Thursday gave us nine games and a clean 2-0 record on totals — our only graded action for the day. No moneyline plays made the card, so the entire betting record for Thursday sits there: no ML decisions, two totals, two wins.

The run environments did most of the work. Oakland and Tampa Bay combined for 11 runs on an over/under of 7.5, and Boston and Texas stayed under 8.5 with a quiet 4-3 final. Both games did exactly what the model suggested they might, in opposite directions. That's a good night.

No money was staked Thursday. card_graded is null — nothing cleared the bar for an official play, and that's the honest record. What we published were leans: tracked edges, nothing wagered.

The Leans

• OAK @ TB — Over 7.5 (lean). Final: Oakland 1, Tampa Bay 10. Total runs: 11. WIN as a tracked lean.
• BOS @ TEX — Under 8.5 (lean). Final: Boston 4, Texas 3. Total runs: 7. WIN as a tracked lean.

Both leans would have paid. We call that out because the board deserves an honest accounting — but these were not official plays, and we won't credit them as such. The leans go 2-0 on totals Thursday, 0-0 on moneylines.

We had 277 players score Thursday, with 136 matched to projections. The range was wide.

Top Scorers

• Kyle Tucker (Los Angeles Dodgers): 33.0 DK pts, 4-for-5. Our projection: 8.9. That's a significant beat — Tucker was nearly four times his projected value.
• Walbert Urena (Los Angeles Angels): 31.35 DK pts, 7.0 IP, 9 K, 0 ER. Our projection: 15.51. A strong outing that roughly doubled his projected floor — clean game from a pitcher we had reasonably priced at $8,000.
• Xander Bogaerts (San Diego Padres): 29.0 DK pts, 3-for-5. Our projection: 7.14. Another big beat.
• Munetaka Murakami (Chicago White Sox): 26.0 DK pts, 2-for-4. Our projection: 6.4.
• Aaron Nola (Philadelphia Phillies): 24.15 DK pts, 7.0 IP, 4 K, 0 ER. Our projection: 16.21. He finished within 8 points of the projection — a reasonable call on a $7,700 arm.

Where We Hit

Tucker, Urena, and Bogaerts were the three clearest projection wins, each producing well above what the model expected. Tucker in particular — 33 DK points against a 8.9 projection — is the kind of outcome you don't predict so much as hope you're in the lineup for. Urena going 7 deep with 9 strikeouts at $8,000 was a legitimate value call that held up.

Where We Missed

Three starting pitchers hurt the slate badly.

• Kyle Harrison (Milwaukee Brewers): projected 15.46, finished at -5.05. He lasted 1.2 innings and allowed 4 earned runs. Those are the kinds of lines that crater GPP rosters built around him.
• Brady Singer (Cincinnati Reds): projected 11.11, finished at -15.5 — the worst single result on the board Thursday. 3.1 innings, 0 strikeouts, 7 earned runs. There's no model that accounts for that kind of early exit gracefully.
• Miguel Ullola (Houston Astros): projected 10.95, finished at -2.3. Two-thirds of an inning, 1 earned run. Another short outing that wiped out the projection entirely.

Starting pitcher variance is always the sharpest edge in DFS, in both directions. Thursday was a good illustration of that: Urena and Nola delivered, Harrison, Singer, and Ullola did not.

The totals model had a clean Thursday. Both leans identified meaningful disagreements with the market, and both resolved in the right direction — one blowup offense in Tampa, one locked-down pitchers' duel in Texas. The fact that neither cleared the bar for an official play is worth noting: the edges existed, but the model's threshold exists for a reason, and we're not going to adjust that threshold retroactively because two leans won.

On the DFS side, the recurring lesson is the same one Thursday reinforced: when you build around starters who don't make it to the third inning, no amount of hitting upside covers the hole. Harrison, Singer, and Ullola combined for 5.2 innings and -22.85 DK points against a collective projection north of 37. That's a brutal spread. The better side of the pitcher ledger — Urena and Nola both going seven innings — shows the upside when the projection is right. The problem is you don't always know which side you're on until pitch 15.

One forward-looking note: Tucker's 4-for-5 game at 33 DK points is a useful reminder that hitter projections carry wide variance too. An 8.9 projection isn't wrong just because the player went 33 — that's within the range of outcomes for a hitter of his caliber on a good night. We track the miss honestly, but we don't overfit the model to individual outcomes.