In 2018, it finally happened. No. 16 Maryland Baltimore County (UMBC) shocked everyone with their 74-54 upset over No. 1 Virginia last year, the first victory ever by a No. 16 seed in the Round of 64, and now everyone is going to be looking for a sign it could happen again.

Upset selections on this seed line have been increasing as of late. According to data pulled from the NCAA’s Bracket Challenge Game, over two percent of brackets picked UMBC over Virginia and a record-high eight percent of all brackets had another No. 16 seed beating a No. 1 seed in the first round; six percent of brackets picked an upset on this line in 2015.

The temptation is understandable but it really is a fool’s errand to try to catch lightning in a bottle two years in a row. After all, the No. 1 seed enjoys a 135-1 record over the No. 16 seed since 1985, the first year the field expanded to 64 teams. Since 2011, the first year for the current 68-team field, the No. 1 seeds have outscored the No. 16 seeds by 20 or more points in 18 out of 32 contests. Just four of those 32 games, not including the win by UMBC last year, were decided by nine points or less.

But if you are going to throw caution to the wind and select a No. 16 seed to do something miraculous again, here is some information to help you make a selection.

No. 1 Virginia vs. No. 16 Gardner Webb

Chance at upset: 0.5 percent

Gardner Webb, fresh off winning the Big South Tournament earlier this week, makes its NCAA tournament debut and is one of the most-talented No. 16 seeds in the Big Dance since 2011, the first year the field expanded to 68 teams: the Bulldogs rank 167th in adjusted efficiency according to Pomeroy. That said, by comparison UMBC was more than twice as likely to beat Virginia last season, sporting a 1.2 percent chance to win.

However, the Bulldogs are awful defensively (253rd in the nation), allowing a whopping 45 percent of field goal attempts against to come beyond the three-point line. And their offense isn’t potent enough to beat Virginia like UMBC did in last year’s tournament: Gardner Webb attempts a low volume of shots from behind the arc (37 percent, 207th) and doesn’t grab many offensive rebounds (23 percent, 330th).

Plus, since 2011 the Big South is 0-7 in the Round of 64.

No. 1 North Carolina vs. No. 16 Iona

Chance at upset: 0.5 percent

Iona is ranked lower than Gardner Webb per Pomeroy with a worse defense (275th) that is terrible at stopping teams in the half court. According to Synergy Sports, the Gaels allow 0.9 points per possession in these sets (22nd percentile) with below-average marks against the ballhandler on the pick and roll (20th percentile), putbacks off offensive rebounds (19th percentile) and vs. spot-up shooters (5th percentile).

North Carolina, meanwhile, excels in spot-up situations (97th percentile), which account for a team-high 25 percent of their offensive opportunities.

No. 1 Duke vs. No. 16 North Dakota State/North Carolina Central

Chance at upset: 0.4 percent for North Dakota State, 0.04 percent for North Carolina Central

Duke, the top seed in the tournament, plays the winner of the First Four game between North Dakota State and North Carolina Central, two teams just happy to be here.

If North Dakota State makes it through to the Round of 64, Duke will get to feast on one of the worst defenses in the country (278th) that also fails to create turnovers (15 percent, 347th) or steals (7 percent, 312th), two traits needed for an upset of this magnitude.

If North Carolina Central prevails, the Blue Devils will have the honor of dispatching the 303rd best team in the nation, according to Pomeroy’s ratings. In fact, it is quite possible the Eagles beat themselves: they turn the ball over 23 percent of the time on offense, the 11th highest rate among Division teams, giving Duke extra possessions that could cause an avalanche of points against.

No. 1 Gonzaga vs. No. 16 Fairleigh Dickinson/Prairie View A&M

Chance at upset: 0.2 percent for both

Fairleigh Dickinson’s lack of depth is a huge liability. Coach Greg Herenda doles out 20 percent of the team’s minutes to his bench, the fourth-lowest rate among tournament teams, forcing starting guards Jahlil Jenkins and Darnell Edge to play timid, resulting in a below-average 0.6 points per possession on drives to the basket (22 percentile). The Knights also allow opponents to grab a third of their misses, one of the worst rates allowed in the country. Imagine what Gonzaga’s offense, the top in the nation, could do with a few extra possessions.

Prairie View will appear in the NCAA tournament for the second time in the school’s history but a lack of size (average height is 75.2 inches, seventh-smallest in Division I) and an inability to shoot (49 percent effective field goal rate, 263rd) will not bode well for them this time around.

Read more: