Ionescu Goes Off and the Depth Chart Did the Work

Two games on the board Monday, and the story in both of them was less about who was supposed to carry the load and more about who ended up with the keys. Across New York and Los Angeles on one side, Portland on the other, injuries reshuffled rotations before tip and the players who absorbed those extra minutes made them count. That is not a coincidence — that is how the WNBA works, a league where role definitions are fluid enough that a few lineup changes can turn a mid-salary player into a slate-winner by the third quarter.

Sabrina Ionescu dropped 27 points, 7 rebounds, and 6 assists in 36 minutes and looked like herself — fully, uncomplicatedly herself. On the LA side, the beneficiaries of bumped minutes spread the wealth in a way that made tournament builders happy: production scattered across four names, a couple of them at salaries where you barely noticed the spend. And in Portland, a couple of depth pieces quietly put up numbers that will make people wonder why they weren't in more lineups. The short answer: sometimes the model sees it and the salary hasn't caught up yet.

The one-number receipt for the night: a mean absolute error of 6.6 DraftKings points across 31 graded players. That is an honest number — not a brag, not a confession, just where the model landed on a two-game slate with significant injury news baked in late. Here is how it broke down by name.

The Top Scorers

• Sabrina Ionescu (NY, $10,300) — 43.8 DK points. We had her at 35.1. An injury bumped her minutes to 36, and she delivered a 27/7/6 line that pushed her well past the projection. The direction was right; the magnitude was better than expected. That is a good problem to have.
• Rae Burrell (LA, $7,700) — 39.8 DK points against a projection of 26.6. An injury bumped her minutes, she played 34 of them, and then she went and added 4 steals on top of the 24-point, 3-assist night. Steals are the most volatile counting stat in the game — you can project the minutes, you cannot project the hands.
• Erica Wheeler (LA, $7,000) — 33.3 DK points, projected at 25.9. An injury opened up her minutes and she found 9 assists in 27 minutes. Nine. That is a number that bends a stat line in a hurry, and at $7,000 it made her one of the better values on the slate.
• Ariel Atkins (LA, $7,100) — 30.8 DK points, projected at 26.0. An injury bumped her minutes to 34 and she produced a steady 16/5/5 line with a steal. Clean, unsexy, exactly what you want from a mid-salary piece.
• Frieda Buhner (POR, $4,500) — 30.3 DK points against a projection of 11.9. An injury bumped her into 21 minutes and she went for 16 points, 5 rebounds, 2 assists, and 2 steals. The miss here was not directional — we knew minutes were moving her way — it was magnitude. A $4,500 salary with that kind of ceiling is the definition of a GPP dart.
• Megan DiLeo (POR, $7,900) — 29.0 DK points, projected at 21.6. An injury boosted her run and she delivered 20 points and 6 rebounds in 27 minutes. Good call, solid execution.

Where We Hit

Rae Burrell is the best call of the night in terms of process. We identified the minutes bump, built in the elevated projection, and while she outperformed even that number, the logic held. Sabrina Ionescu is the best call in terms of outcome — 43.8 actual against 35.1 projected, with the model flagging the minutes increase as the catalyst. And Megan DiLeo is the quiet one: not the flashiest number on the slate, but a clean hit on a player whose opportunity was correctly identified before tip.

Where We Missed

Let's not bury it. Carla Leite (POR, $8,800) — projected at 36.8, delivered 16.0. An injury had bumped her minutes in our model, but she played only 22 and finished with 11 points, no rebounds, and 3 assists. That is a 20.8-point miss on a player priced like a centerpiece. We own that. The injury news created a projected opportunity; the actual rotation did not cooperate. It happens, it stings, and it is the kind of result that is already informing how we weight minute-bump signals for players in murky depth-chart situations going forward.

A few things that will carry into the next projection cycle.

1. Not all minute bumps are created equal. The LA bumps — Burrell, Wheeler, Atkins — produced four players who outperformed projections on a slate where the minutes increase was clean and the role was well-defined. The Portland situation around Leite was murkier, and the model paid for trusting the signal too heavily. The distinction going forward is how clearly the new opportunity is defined, not just that minutes are available.
2. Steals and assists are the projections that age the fastest. Wheeler's 9-assist night and Burrell's 4-steal game were the biggest individual gaps between projection and outcome. Both came from genuinely volatile counting categories — the kind that are nearly impossible to project with precision but can completely reshape a DK score. That is not a flaw in the model; it is an honest feature of the sport. Know it going in.
3. The low-salary, minute-bump play is a GPP play, full stop. Frieda Buhner at $4,500 going for 30.3 DK points is a beautiful outcome, but the projection was 11.9. That gap — 18.4 points — is not a model failure, it is a reminder that players at that salary tier have enormous variance. You use them to create salary relief and swing for a ceiling. You do not build your floor around them.
4. The walk-forward validation that underpins this model holds up on nights like this precisely because we are not pretending the Leite miss did not happen. A 6.6 MAE on a volatile, injury-disrupted two-game slate is a number we will take — and the miss is already in the next training window.