Cracking the NCAA Tournament Bracket, Part 1: Myths and Common Beliefs

March Madness? Is it really “madness?” Or maybe, just maybe, is it actually “March Rhyme and Reason?” 

People have been trying to figure out how to predict the NCAA Tournament for decades. So far, so wrong. We all enter pools with friends, family, co-workers and complete strangers. And we all wish we had the magic formula. Some of us have a specific way of making our picks, and we never waver. A lot of us change it up every year, or submit multiple entries, hoping that it’s a numbers game.  

I’m someone who watches the maximum amount of college basketball possible. I read about it, write about it, talk about it, think about it…a lot. From about December 1 through Selection Sunday, it’s fair to say I’m in a serious, committed relationship with college basketball. Yet I’ve never won a bracket contest. Last season was the first time I remember picking the correct champion.

Last year, I teamed up with a revered Excel Spreadsheet Master, a highly sought-after engineer in his field and an all-around brilliant fella (those accolades all describe the same friend of mine) to see if we could make heads or tails of the brackets. I spent countless hours compiling statistics from the past ten years of NCAA Tournaments; looking at the Final 4 teams, the teams who got upset, the teams doing the upsetting, and more. Then my ‘numbers guy’ ran a series of regression analyses. We looked at the data, compared it with the teams in the 2013 Tournament, and then decided which upsets were most likely to happen and which teams were most likely to reach the Final 4.

And, we were way off. Wayyy off.  

A few weeks ago, I googled “Predicting NCAA Tournament” and landed on this article about a couple of hotshot Yale professors who, back in 2007, attempted to use a gang of statistical data to predict that year’s Dance. After all of that number crunching, that team came up with a 58% success rate.  

In other words, those big shot Yale dudes did about as well as your pen pal in the United Arab Emirates would do.

This season, I’ve undertaken a different approach. Over the course of the next few posts, culminating with my predictions once the brackets are announced, you’ll get a behind-the-scenes look at what really happens in these brackets. We’ll cover myths and trends. You’ll see that “huge upsets” aren’t as random as you might think. You’ll know which teams are dangerous and which teams are vulnerable. You’ll have a game plan to maximize your bracket’s points. And, hopefully, you’ll be submitting the winning bracket in your group. (Or winning that billion bucks.) 

Let’s get started with myths and trends.

Myths

 Myth #1 – Girls are better than guys. You hear all the time that “My wife always does better than me in those things.” Well, that may be the case some of the time, but there’s no evidence available that girls are better than guys at this. Sometimes a girl can get lucky based on her picking strategy. For example, she could say, “Ooh, Pittsburgh. Yuck. I would never go to that city. I’m picking them to lose.” Or “Oh I love those Oregon uniforms! The Ducks! I’m picking them.” Or “Aww, Kent State. They had that shooting tragedy. They deserve to win.” But generally, I’ve seen girls do well, and I’ve seen girls do poorly. So guys, if you think the winning strategy is simply to “let my girlfriend make my picks,” you’re not assured of a damn thing.

Myth #2 – People who watch a bunch of college basketball are at a disadvantage against people who watch very little or none. This one probably has some truth to it. I’m living proof, considering who’s won some of the pools I’ve been in. I think, generally, the avid watchers simply overanalyze the matchups and focus on the wrong criteria when deciding which team will beat another. Biases play a role, too—like if you watch a team play a lot one year, you either focus too much on its shortcomings or give too much credit for a specific game you watched. That said, the “let my dog make my picks” versus doing it yourself will only yield better results in rare instances.

Myth #3 – If everyone and their mama are picking a specific upset to occur, it probably won’t occur. An example I remember was back in 1994 when Ohio University had Gary Trent, “The Shaq of the MAC,” and they were a 12-seed matched up against 5-seeded Indiana. Everyone thought OU would pull the shocker, but IU won by 12. Truth be told, upsets are unlikely to occur anyway. It has nothing to do with whether it’s a popular one to pick or not.

Trends

A common mistake we make (myself very much included) is that when we think back over the last few years, we think we have some things figured out. Things like:

• Team X always goes far

• This conference always does well/poorly

• This team went out early last year, so they’ll be extra-motivated this year

• This coach’s teams never go far

Then, those of us who watch a bunch of hoops and who really dive into each team’s stats and results start to focus on certain criteria like: 

• Team X won that game at Team Y. Nobody ever wins in that building. Team X is tough!

• Team X won its conference… or Team Y only went 10-8 in their league?

• Team Z is loaded with experience; those guys will be ready, and they won’t want their careers to end/ Team Y has too many freshmen; they won’t be able to handle the pressure

• Team X is in the Top 10 nationally in points allowed. And “defense wins championships!”

• Team Y makes their free throws/Team Z doesn’t make their free throws

• Oh man, Team Z is on a roll. They’ve won 14 of their last 15. They’re hot!

• What is up with Team X? Only 5-5 in their last 10? Not playing good basketball.

• It just seems like Team Z’s year 

When I went back and looked at the past five NCAA Tournaments, I realized that most of these “trends” were false beliefs.

Let’s return for a moment to statistics. Earlier I covered how far off my colleague and I were last season when we based our predictions solely on raw data; and the fact that those two Yale math whizzes didn’t prove much of anything with their methods. That said, my approach this season does not completely ignore stats. In some cases, a specific stat tells the story. A single stat that was overshadowed by the accomplishments and make-up of a team throughout a season would have, if identified before the games began, been a big shiny red flag. 

Take the stunning 2012 first round upset when 15-seed Norfolk State knocked off 2-seed Missouri. Missouri was 30-4 that season and had the best offense in the nation according to KenPom.com. The Tigers didn’t have a single bad loss all season. Mizzou won the Big 12 conference tournament, a strong feat considering the caliber of that league. And, probably most importantly, this team was loaded with tremendous guards—guards with quickness and inside-outside capabilities. As most experts will tell you, great guardplay is generally the common theme among teams that make deep runs.  

So how in the bald head of Anthony Peeler did Missouri lose to NORFOLK STATE in the first round? Well, the Tigers, despite all of their success and talent, forgot about one thing: Defense. According to KenPom, Mizzou ranked 146^th in the nation in adjusted defense. In other words, they couldn’t get stops down the stretch. Norfolk State continued to score, and as the game drew to a close, the pressure fell solely on Missouri. In a game that shouldn’t have been close, Norfolk pulled it out against Missouri’s shoddy D. (By the way, in the next round, Norfolk scored just 50 points against 7-seed Florida, a team perennially in the Top 10 in adjusted defense—just so you know.)

So, was that a random upset after all? I don’t think so. Easy to predict? Of course not. But, had you looked at the entire picture with Missouri, and discovered that Swiss cheese defense, you may have predicted they’d lose to Florida in the 2^nd round, and that could have saved your bracket from losing a Final 4 pick in the first weekend. (You get more points for the later rounds than the first couple.)

Alright, I’ll end Part 1 here. Digest this for a few days. Over the weekend, I’ll release the next post, which will cover the following:

• The actual trends we’ve seen recently in the NCAA Tournament

• Beware of overseeded and underseeded teams (and how to tell)

• Why certain TYPES of teams generally do not advance

• Why the TYPE of guards a team has makes a difference

• Why the TYPE of offense and TYPE of defense a team employs can spell vulnerable or dangerous

• Why what a team does in conference play doesn’t mean a hill of beans

• Foreshadowing/Which stats actually matter

And then, after the field is selected, I’ll release my predictions to the public. Have a great week, thanks for reading, and enjoy your March Madness!