March Madness predictions
For sports fans, March Madness and the NCAA Tournament is one of the most fun times of the year. Analytics junkies and new bandwagon fans can all fill out their brackets and try and predict what craziness will unfold this spring.
Creating a perfect bracket is practically impossible, but it’s worth a shot. Like everyone else, my bracket has to include some upsets and expected favorites living up to the hype. Using some statistics, as well as gut feelings (all we can really trust in an event as unpredictable as the NCAA Tournament), here are some takeaways from what I believe to be the best bracket I made this year.
South Region: This region felt pretty predictable to me. The upsets, UC San Diego over Michigan and UNC over Ole Miss, are two of the most popular upset picks in the tournament, so nothing too crazy here.
UC San Diego is a very good basketball team that plays great defense and shoots a lot of threes, and I think they’ll be able to slow Michigan down. I think UNC, in typical UNC fashion. is getting hot at the right time and will be able to get past Ole Miss.
I have Texas A&M getting the win over Auburn to make it to the Sweet 16. They’ve done it before and I’ll give them the benefit of the doubt, with Auburn cooling down at the wrong time.
Nobody in the South will be able to get past Michigan State, however. They sit very highly in Kenpom and place well within the trapezoid of excellence (get used to hearing that). With the Aggies knocking off Auburn, they’re the next up out of this region. MSU vs ISU should be a great one.
East Region: This region is a little more fun in my opinion. While Duke is always a popular favorite in March, and they very well may be the best team in the country, 11 out of the last 13 Final Fours have included a 7 seed or higher. This year I think that team is VCU.
Before discussing their run, let's look at the rest of the region. The only upset I have in the first round is VCU (crazy I know). Liberty over Oregon is a popular upset pick this year but I don’t really see that happening. Looking a bit into St. Mary’s vs Vanderbilt, I don’t see a real chance of an upset there either.
VCU gets to avoid Duke, as I have Arizona pulling off the upset in the Sweet 16 there. Both teams find a spot in the trapezoid of excellence, and I think a younger Duke team falls to a deep and experienced Wildcats team.
VCU plays great defense which will help them slow down a lightning-quick Alabama team, and I think they’ll capitalize on their ability to get offensive boards in a victory over Arizona to take them to the Final Four.
West Region: Keeping with the trend, I don’t see many major upsets in the West. Colorado State over Memphis is a very popular pick and with good reason. Kenpom and other stats say CSU is just the better team, so I’m leaning toward the Rams for that one.
Another popular upset pick is Drake over Missouri. But similarly, the stats sway against the upset here. Especially on the graph showing the trapezoid of excellence, Drake is the furthest team away. They play extremely slow and do not possess a great net rating.
The big story for me in this region is Maryland. They have the most effective and efficient starting 5 in the nation (granted they pretty much only play 5 guys), and play great defense. In March, with shortened rotations and referees letting kids play, I have them knocking off Florida.
St. Johns shouldn’t have too tough of a time making it to the Elite Eight, but Texas Tech should give them a fight. The Red Storm should be able to get by with their great defense, but their shooting struggles will catch up with them eventually. Give me the Terrapins to make it to the Final Four.
Midwest Region: Unfortunately, this region features two of the favorites in the tournament to clash early on: Gonzaga and Houston. Gonzaga is being referred to as the most underseeded team in the tournament, and it’s obvious why. Kenpom has them 9th overall (as an 8 seed!), and the Bulldogs can be found in the bottom right in the trapezoid of excellence.
However, they have to run into the team (spoiler alert) that I have winning the whole thing. In a clash between contrasting playstyles of fast and slow, in March, I give the edge to the defensive juggernaut in Houston.
In the Elite 8, Houston vs Tennessee should be a defensive war, but I give the edge to the Cougars. They just beat out Tennessee in both offensive and defensive rating.
Elsewhere in the region, Xavier would be a nice team to pick for an upset, but I think Illinois also has good upside. I have the Illini advancing to the Sweet 16, knocking off Kentucky along the way, again in a battle of pace.
Final Four and National Championship: Michigan State and Maryland grade out very similarly in Kenpom, and both sit in the trapezoid of excellence. However, MSU sits just slightly higher in most categories. In another game with contrasting paces of play, I again give the edge to the slower team in Michigan State. I also believe they’ll be able to take advantage of their excellent rebounding, a stat Maryland was just average in.
VCU put together a good run, but it will not include making it to the National Championship. Houston is too good, too efficient, and VCU’s offense doesn’t stand a chance against a dominant Houston defense.
In fitting college basketball fashion. The National Championship will feature two defensive-minded teams, both with slower paces of play. Michigan State excels at preventing teams from making 3s, they are 2nd in the country in opponent 3-point percentage. They hold their opponents to just 28% from behind the arc. On the other hand, Houston is one of the most efficient teams from 3, shooting almost 40% and coming in at 4th in the country in 3-point percentage. It will surely be interesting to see who wins that battle in this championship game.
Ultimately, I have Houston as my National Champion in 2025. They are one of only 3 teams top 10 in both offensive and defensive rating. They force a ton of turnovers, and shoot the ball well.
We could easily look back at this in a couple of weeks and have my bracket be 40% accurate, and with the way things typically go in March, I wouldn’t be surprised. If that does happen though, don’t blame me, blame the numbers.
