The conventional wisdom among those that enter bracket pools is if you are looking for an upset in the round of 64, pick one of the No. 12 seeds to knock off a No. 5. Since 1985, when the tournament expanded to 64 teams, No. 5 seeds have been upset 47 times in 132 games, so that line of thinking by bracket pool pickers is justified. However, this year it could be the No. 10 and No. 11 seeds that bust brackets the best.

According to the Pomeroy rankings leading up to the tournament, which consistently have been one of the best metrics at predicting future wins this season, with the higher ranked team winning 72 percent of matchups, the average No. 12 seed is expected to outscore an average team by 13.1 net points per 100 possessions. The No. 10 and No. 11 seeds, by comparison, would be expected to outscore opponents by 14.7 and 15.4 net points per 100 possessions, respectively.

That becomes more important once you look at their first-round opponents. The average No. 5 seed has an adjusted net efficiency rating of 21.3, which means that on a neutral court, the 5-seed would be 8.2 net points per 100 possessions (21.3 minus 13.1) better than the 12-seed, implying a 72 percent win probability for the 5-seed. The average 6-seed has an average rating of just 17.1, which, when compared to an 11-seed, gives them an implied win probability of 53 percent. The 7-seeds have an average rating of 17.8, which, when compared to an 10-seed, gives them an implied win probability of 57 percent. Those last two are basically coin flips, which can pay off big in larger bracket pools.

This is where DAViD, the Data-Assisted Victory Detector, becomes so valuable. These win rates — adjusted for other factors such as any additional possessions to be had via offensive rebounds and turnovers — are compared to what we would expect based on looking at seed matchups in a vacuum, with stronger upset potential reflected in a higher DAViD score. This helps you separate from the pack in your bracket pools when you pick correctly.

For example, No. 10 Butler has a 57-percent chance of beating No. 7 Arkansas in the round of 64, significantly higher than the 38 percent chance we would assign to a 10-seed based on historical results. That gives Butler a DAViD score of 152, a very high mark, and one worthy of pushing the Bulldogs through to the next round on your bracket.

DAViD score Description
0-50 Not a great upset pick
51-100 Below-average chance to upset/poor value
101 to 125 Above-average chance to upset/good value
126+ Potential bracket buster

Here are the win probabilities and DAViD scores for the first-round matchups for the rest of the teams seeded at No. 10 or worse:

Read more on the tournament:

NCAA tournament cheat sheet: Everything you need to dominate your pool

Forget Kentucky and Arizona, Virginia has easiest road to the Final Four

Kansas, Xavier make No. 1 seeds the weakest since 2002. Will upsets follow?

The most vulnerable top seeds in the NCAA tournament

The three best bets to win the 2018 NCAA tournament