A'ja Went to Work and Brochant Blew the Roof Off

Two games on the board Tuesday, and the slate didn't waste anybody's time. Las Vegas leaned on their cornerstone — as they do — and Phoenix got a two-headed night from veterans and a role player who had absolutely no interest in staying in her lane. Los Angeles and Washington gave us the second game, and both produced enough fantasy-relevant performances to make the slate feel bigger than two games usually do.

The injury news that broke before tip reshaped a couple of usage pictures, and if you were monitoring the model's live updates, you had a head start on exactly who would benefit. That's the whole game, really — it's rarely the starting five you projected on Monday that plays Tuesday. It's the adjusted version. The teams adjusted. The model adjusted. Let's see how it graded out.

The honest one-number receipt first: a mean absolute error of 5.4 DK points across 38 graded players. The model is walk-forward validated — it beats naive baselines out of sample — and 5.4 is a number you can actually build with. Not every night looks like this. Some nights it's worse, and we'll tell you when it is. Tonight it earned its keep.

The Slate's Top Scorers

• A'ja Wilson (LV, $12,500) — Projected 50.4, delivered 56.8. Line: 26 pts, 13 reb, 2 ast, 2 stl, 4 blk in 35 minutes. An injury had bumped her projected minutes heading into the night, and she used every one of them. Four blocks. Thirteen boards. This is what a generational player does when the door opens a little wider — she walks through it, then takes the door off the hinges. The model had her as the top projection on the slate and she cleared it by 6.4. That's the job done.
• Noemie Brochant (PHX, $5,100) — Projected 12.9, delivered 42.3. Line: 24 pts, 3 reb, 3 ast, 5 stl in 30 minutes. This is the number we have to own plainly. A 29.4-point miss is a 29.4-point miss, and we're not going to dress it up. Five steals and 24 points at $5,100 is the kind of line that wins tournaments by itself. The model did not see it. Worth understanding why: Brochant's projection was built on her established role and usage history, and she simply went well beyond it. No injury bump flagged here — this was organic outperformance, the kind of night that happens in a small sample and is nearly impossible to project with confidence at that salary. It stings, especially from a chalk-avoidance standpoint in GPP construction. File it, move on.
• Jackie Young (LV, $10,300) — Projected 37.7, delivered 41.5. Line: 25 pts, 2 reb, 8 ast in 32 minutes. An injury had bumped her minutes, and eight assists alongside 25 points is exactly the kind of usage you were hoping to see. The model was 3.8 points light — a clean hit by any reasonable standard.
• Nneka Ogwumike (LA, $10,700) — Projected 36.3, delivered 40.3. Line: 18 pts, 9 reb, 8 ast in 33 minutes. An injury had bumped her minutes, and she turned it into a near triple-double. The model called her direction correctly and was 4.0 points light. Another min-bumped player who delivered on the expanded role.
• Alyssa Thomas (PHX, $11,700) — Projected 40.4, delivered 39.8. Line: 16 pts, 7 reb, 9 ast, 1 stl, 1 blk in 33 minutes. Six-tenths of a point off the projection. That is, for lack of a better term, ridiculous accuracy. Alyssa Thomas is a model-friendly player — high-floor, multi-category contributor, consistent minutes — and the 40.4 projection was about as clean a number as you can put out. The slate's quietest great call.
• Shakira Austin (WSH, $11,100) — Projected 36.0, delivered 39.0. Line: 20 pts, 8 reb, 3 ast, 2 stl, 1 blk in 30 minutes. Three points light, and a genuinely useful outcome for anyone who rostered her. The multi-category production held up exactly as projected.

Best Calls

• Rae Burrell (LA, $8,300) — Projected 23.2, delivered 37.3. Line: 19 pts, 3 reb, 6 ast, 3 stl in 35 minutes. An injury had bumped her minutes, and at $8,300 she delivered 37.3 DK points — that's genuine value-per-dollar production. The model flagged the min-bump and adjusted her projection upward relative to her baseline. The 14.1-point beat over projection is a beat, but the directional call — play the min-bumped Burrell — was the one that mattered for lineup construction.
• DeWanna Bonner (PHX, $8,900) — Projected 26.1, delivered 35.3. Line: 11 pts, 7 reb, 2 ast, 3 stl, 3 blk in 27 minutes. Eleven points doesn't jump off the page, but three steals and three blocks absolutely do. The model projected her multi-category upside and she delivered it in a different box-score shape than expected — lower scoring, higher defensive production. A 9.2-point beat on a player who doesn't need the ball to score is a good reminder that the DK scoring system rewards the full line.
• A'ja Wilson again — already covered above, but she belongs here too. Best projection on the slate, best outcome on the slate. The min-bump edge combined with the player being A'ja Wilson is a combination that tends to work out.

Worst Calls

Beyond Brochant — which we addressed above and which stands as the clearest miss of the night — the brief carries no additional flagged worst calls from the model's own ledger. That's either a good night or incomplete grading, and with a two-game slate still marked in progress, we'll take it at face value while reserving the right to update. The Brochant miss is the one to sit with. It's real, it was significant, and we'd rather say so clearly than footnote it.

Three things worth carrying forward.

1. The min-bump edge is real and it works. Four of the slate's top six scorers were flagged as min-bumped heading into the night — Wilson, Young, Ogwumike, and Burrell — and all four delivered at or above projection. That's not luck, that's the mechanism working. When a teammate's injury redistributes usage, the model catches it in near-real time and adjusts. The players who benefit tend to outperform their pre-news baselines. If you're building lineups and ignoring the injury adjustments, you're leaving the most actionable edge on the table.
2. Role-player ceilings are legitimately unforecastable at the low end of the salary range. Brochant at $5,100 was a 29-point miss. That's not a model failure in the traditional sense — it's the inherent variance in projecting players with limited established usage. The floor is low, the ceiling is occasionally spectacular, and the probability of landing on the spectacular night is, by definition, low. In GPP construction, sprinkling in a low-salary player with upside is a valid strategy. Just understand that the projection on those players is wider by design — the model's confidence interval is honest about that even when the column number isn't.
3. Thomas-type players are the model's sweet spot. High-usage, multi-category, consistent-minutes players project cleanly because there's less variance in their role. A six-tenths-of-a-point miss on Alyssa Thomas is the model doing exactly what it's supposed to do. In a smaller slate where every rosterable player matters, leaning into the high-floor, well-projected names and letting the tournament construction handle the upside swings is a reasonable framework to carry into the next slate.

MAE of 5.4, a min-bump edge that held up across four players, one legitimate miss to own, and a two-game slate that gave us more to work with than most. We'll be back with projections when the next slate locks.