Correctly picking the national champion is a critical part of winning your bracket challenge. And in a 68-team tournament known for unpredictability, that's typically a difficult task.

Favorites lose, sleepers rise and brackets―whether meticulously researched or hastily thrown together―can become crossed-out, pen-marked disasters after the first two rounds.

But even if your selections are exiting the tournament left, right, up and down, a decent consolation is having still been able to pick the champion.

How do you figure it out?

Pre-tournament analysis often focuses on several factors: roster experience, three-point shooting, rebound rate, turnover rate and tempo. Those measurements all provide valuable insight and are particularly useful in identifying the most likely early-round upsets.

Individually, though, they're not good indicators of championship success.

Over the last 17 years, one trend has shown to be unmistakable. In that span, only one national champion (2014 UConn) didn't finish in the top 20 for both offensive and defensive efficiency on KenPom.com.

Entering the 2019 tournament, seven teams fit the description.

Now, the immediate takeaway might be "hey, that's easy! Pick one team from among Virginia, Gonzaga, Duke, Michigan State, Michigan, North Carolina and Kentucky—and prosper."

But the trend is that the national champion finished the season in the top 20. That asterisk of "finished the season" keeps a handful of schools in the conversation.

Tennessee owns the nation's third-best offensive efficiency rating but is 34th on defense. Teams with similar profiles include Purdue (fifth, 32nd), Virginia Tech (11th, 25th), Buffalo (19th, 29th), Marquette (23rd, 35th) and Nevada (26th, 33rd).

Texas Tech highlights the other group, ranking first defensively but 36th on offense. Programs closer to that mold are Houston (24th, 13th), Florida State (31st, ninth), Kansas (34th, 15th) Maryland (27th, 22nd) and even Louisville (29th, 17th).

So, that leaves a list of seven favorites and 12 possibilities.

Let's get trimming.

During the last two decades, the national champion has held a top-three seed in every tournament except one. (That 2014 Shabazz Napier-led UConn team, a No. 7 seed, is a heck of an outlier.) Wave goodbye to Buffalo, Florida State, Kansas, Louisville, Marquette, Maryland, Nevada and Virginia Tech.

No national champion has ever recovered from a loss in the opening round of the preceding conference tournament. That eliminates Purdue and Texas Tech, pruning the field to nine.

This is where the blueprint gets tricky.

Every champion and all but four national runners-up in the KenPom era (since 2002) has finished with a top-30 strength of schedule. However, a deep run in March Madness organically offers an SOS bump. Plus, the four runners-up that didn't have a top-30 SOS were 2008 Memphis, 2010 Butler, 2011 Butler and 2017 Gonzaga. Playing in a non-high-major league shouldn't preclude Gonzaga or Houston.

Additionally, NBA-caliber talent, while beneficial, is not a foolproof factor. The absence of a projected top 2019 first-round pick shouldn't bounce Michigan, Michigan State or Tennessee.

What's next? Try history.

As valuable as defense is, the championship trend favors top offenses. In the KenPom era, only 2003 Syracuse, 2011 UConn and 2014 UConn ended the year lower than 10th in offensive efficiency. Yes, a deep tournament can swing the ratings, but Michigan (18th) and Houston (24th) are outliers for now.

Despite all of the program's recent success, Virginia has reached the Elite Eight once in Tony Bennett's previous nine years. Tennessee has never made a Final Four, and Gonzaga's streak of 20 straight NCAA berths has resulted in one Final Four appearance.

If you disagree with that method, it's understandable. History is not a perfect indicator of future success.

Nevertheless, we're down to a Final Four: Duke, Kentucky, North Carolina and Michigan State. Incidentally, Duke and Michigan State are the top-seeded teams in the East Region, while UNC and Kentucky are seeded first and second in the Midwest.

At this point, you might feel throwing a dart would be most accurate. Considering the balanced excellence of those four programs, there certainly must be worse strategies.

But in the 34 tournaments since the field expanded to 64 teams in 1985, a No. 1 seed has earned the national title 22 times. Ignoring that 64.7 championship rate would be foolish, so get ready for Round 4 of Duke vs. North Carolina.

Anyone have a coin we can borrow? Asking for a friend.

North Carolina swept the regular-season series, but Duke won the ACC tournament clash 74-73 when Zion Williamson was healthy. On a neutral court, it's practically a dead-even matchup. For the sake of making a pick, give us Duke and its ever-so-slightly more efficient offense and defense.

This result isn't surprising. Mike Krzyzewski's team is the betting favorite and the No. 1 overall seed. Duke also happens to most closely follow NCAA tournament trends as well.

Don't get too carried away.

Again, outliers happen. Last year, UMBC―the least-picked team in the first round on ESPN's Tournament Challenge―became the first No. 16 seed ever to eliminate a No. 1 team. No. 3 Michigan State's terrific shooters couldn't hit water if they fell out of a boat during a second-round loss to No. 11 Syracuse, going 8-of-37 from the perimeter.

We can hardly overstate the randomness of one-game samples. While that's what makes March Madness so captivating, even a perfect historical blueprint can be foiled on the opening weekend.

But if you like trends, stick to Duke and Carolina.

Statistics courtesy of KenPom.com or Sports Reference, unless otherwise noted. Follow Bleacher Report writer David Kenyon on Twitter @Kenyon19_BR.