remaining tallest midgets. The 12 seeds are still pretty good because the top few tallest midget conferences have at least one team each that is reasonably close to tournament-quality. However, the committee quickly runs out of those and starts putting in increasingly short tallest midgets. Here are the NET rankings by seed in the 2019 Tournament:
- #10: Range of 31-61, mean of 48, median of 50
- #11: Range of 32-73, mean of 73, median of 55.5
- #12: Range of 40-58, mean of 48.25, median of 47.5
- #13: Range of 68-103, mean of 80, median of 74.5
- #14: Range of 86-121, mean of 105.5, median of 107.5
Note the HUMONGOUS drop-off between the #12's and the #13's. THAT is why #13's win about half as often as #12's, not because #4's are substantially better than #5's.
Another proof, the next round:
When 10-12 seeds do win their first game (upsetting 5-7) they almost always have a tougher game in the second round (vs 4/13, 3/14, 2/15 and it is almost always 4, 3, 2 as demonstrated above). Despite that, their record is actually better in second round games:
- #10 is .393 in first round games against #7 (55-85) but improves to .418 in second round games against 2/15 (23-32).
- #11 is .371 in first round games against #6 (52-88) but improves to .423 in second round games against 3/14 (22-30).
- #12 is .357 in first round games against #5 (50-90) but improves to .420 in second round games against 4/13 (21-29).
All of the bottom four seeds stay the same or get worse in the second round except the #15's and the data-set there is too small to be useful:
- #13 is .207 in first round games against #4 (29-111) and stays the same .207 in the second round games against 5/12 (6-23)
- #14 is .150 in first round games against #3 (21-119) and declines to .095 in the second round against 6/11 (2-19)
- #15 is .057 in first round games against #2 (8-132) and improves to .125 in the second round against 7/10 (1-7). The problem here is that it is just one win in just eight tries so the data-set isn't large enough to rely on.
- #16 is .007 in first round games against #1 and declines to .000 in the second round against #9 (the one time it happened, they lost to a #9).
I *THINK* I can understand why 10-12 seeds get better in the second round. I *THINK* it is for two reasons:
- The overrated #10-12 seeds are underrepresented in second round games because they all lost in the first round, and
- The underrated 10-12 seeds are overrepresented in second round games because they are more likely to win in the first round.
The thing is that both of those things *SHOULD* apply to #13-#16 seeds as well. In addition, #13-16 seeds that do make it to the second round have two major advantages that *SHOULD* help their record in second round games:
- The games get easier even if the higher seed wins. A #12 that beats a #5 usually then has to play a #4 that is, at least in theory, even better than the team that they just beat. Same for a #11 that beats a #6 and gets a game against a #3 or a #10 that beats a #7 and gets a game against a #2.
- There is a VASTLY better chance that a #13-16 seed that does win their first round game will NOT get the chalk opponent. Consider:
- A #10 that beats a #7 has a 94% chance of getting a #2 in the second round and only a 6% chance of getting #15.
- A #11 that beats a #6 has a 85% chance of getting a #3 in the second round and only a 15% chance of getting #14.
- A #12 that beats a #5 has a 79% chance of getting a #4 in the second round and only a 21% chance of getting #13.
- A #13 that beats a #4 has a 64% chance of getting a #5 in the second round and a 36% chance of getting #12.
- A #14 that beats a #3 has a 63% chance of getting a #6 in the second round and a 37% chance of getting #11.
- A #15 that beats a #2 has a 60% chance of getting a #7 in the second round and a 40% chance of getting #10.
- A #16 that beats a #1 is obviously playing either #8 or #9 which are basically interchangeable. Admittedly, since this has only happened once this is insufficient data as well.
The point is that bottom-4 seeds that DO win in the first round always get an easier game in the second round and that second round game is significantly easier more than 1/3 of the time while 10-12 seeds to pull off a first round upset almost universally get a tougher opponent the next time out. In spite of that, #10-12 seeds improve from the first to the second round while #13-16 seeds decline, stay the same, or have insufficient data to make a determination.
The #12 seeds are reasonably competitive up until they run into a buzzsaw known as a #1 seed in the S16. In their S16 game a #12 seed could play a #1, #8, #9, or #16, of those:
- 85.71% of the time it is going to be a #1
- 9.29% of the time it is going to be a #8
- 5% of the time it is going to be a #9
- So far no #16 has gotten to the second weekend.
- #12 Seeds crash and burn in the S16. Their record there is an abysmal .048 (1-20).
- #11 Seeds that make the S16 play 2/7/10/15. Their record is .364 (8-14).
- #10 Seeds that make the S16 play 3/6/11/14. Their record is .348 (8-15).
Note that while the #12's crash and burn in the S16 because they invariably run into a #1, the #10 and #11 seeds that do make the S16 are still reasonably competitive even there. They win more than 1/3 of their games.
What about the #13-#16 seeds? Well we already know that no #16 has ever made the second weekend, no #13-#15 has ever won a second weekend game:
- #13 Seeds play 1/8/9/16. They are 0-6.
- #14 Seeds play 2/7/10/15. They are 0-2.
- #15 Seeds play 3/6/11/14. They are 0-1.
- #16 Seeds, if they made it, would play 4/5/11/13. They are 0-0.
In the 35 years of the 64+ team tournament there have been 560 #13-#16 seeds. They have had 560 chances to show us that they could win a S16 game and they have come up short EVERY SINGLE TIME. The #12 seeds have one S16 win. The #9 seeds have four and every other seed has at least eight. The #13-16 seeds have NONE.
The tallest midgets that make up the bottom of the tournament have absolutely no business being in a National Championship Tournament and 35 years of experience has demonstrated that beyond argument, as I have laid it out above. They flat out suck.
My personal preference is in this order:- Go to a 32 team tournament. As @bwarbiany pointed out above no #9 or below has ever won a NC but it is actually worse than that. No #9 or below has ever even made it to the NC game. Six (4 #11's, 1 #10 and 1 #9) have made it to the F4 but they all lost in the semi-final. The top-32 have made it to and won the NC, the other 32 (or 36 as it is now) teams are all superfluous window-dressing.
- Put in some kind of threshold that a conference needs to meet to get an auto-bid. Maybe "have at least one team ranked in the top-64" or even "have at least one team ranked in the top-100". The first of those would have eliminated 17 tallest midgets in 2019 while even the second would have at least cleared out the worst 12.
- Keep the auto-bids for every conference but expand the tournament to 80 teams. Using the NET Rankings from 2019 that would have meant that (assumes the committee went exactly by NET Ranking) the top-65 teams would have gotten in (getting the #1 through #16 seeds and one of the #17 seeds) along with 15 tallest midgets ranging from #68 to #260. Each of the 15 tallest midgets would have been placed in a game against a legitimate tournament-quality opponent but none of them would have been unwinnable at the level of the #1/16 games now. Instead, you'd see upsets probably around 1/4 to 1/3 of the time in each of the 16/17, 15/18, 14/19, and 13/20 games then those winners would have been MUCH better than the current #13-#16 seeds so the first full round would be more entertaining because there would be a realistic chance of an upset in every game instead of only about half of them.
My personal preference is #1 followed by #2. I push for expansion to 80 teams instead because I am pragmatic and realize that the NCAA isn't going to reduce their field by more than half nor are they going to exclude some conferences entirely so my third option is the only viable one.
