Everybody loves an underdog, but it takes two teams to dance the toxic tango that leads a Giant to fall into a Killer’s clutches. So it’s high time that we train our statistical model on teams that could land in the top third of this year’s NCAA tournament brackets, and analyze the tendencies of those that could be vulnerable to upsets.

In general, to pull off unexpected wins, Giant Killers must play a high-risk, high-reward style, seeking to maximize the volume and value of their possessions by forcing turnovers, shooting 3-pointers and/or crashing the boards. And it’s precisely those tactics that successful Giants must squelch. Goliaths who stop Davids from going on runs can assert their own superiority and ward off upsets.

Statistically, that means offensive rebounding is the most important trait for overdogs looking to stay safe: Grab your own missed shots, and you not only keep the ball away from the other guys, but you can survive a subpar shooting night. Teams whose game centers on dominating the offensive boards (like North Carolina this season) have suffered the fewest upsets historically, and we call them Power Giants. But other styles work, too. Gambling Giants (such as Iowa) essentially give up on defensive rebounding to work the perimeter and force turnovers. Conversely, Pack-Line Giants (like Arizona) focus on controlling the defensive boards. Where high seeds tend to stumble is when they’re not strong at rebounding at either end or winning turnover battles. Such Generic Giants -- a group that includes both Maryland and Iowa State in 2016 -- lose nearly 40 percent of their matchups to Killers.

Our model rates Giants on a scale from zero to 100, based on the percentage chance it estimates they have to beat an average Killer. Before we get to the shaky Goliaths, a few notes:

West Virginia, whose entire game is built around generating offensive rebounds and turnovers, is the safest Giant of any team that will make the field of 68, with a Giant Rating of 97.38.

But Louisville (Giant Rating: 97.41) would have been the safest of all. Looks like the Cardinals may have given up more than they realized when they self-imposed a tournament ban for this season.

Virginia is the most vulnerable No. 1 seed, as projected by ESPN bracketologist Joe Lunardi, but still has a Giant Rating of 91.2. Which means that even in this season of parity, it’s unlikely, as in more than 97 percent unlikely, that a 1-seed will lose to a 16-seed. Especially since High Point, our favorite sleeper on the ocean floor of brackets, lost in the Big South semifinals on Saturday night. Farewell to John Brown, a great dunker who never got to play Giant Killer.

And now, on to the five most vulnerable potential Giants:

California Golden Bears (Giant Rating: 71.4): Subjectively, we would rather be rooting for Cal’s fantastic freshmen than against them, but our model sees several reasons to worry. The Bears are highly efficient at both ends because they can shoot from anywhere and work the boards, but turnovers are a problem. Cal loses the ball on nearly 18 percent of possessions, while forcing turnovers on just 13.8 percent of opponent possessions. That’s largely a result of their style -- their players talk about “getting our safeties back” after they get shots off -- but it’s extreme: The Bears rank 348th in the NCAA in defensive turnover percentage.

They’re also unusually reliant on free throws (41.7 percent of field-goal attempts) even though their foul shooting is poor (65.7 percent, ranking 317th). All of which means they have a hard time coming back from deficits: Oregon State, Stanford and Colorado all took early leads against Cal and went on to win, because the Bears couldn’t win turnover or free-throw battles. Though teams have taken few 3-pointers against Cal, the Bears are vulnerable on the perimeter, too, which Richmond exploited to great effect in a November upset (9-18 from long range in a 94-90 win). California’s cubs, who have won eight of their past nine games, may be coming together at the right moment. But the formula is there for an underdog to take them out.

Notre Dame Fighting Irish (Giant Rating: 77.0): Like fellow vulnerable Giant Maryland, Notre Dame has lost three of five, and those results might keep the Irish from being a Giant altogether. Lunardi lists the Irish as a 7-seed in his latest bracket, but a decent showing in the ACC tournament could easily vault them back into a first-round GK matchup. Our model says that wouldn’t be a relaxing experience for Notre Dame.

It’s easy to forget after last season’s run to the Elite Eight, but Notre Dame hasn’t fared well in recent NCAA tourneys. The Irish lost as a 7-seed in the first round in 2012 (vs. Xavier) and 2013 (vs. Iowa State). In 2011, they made it through the first round as a No. 2 seed, but got nailed by 10th-seeded Florida State’s slingshot in the second round. And in 2010, led by the great Blaine Taylor, 11-seed Old Dominion turned the Irish into a first-round slain Giant as well.

