There’s a version of Napheesa Collier that’s more disruptive than the one who tore three ligaments in her ankle during last year’s WNBA semifinals, and Minnesota didn’t need to wait for full health to get her back. Three games into her return from double ankle surgery, the Lynx forward is scoring, rebounding, and stealing at a higher rate per minute than she did during the 50/40/90 season that finished second in MVP voting. Fewer minutes. Better math.
The injury itself was ugly. A collision with Phoenix’s Alyssa Thomas in Game 3 of last year’s semifinals left Collier with three torn ligaments and a Grade 2 sprain, and she had surgery on both ankles over the offseason. She missed the first 27 games of this season, close to ten weeks, while the Lynx built a season around her absence instead of her return.
Her actual comeback broke as news before it broke as basketball. Here’s the confirmation beat writers had queued up before most fans found a stream:
https://x.com/AnnieCostabile/status/2079943491228938647
The box score explained why it mattered more than a routine one-game return. Collier’s line against Seattle read 24 points and 10 rebounds in just 22 minutes. Two games against the Toronto Tempo followed: 15 points and six rebounds in 25 minutes, then 16 and five in 22, pushing her three-game averages to 18.3 points, seven rebounds, two assists, and 1.7 steals a night, all in barely 23 minutes a game.
How good was Napheesa Collier before the injury?
Before the injury, Collier was the WNBA’s most complete offensive weapon: 22.9 points, 7.3 rebounds, and 3.2 assists a night across 33 games, on 50/40/90 shooting splits, only the second such season in league history and the first at 20-plus points per game. Hitting 50 percent from the field, 40 percent from three, and 90 percent from the line in the same season means there’s no easy way to defend her; sag off and she shoots over you, close out hard and she’s already blown past you to the rim. She finished second in 2025 MVP voting off that profile. That’s the bar her three-game sample is being measured against, and on a per-minute basis, it’s clearing it.
What do the per-minute numbers actually show?
Adjusted to a per-36-minute rate, Collier’s scoring, rebounding, and steal numbers are all up from her pre-injury season, while her assist rate has dipped slightly. Three of four core categories point in the same direction, which is the more surprising finding than her health clearing at all.
Scoring: her per-36 rate has climbed from 25.5 points to 28.6, roughly the gap between an above-average starter’s workload and what an MVP finalist looks like once you strip out a full 32 minutes of fatigue. Rebounding: 8.1 boards per-36 before the injury, 11.0 now, close to three extra possessions ending in her hands every 36 minutes. Steals: 1.8 up to 2.7, nearly a full additional takeaway. The lone number moving the other way is assists, down from 3.6 to 3.1 per-36, a small dip that lines up with a Lynx offense leaning more on Olivia Miles’ All-Star rookie season and the Lynx’s historic backcourt pairing to create shots while Collier reacclimates.
Is three games enough to believe this?
No, not on its own. Three games is 69 minutes of basketball, a sample size that would get you laughed out of an actual stats class, and it comes with an obvious asterisk: rust looks different for every player, and short hot stretches happen constantly across a 40-game season.
But the honest read gets more interesting once you look at the direction of the data, not just the size of the sample. Rust usually shows up as decline: missed rotations, tentative closeouts, hesitation on shots a player used to take without thinking. Collier’s numbers are running the opposite way in three of four categories, and that consistency is what earns a second look before anyone extends her minutes. I ran the comparison expecting a mild uptick in one category and a quiet stretch everywhere else. That’s not what came back, and the direction, not the sample size, is the part worth tracking as Cheryl Reeve loosens the leash over the next few weeks.
What does this mean for the Lynx’s stretch run?
Minnesota enters August at 24-6, the best record in the WNBA, riding a nine-game winning streak that started before Collier ever checked back in. Her per-minute spike doesn’t explain that record. It raises the ceiling sitting on top of it, which is the more inconvenient problem for everyone else chasing them down the stretch.
Reeve has kept Collier in the 22-to-25-minute band through her return, nowhere close to the 34.7 minutes a game she averaged during her 2024 peak season. That’s not caution born of doubt. It’s caution applied to a formula that’s already working, which means the real test arrives whenever Reeve decides the ankles can handle a normal workload again. If a 23-minute version of Collier is outproducing her MVP-runner-up self on a per-minute basis, the math for October changes the moment that restriction lifts, and every other team building a stretch-run case around health, chemistry, or a hot streak now has to account for a Lynx team that hasn’t even unlocked its best player yet.
Give the Lynx a healthy Collier at 30-plus minutes a night in the playoffs, running these per-minute numbers instead of leveling off as the workload grows, and this stops being a nice recovery story. It becomes the reason Minnesota is the team nobody left in the WNBA actually wants to see in October.
