2022 NCAA Tournament Bracket Strategy from TeamRankings
This is a guest post from TeamRankings.com.
Every year, an average of 57% of their subscribers win a prize using their NCAA bracket picks.
To win a bracket pool, you have to score more points than your opponents. You need to get at least some picks right that your opponents miss.
However, a common misconception with bracket pools is that being contrarian requires picking upsets. In multi-round contest format like the NCAA tournament, if you make a bold early upset pick and get it wrong, missing out on points that your opponents continue to rack up can doom your entry. Being a bracket pick contrarian doesn’t always mean picking longshots.
At the same time, a contrarian set of Final Four picks may offer great value against the public, but the probability of it happening may be so low that you reduce your chances to win. The traditional bracket pool scoring system (1-2-4-8-16-32) puts a huge emphasis on late-round picks, and you typically need to get a few of them right to have a shot at winning.
How the Public Picks the Top Seeds
Thankfully, it’s not always the longshots that provide contrarian value in bracket contests, for two reasons. First, some top seeds have been great value plays in recent years. Second, some lower-seeded teams that players might perceive to have contrarian value have actually been overvalued.
The table below is based on bracket pool pick popularity data that we have accumulated for every tournament since 2010. It shows the percentage of time the public picks a team to win in each round by seed number, for No. 1 through No. 6 seeds:
| Seed | Round 1 | Round 2 | Sweet 16 | Elite 8 | Final Four | Champ |
|---|---|---|---|---|---|---|
| 1 Seeds | 98% | 89% | 70% | 46% | 27% | 15% |
| 2 Seeds | 97% | 82% | 54% | 25% | 12% | 5% |
| 3 Seeds | 93% | 68% | 30% | 11% | 5% | 2% |
| 4 Seeds | 89% | 58% | 17% | 8% | 4% | 2% |
| 5 Seeds | 75% | 32% | 8% | 3% | 1% | 1% |
| 6 Seeds | 71% | 23% | 7% | 2% | 1% | 0% |
The data above shows that an average No. 1 seed was picked to win the NCAA title 15% of the time in bracket pools nationwide.
(Granted, there is plenty of variation around that average, with some No. 1 seeds picked as NCAA champion by 30% or more of the public, and other No. 1 seeds picked even less often as champion than some No. 2 seeds.)
In the last four NCAA tournaments, here’s how popular some specific subsets of teams were as NCAA champion picks in bracket pools:
| Type of Team | Frequency as Champion Pick |
|---|---|
| Most popular No. 1 seed | 29% |
| 2nd most popular No. 1 seed | 15% |
| 3rd most popular No. 1 seed | 10% |
| Most popular non-No. 1 seed | 8% |
| Least popular No. 1 seed | 7% |
| 2nd most popular non-No. 1 seed | 6% |
As it turns out, the group in second-to-last row in the table above (the least popular No. 1 seeds) has produced both a national champion (Virginia in 2019) and a finalist (Gonzaga in 2017) in the past four years.
In comparison, none of the most popular No. 2 seeds have even reached the Final Four in that same time period.
Finding Value at the Top of the Bracket
With our data set, we can also calculate the number of wins that the public has expected a team at each seed level to get, on average, since 2010.
When we compare that public expectation to actual performance, the results are pretty eye-opening:
| Seed | Actual Wins | Public Expectation | Difference | Percent of Expectation |
|---|---|---|---|---|
| 1 Seeds | 3.18 | 3.45 | -0.27 | 92.2% |
| 2 Seeds | 2.18 | 2.74 | -0.56 | 79.6% |
| 3 Seeds | 1.81 | 2.08 | -0.27 | 87.0% |
| 4 Seeds | 1.59 | 1.77 | -0.18 | 89.8% |
| 5 Seeds | 1.09 | 1.19 | -0.10 | 91.6% |
| 6 Seeds | 0.73 | 1.04 | -0.31 | 70.2% |
One would probably expect there to be a “value tax” for picking the very top teams in the bracket, compared to taking some additional risks on lower seeds. That rightmost column in the table above isn’t a perfect way to measure that tax, but it certainly throws a question mark on that hypothesis.
What’s most interesting is by this metric, it’s not the No. 1 seeds that have been most overvalued as a group, but the No. 2 seeds, and by a pretty big margin. And while the No. 3 and No. 4 seeds have a more similar value tax to No. 1 seeds, they also have less upside since their total wins expectation is lower.
A lot of bracket pool players blindly assume that putting two No. 1 seeds in the final game is nowhere near contrarian enough. Instead, they end up picking a team that not only has a lower chance of making it, but is also a worse relative value when you factor in pick popularity.
(By the way, in the last six NCAA tournaments, the final game has featured a No. 1 seed vs. a No. 1 seed half of the time. If picking that outcome is on your “never do” list, you’re really shooting yourself in the foot.)
How Contrarian Do You Need to Get?
A lot of people think you have to make a contrarian NCAA champion pick to have the best shot to win a bracket contest.
That’s often the case in huge pools where you need to beat hundreds or thousands of people. But in the friends-and-family pools and smaller office pools many people play in, it’s far from the truth.
Look back at the first chart in this article showing the percentage of time the public picks certain seed numbers to win in each round. The average No. 1 seed was picked 15% of the time to win the title game, and another 12% of the time to lose in the title game (since it was picked to win in the Final Four 27% of the time).
Now imagine you pick two No. 1 seeds, two of the favorites to win the tournament, to meet in the final game.
You would expect, on average, that about 2% of entries in your pool would share that exact same combination of title winner and title loser (15% x 12% = 1.8%). Even if you went with the most popular NCAA champion pick and another one of the top seeds, you would expect only about 4-5% of brackets in your pool to share the same two Final Four winners and NCAA champion winner.
In the common 1-2-4-8-16-32 scoring format, those last three games represent one-third of the total available points in the pool. If those “super chalky” picks of yours come through, even in an 100-entry pool, you’ll effectively be heads up against only a handful of other entries you need to outscore in the earlier rounds.
The public likes to enforce arbitrary quotas on how many top seeds they pick to make the Final Four (two of them on average) and the title game (barely over one on average). Because of that irrational bias, you can sometimes find contrarian value sticking with No. 1 seeds, while also having some of the best odds to score points.
The Best Bracket For Your Pool
While the research we compiled for this post helps to dispel some popular yet ill-informed “golden rules” of bracket picking, it’s even more critical to understand that every NCAA tournament is different.
Even though recent data shows that No. 2 seeds have been the most overvalued seed line, it could be the case that in 2022, a particular No. 2 seed has a great chance to win it all and is flying somewhat under the public’s radar.
That’s where we come in. Our data-driven expert bracket picks have generated over $1.75 million in bracket pool prizes since 2017, using technology so sophisticated that WIRED Magazine wrote about it.
In minutes, our bracket optimizer tool generates ready-to-play brackets that adjust for team advancement odds and pick popularity, plus key strategy factors like the size of your pool, its scoring system, and its payout structure.
We also offer the most comprehensive set of objective predictions, analysis tools, and data on the 2022 NCAA tournament, plus tools for competing in NCAA survivor and Calcutta pools.