The problem in nearly each of those seasons was defense. And so it is again for the Irish, who rank 193rd in defensive efficiency, giving up adjusted 104.7 points per 100 possessions. A team can get away with bad overall defense if it finds ways to minimize opponents’ possessions. But the Irish don’t force turnovers—like, ever—with a 14.8 rate (337th in the nation). Nor do they do a particularly good job cleaning up after their opponents miss a shot, allowing a 30.3 percent offensive rebounding rate.

So, yes, Notre Dame is an offensive machine, ranking third in the nation with an adjusted 119.8 points per 100 possessions. Sure, Demetrius Jackson is playing as well as almost any point guard in the country, Zach Auguste is averaging a double-double, Steve Vasturia and V.J. Beachem are knocking down treys and Bonzie Colson gives them five starters scoring in double-figures. But the Irish have to get stops on a consistent basis, especially since every team faces a tough shooting night at some point in the tourney. Based on what the Irish have shown this season, that’s not about to happen.

Maryland Terrapins (Giant Rating: 77.2): If you plan to pick against Maryland next week, you’re likely basing that decision on the Terps’ late-season swoon. They’ve lost four of their past six, including an embarrassing performance at Minnesota. Simply put, Maryland isn’t playing well. But even if you eliminate recency bias, Maryland is plagued by much deeper problems. None matters more than the fact that the Terps get absolutely blasted in the turnover battle. You’d think that a team with a supposedly elite point guard in Melo Trimble would take care of the rock. But Maryland gives it away on 19.3 percent of its possessions. That’s 251st in the country. If our model had an empty gum wrapper, it still wouldn’t trust Maryland to hold onto it.

The Terps compound that problem with a defense that doesn’t pressure opponents. Maryland generates turnovers on just 16.6 percent of defensive trips, meaning that in most games, they are at a severe possession advantage. You’d hope that a big frontline that features Robert Carter, Diamond Stone and Jake Layman would offset those ball-handling problems with dominance on the boards, but Maryland is right around the national average on both ends, grabbing offensive rebounds on 29.9 percent of misses and allowing them on 29.3 of opponents’ bricks.

Now, we know you don’t always handle bad news very well, Terps fans, but we’re just the messenger here. And the message is simple: Your team is ripe for the picking against a viable underdog.

Utah Utes (Giant Rating: 77.6): It has been a fine campaign for the Runnin’ Utes, who finished the regular season with 24 wins and a second-place finish in the Pac-12. And for their efforts, Lunardi says they are in line for a 3-seed. There’s just one problem: Utah isn’t that good.

At least that’s what our model says. The basic power rating section of our GK formula says Utah is the 23rd-best team in the country, which means they are more the caliber of a 6-seed. And that’s important, because an underappreciated factor in generating upsets is that teams are often just seeded wrongly.

But Utah also faces some problems specifically tied to Giants that fall early in the tourney. Particularly concerning is a defense that doesn’t force turnovers, as Utah takes the ball away on just 15.9 percent of opponents’ possessions. Utah also doesn’t take advantage of its size as effectively as it should. Because the Utes try to control the pace and limit opponents’ transition game, they don’t crash the offensive glass particularly hard. And that’s a shame, because 7-foot Jakob Poeltl is as good a big man as you’ll find in the country, and he’s flanked by 6-9 Kyle Kuzma, who can go get the ball and put it right back in the basket. Utah has shot the ball extremely well this season, ranking 16th in effective field goal percentage (55 percent). But, again, being a safe Giant means being prepared for the worst-case scenario. And if they struggle to connect on some of those shots that they normally hit, it’s not clear that they can adjust and find another way to win.

Iowa State Cyclones (Giant Rating: 79.5): Make no mistake, the Cyclones can shoot from anywhere: All seven Iowa State players who get significant playing time have an effective field-goal percentage of 54.8 percent or better, which is amazing. But they pretty much have to. Steve Prohm’s crew doesn’t rebound at either end, and they don’t force turnovers, either, which means they’re allowing opponents an average of about 7 shots more than they take, night in and night out. So if Iowa State doesn’t execute at the very highest level for an entire game, it’s easy for them to trip. This season, for example, they’ve lost to Northern Iowa on a night when they shot 3-for-17 on 3s, and blown late leads against Texas Tech and Baylor in contests that went into overtime. Even when the Cyclones are in super sharpshooting mode, they can’t always blast out of the hole bored by losing the possession battle: They shot 11-22 from long distance, but still lost to West Virginia (which is basically Iowa State’s negative image) last month. 121.2 adjusted points per 100 possessions, the second-best scoring rate in the country, buys Iowa State competitiveness in nearly every game and a lot of fun, but not insurance.

Thanks to Liz Bouzarth, John Harris and Kevin Hutson of Furman University for research assistance.