Advanced metrics have made a strong case that long two-pointers are the least efficient shots in basketball.

But they've probably never been less efficient than they were at a recent Timberwolves scrimmage, when new interim head coach Ryan Saunders had a rule: Any long two-pointer attempted — make or miss — was worth minus-1 points.

Sink what you thought was a game-winning 22-footer with the score tied? Sorry, you just lost.

Saunders said he's tried a few different things in scrimmages so far to shake things up and "try to condition ourselves a little bit and try some new things" to keep things fresh.

"We're in January. The season can get long at times, so we're giving guys different things that they can think about in terms of scrimmages," Saunders said after Monday's practice. "But we do want to emphasize more three-point shots as well as layups, as those things statistically are better shots."

The math on long twos vs. threes is pretty simple. According to Basketball Reference, NBA teams this season are shooting 39.4 percent on long twos (defined there as shots between 16 feet and the three-point line) and 35 percent on threes. Three-pointers are worth one more point (you're welcome), which more than makes up for the small negative difference in percent made.

The Wolves this season have actually shot better from three-point range (35.5 percent, No. 13 in the league) than they have on long twos (34.6 percent, second worst in the NBA).

The Wolves already have started making some positive strides away from inefficient shots this season before Tom Thibodeau was fired. They were attempting about 28 threes per game, up from their league-worst 22.5 last season, and had cut down on their long twos.

But they still sit in the top 10 this season of long twos attempted by percentage of overall shots and lag in the bottom 10 of three-pointers attempted by percentage of overall shots.

So clearly this emphasis is one difference between Saunders and Thibodeau — as well between Ryan and his father, the late Flip Saunders.

In Flip's final season as head coach in 2014-15, the Timberwolves were tied for the second-highest percentage of overall shots taken that were "long twos" and last in the NBA and percentage of shots from three-point range.

That said, Ryan Saunders defines "long two" a little differently than most stat sites, essentially seeing it as a jumper from between the college three-point line (20 feet, 9 inches) and the NBA line (23 feet, 9 inches at the top of the key).

"An elbow jump shot isn't necessarily considered a long two," he said.

It's early, but perhaps the message is getting through. In Saturday's 110-106 victory over New Orleans, the Wolves attempted 39 three-pointers. Examining boxscore data indicates there were only five attempts that fit Saunders' criteria as a long two (the Wolves made just one of them, but it did count for two points instead of minus-1).

The goal isn't to make players question their shot selection but to drill into their heads that efficiency matters.

"Ultimately these guys have been playing basketball their whole lives and if they have a good shot, we're confident in their shots," Saunders said.