NCAA Tournament Bracket Picks: How to Use a Contrarian Strategy That Can Win

Filling out an NCAA Tournament bracket has become interwoven into the fabric of society. In January you make resolutions, in February you buy (or receive) chocolate and roses, and in March you fill out a bracket.

While this is a fun activity that has become a national pastime, this article is about finding an edge to win your bracket pool.

It may seem obvious that you want to win your pool, but do you actually fill out your bracket with that intention? We all aspire to finish on top when they cut down the nets at the end of the tournament, but a large majority of people are filling out their brackets using gut feeling or other irrational methods.

What a lot of people ignore is the dreaded four-letter word that many avoid like the plague: math. Given we know the percentages each team is being picked, we can compare those to our percentages in our Bracket Simulator.

I also assume that many of you are filling out your bracket incorrectly by picking the first round games, then second round games, and so on. Given the points system for most bracket contests, you should put most of your weight on the champion and go backward.

Most bracket systems double the points for each subsequent round, meaning that if a correct first-round pick is worth one point, then a second-round pick is worth two points and so on. This means that picking the correct champion is worth the same amount as getting all 32 first round games correct.

So while fretting over which 5 vs. 12 matchup is fun to do, you should try to get the most leverage by working backward through your bracket.

That is the strategy, now let's get to the picks, starting with the champion.

Fading Duke is the obvious play here. Even though our projections have Duke as the most likely winner, with nearly half of all brackets picking Duke to win it all you won't be gaining an edge against the rest of the field.

The choice comes down to Gonzaga or Virginia as both are projected to win the tournament more than people are picking them. In the end, I'll go with Gonzaga because it is projected to win slightly more often and I can still go heavily contrarian by putting Virginia in the championship game as well.

Note that we have two 1-seeds in the finals even though we are being contrarian. We don't have to go crazy and pick a Loyola to get to the Final Four to differentiate our bracket from the rest of the pack. By picking Gonzaga and Virginia, we have a final that we would expect to share with only 2.5% of all brackets entered.

We still need two more teams in our Final Four. Here are the top options in the East region:

Once more, we'll pick against Duke as it is being picked more than it should. We simply move down to the next line and take Michigan State who we would expect to get there 24.8% of the time, but are only being selected in 13.5% of brackets.

Here is the same data to win the Midwest Region:

This isn't exactly rocket science here. Our projections actually have Kentucky as the favorite to come out of this region, but most entries are lining up to take the chalk in North Carolina. It's rare, but Kentucky is actually being overlooked in this spot.

Now I repeat that process for each region, looking for spots in which the overall field is overweight or underweight on certain teams. Here is how the East Region ended up:

One of the key decisions is how early to knock Duke out of your bracket since it is the heavy favorite. Our projections have the Blue Devils getting to the Elite Eight nearly 80% of the time, but almost 95% of brackets have them getting at least that far. If Duke slips up early, we'll gain a ton of equity in our pools.

No real upsets in the first round as UCF is actually favored in our model. The masses are high on LSU and low on Maryland, so the Terrapins move on to the Sweet 16 in this scenario.

Users are heavy on Michigan in this region, so it goes down early to Nevada. Nearly 80% of brackets have Buffalo winning in the first round so it is a prime upset candidate, although it will be worth monitoring how those percentages fluctuate once the winner of the First Four game is known.

Marquette or Florida State in the Sweet 16 was the toughest decision in this bracket. Almost 58% of brackets have Florida State winning at least two games, which is just slightly higher than our projection of 55.5%

We have a couple of upsets early with Saint Mary's and Iowa being undervalued. Other than that, we have a relatively chalky region. Remember, we have already done most of the work being contrarian with our Final Four picks.

North Carolina and Duke are rivals on the court and in popularity in brackets this year as users are overwhelmingly plugging both in to make deep runs. Just like their Tobacco Road counterparts, I have the Tar Heels out in the Sweet 16.

Iowa State has nearly as good of a shot to get to the second week of the tournament as Houston, but entries are flocking to Houston so we'll go contrarian and take the Cyclones to win two games in the tournament.

That completes the contrarian bracket for this season. I want to reiterate that my plan is not to get every game correct, it is to set myself up for the best chance to win my entire pool. By knocking off some of the most popular teams early and going with good teams that are being overlooked, I am maximizing my opportunity to win while not having to pick a ton of upsets.

While it is fun to try to predict the next Loyola or Butler, those teams are likely to bust your bracket early while the majority of your opponents are going to be in decent shape with the chalk.