How stats and history would pick the 2021 NCAA Tournament

Sweet Sixteen

West Region

(1) Gonzaga over (5) Creighton. Or literally anyone left in this region, tbh. Gonzaga is at 76-77% to make the Elite Eight; only three teams ever have topped 72%, and two (2000-01 Duke, 2007-08 Kansas) went on to win the title.

(2) Iowa over (6) USC. Iowa actually cracks 50% to make the Elite Eight, which is fairly unique. Only 14 teams have ever hit this mark before, and nine of them made the Elite Eight. Even if you go down to 40% or greater, it’s still an excellent 23-for-34 success rate (all others 12-for-46).

South Region

(1) Baylor over (5) Villanova. I HATE THIS PICK. If Villanova were at full strength, I’d legitimately consider pulling the trigger and making a run at it with the Wildcats. Sadly, Baylor is just above the 44% cutline for removing a 1 seed entirely. 1 seeds at 44% or better to get to the Elite Eight are 47-for-61 in getting there. Baylor easily has the lowest odds of any team to get there on KenPom, which makes me think they’re going to fall at some point. As I type, 71% of ESPN users have Baylor in their Elite Eight and 51% in the Final Four. There is massive value to bet against Baylor here. I just need more convincing.

(2) Ohio State over (6) Texas Tech. I can’t bet against Ohio State, who is around 42-44% to make the Elite Eight. Refer to the Iowa stat of 40% or above for why I regrettably like OSU.

Midwest Region

(5) Tennessee over (8) Loyola Chicago.

Written on March 15, 2019, after I initially picked Loyola:

Please shoot me in the balls. Oh my God.

Here’s the problem. As it stands, Loyola Chicago is ranked above Tennessee on both KenPom and Torvik. If they beat 1-seed Illinois, they’re obviously going to remain in the KenPom top 10 if not elevate themselves further. They currently have the third-best Elite Eight odds of any 8 seed in the last 20 years. The only 8 seed comfortably ahead of them in this category – 2000 Wisconsin – made the Final Four.

I’m also going to say this: it’s a reverse jinx. If I pick an 8 seed to get this far, the odds that they blow it in the Round of 32 escalate greatly. Perhaps Baylor is the 1 seed that loses instead. Perhaps this matchup does happen and Tennessee wins. You have to do really stupid things sometimes, and this is one. But I’m making myself do it.


Changed my mind. I don’t love this pick at all, and it makes my stomach a tad upset…but it’s just safe enough and contrarian enough to work. Tennessee’s Sweet Sixteen odds are 17.4% compared to Loyola’s 13.5%, so you’re about 3.9% more safe here. It’s not much, but it works. Along with that, of the seven 5 seeds to make the Elite Eight since 2000, five have been at 17.5% or better. Tennessee’s right on that cutline. Loyola may very will win this hypothetical game, but look: I’ve gotta go down swinging.

(2) Houston over (3) West Virginia. 2 seeds at 50% or better: 9-for-14 in getting to the Elite Eight. (40% or better is 23-for-34, again.) Houston is around 50-52%. West Virginia is at 17%; only one 3 seed below 20% has ever made the Elite Eight out of 17 applicable teams.

East Region

(4) Florida State over (1) Michigan. If you see me on Sweet Sixteen weekend, please know that I want to be thrown into a wood chipper if the Tennessee and Michigan scenarios come true.

This is a precautionary pick. Florida State is a very good 4 seed with anywhere from 17-23% Elite Eight odds. No 4 seed has ever made it sitting below 15%, and generally, you want 22% or better. My odds have FSU at 23.4% after I factor in a small detraction for the Livers injury for Michigan. Along with that, 1 seeds below 55% to make the Elite Eight are only 21-for-40 in getting there. Even a pre-adjustment Michigan tops out at 53%. I’ve changed my mind. I don’t want the NCAA Tournament back.

(7) Connecticut over (11) UCLA. UConn is a neutral-court favorite over UCLA, but more importantly, it’s very hard to push the button on an 11 seed. UConn’s odds aren’t good enough to make the Sweet Sixteen by my traditional metric of 20% or better, but 16% is close and UCLA doesn’t meet the 11 seed metrics for Elite Eight-ship.

Elite Eight

(1) Gonzaga over (2) Iowa. Gonzaga already won this game once, but this one will be more meaningful. Gonzaga sits at 59-60% to make the Final Four. 1 seeds with 46% or better odds to get to the Final Four are 14-for-15, and no team at 55% or above – as Gonzaga is – has missed it. Iowa, meanwhile, is right at 20%. No 2 seed has made the Final Four (0-for-44) with worse than 20% odds of making it, and only four out of 59 have made it at 25% or lower.

(2) Ohio State over (1) Baylor. Another value play. KenPom has Baylor under 32% to make the Final Four; 1 seeds below 33% are only 3-for-37 in making it. I personally have Baylor closer to 33-34%, but it’s still a borderline case, and in general, 1 seeds below 46% are 16-for-65 in making it. The odds aren’t in their favor. I don’t think Ohio State is a fabulous long-term value, but at 22.8% (21% on KenPom), they’re the only team from the bottom half of the bracket that doesn’t fail to meet any potential Final Four percentage guidelines. Only the East Region has lower total odds of a top two seed getting to the Final Four. I’m very intrigued by Villanova as a value play here, along with Wisconsin.

(2) Houston over (5) Tennessee (or Loyola Chicago) (or Illinois). Three things. Firstly, Houston has nearly-equal Final Four odds as Illinois does in my metrics composite: 31.7% vs. 31.3%. Secondly, if it is Illinois, Houston would only be a one-point underdog on a neutral court. Lastly, none of Houston’s potential opponents meet historical guidelines for a Final Four appearance. Illinois is either below or 33% threshold (per my numbers) or barely above it (KenPom alone). Loyola Chicago is at 6%, which is the sixth-highest ever for an 8 seed but there’s not a ton of correlation for 8 seeds getting to the Final Four beyond being at 2.2% or above. We’ve covered Tennessee, Oklahoma State, etc. It’s Houston.

(4) Florida State over (7) Connecticut. This is an impossible choice, because neither 4 seeds nor 7 seeds have any true statistical rhyme or reason on who makes the Final Four. There’s some value in sticking with 4 seeds who are at 14% or better to make it (2011 Kentucky, 2006 LSU, etc.) and giving serious thought to taking 7 seeds at 6% or better (2014 UConn, obviously, but 2012 Florida nearly made it as well), but neither has much correlation beyond consistent early-round success. If you’ve made it to this matchup, you’re either nuts or really shooting the moon, of which I am both. So I’m going to take the higher-rated metrics team and hope for the best.

NEXT PAGE: Final Four, title game

One thought on “How stats and history would pick the 2021 NCAA Tournament

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s