March Madness bracket tips: Are making massive upset picks smart?
The best part about March Madness can be its unpredictability.
Yet, it’s that same quality that makes filling out a bracket so difficult.
There’s a reason it seems like every year a team you have slated to advance far in the tournament goes out in the first round: It actually happens.
We looked through the data from the past five years of the Capital One Bracket Challenge Game and examined how many users picked each team to advance in the traditional first round of the NCAA tournament. Interestingly, every year there was at least one team picked by the overwhelming majority to win its first-round matchup that ultimately lost.
In total, five teams that were among the 10 most selected teams to advance past the first round in a given year were eliminated -- 2012 Duke, 2012 Missouri, 2013 Georgetown, 2014 Duke and 2015 Iowa State.
Let's take a closer look at last season's first-round action.
|Team (in order of most-picked by users to win first-round game)||Result|
|Notre Dame||W, 69, 65|
|Kansas||W 75, 56|
|Iowa State||L, 59-60|
Bracket Challenge Game users showed supreme confidence in Iowa State, as only nine teams were more popular to advance past the first round than the Cylcones. However, unfortunately for the majority of Bracket Challenge players, UAB took down Georges Niang and Co.
So is the takeaway that we need to pick one of the tournament’s favorites to be knocked out in the first round? It happens every year and you’ll gain a big advantage since such few people are making these, right?
It’s still way too risky to make a selection of that nature.
|MARCH MADNESS ON SOCIAL MEDIA|
JOIN THE TEAM.
If you’re able to buck the odds and pick one of those low-probability upsets, well done. You can rightfully brag to all of your friends that you correctly predicted the tournament’s biggest upset.
But, even then, that’s all you’re going to be able do with that that amazing prediction; At the of the day, it really isn’t going to help you have a great bracket. As our data shows, it’s correctly predicting the final rounds of the tournament that brings about the best score rather than nailing the first round.
If these types of picks paid dividends in the final rounds, they could be worth it. But of our five upsets, three came from 15 seeds and two came from 14 seeds. And as crazy as March can get, only one 15 seed and two 14 seeds have made it to the Sweet 16 in NCAA history, and none have advanced further. On the other hand, the teams that you’re getting rid of in your bracket to make these massive upset picks are making it deep in the Big Dance.
Looking through all of our data, nearly half the Elite Eight teams, 40 percent of the Final Four and two of our champions from the past five years were made up of the top-10 teams selected to advance in the first round each year. As seen below, in 2015 alone, three Final Four teams and both the runner-up and champion were top-10 selected teams.
|Team (in order of most-picked by users to win first-round game)||Tournament finish|
|Notre Dame||Elite Eight|
|Kansas||Round of 32|
|Villanova||Round of 32|
|Virginia||Round of 32|
|Iowa State||Round of 64|
It turns out picking a massive upset is a lose-lose proposition. If you’re simply just picking, say, a No. 15 seed to beat a No. 2 seed and have them advance no further, you’re making a low percentage pick (10 percent) for something that’s not going to play a huge role in how your bracket does (due to the final rounds having much greater importance).
And if you're making this upset pick and then also predicting that team to advance deep in the tourney, well, that's not wise either. As we’ve talked about, there have been 248 No. 14 seeds and No. 15 seeds in NCAA history, and none of them had made it to the Elite Eight. Furthermore, you’re ridding yourself of a team that is very likely to advance further.
Every college basketball fan remembers where they were when Lehigh toppled Duke in 2013 and Florida Gulf Coast took down Georgetown in 2014. Enjoy these upsets, but just don’t go out and try to predict them. Yes, they will help you differentiate your bracket, considering our data shows that only a small percentage are making them. But your chances of being correct are very slim and, in regards to building the best bracket, the benefit is minimal and the risk is massive.