Team

Driver	Team	Price	Boost	Pts
Kimi Antonelli	Mercedes	$26.6M	⚡ 2×	56
Charles Leclerc	Ferrari	$24.0M	—	20
Arvid Lindblad	Racing Bulls	$8.2M	—	17
Nico Hülkenberg	Audi	$4.8M	—	15
Lando Norris	McLaren	$27.0M	—	−11

Constructor	Price	Delta	Pts
Ferrari	$27.5M	↑$0.3M	45
Racing Bulls	$14.7M	↑$0.6M	23

$132.8M squad value · $0.1M in bank · Chip: None · Penalty: None — 1 of 3 free transfers used, 1 carried to R16 · 165 pts

Race Summary

Russell won Baku from pole with the fastest lap, 0.196 ahead of Verstappen, who came through from P8. Hadjar finished P3 on his return, Leclerc P4 from the front row, and Antonelli P5 from P16 after a Q1 exit. For us, the boost held up despite that qualifying. Antonelli’s 28 base, doubled to 56, was still the best driver score in the squad, and the cheap tier recovered from the back (Lindblad P15 → P7, Hülkenberg P18 → P11). The round’s one loss was the transfer: Norris, bought in for Hamilton, retired on lap 35 for −11, while Hamilton finished P6 for +15. 165 pts is our best round since R12, 51 above R14.

Driver Notes

All five driver scores fit the platform scale (see Key Intel): qualifying P1 10 → P10 1, race P1 25 → P10 1, ±1 per grid place gained or lost, overtakes +1, DNF −20. Overtakes are the only part the classification can’t show, so each overtake count below is the gap between the reported score and the rest.

Driver	Grid	Finish	Pts	Notes
Kimi Antonelli ⚡ 2×	P16	P5	56	Base 28 × 2× = 56, shown on the platform as “28 X 2x”. That is the fourth confirmation that the boost doubles the full weekend. He went out in Q1 in P16, and the results file does not record why (confirm in R15 research). He then gained 11 places to P5. Platform scale: quali P16 0 + P5 10 + 11 places = 21, plus 7 overtakes = 28. The post-qualifying check was applied and correctly kept the boost on him. The picks-file rule moved the boost to Norris only if Norris started on the front row and Antonelli outside the top four. Antonelli met his half, but Norris started P5. 28 was the highest driver base in the squad by 8, so no other boost choice would have scored more.
Charles Leclerc	P2	P4	20	No engine penalty at Baku. The picks file held Leclerc “through a likely back-of-grid penalty”, but he qualified P2, 0.837 off pole, and started there. He lost two places in the race to finish P4. Platform scale: quali P2 9 + P4 12 − 2 places = 19, plus 1 overtake = 20. The penalty is presumably still to come, and that matters at R16 (see Key Intel).
Arvid Lindblad	P15	P7	17	His Q3 run ended, but he went from P15 to P7. He went out in Q2 in P15, his first qualifying outside the top ten since R07. Before that he had made Q3 seven rounds running, R08–R14 (R13 and R14 called it “fourth” and “fifth”; the qualifying tables show seven). He gained 8 places to P7 and beat Lawson (P12) by five. Platform scale: 0 + P7 6 + 8 places = 14, plus 3 overtakes = 17. The points came from the race, not from qualifying.
Nico Hülkenberg	P18	P11	15	15 points from a finish outside the points. Q1 exit in P18, seven places gained to P11, ahead of Bortoleto (P15). Platform scale: 0 + 0 + 7 places = 7, plus 8 overtakes = 15, the most in the squad. On this scale a back-of-grid start at an overtaking circuit is worth more than P9 or P10 from a Q3 slot. Compare his Madrid weekend: P11 → P10, 2 points.
Lando Norris	P5	DNF	−11	Out on lap 35, from P5 on the grid. Gasly retired on the same lap and Colapinto one lap later. The results file does not record the causes, so whether the three were linked is unknown (confirm in R15 research). Platform scale: quali P5 6 + DNF −20 = −14, plus 3 overtakes = −11. A DNF does not also cost places lost. McLaren’s revised sidepods, the main reason he was bought, did not show in either car’s result. Piastri dropped from P3 on the grid to P13.

Constructor Notes

Constructor	Pts	Notes
Ferrari	45	Back above 40 after 18 in Madrid. Leclerc P4 and Hamilton P6, with both cars in Q3 (P2, P6). Drivers 20 + 15 = 35, plus both-in-Q3 +10 = 45 exactly, a zero residual, which confirms the platform’s constructor qualifying bonus. Hamilton contributed +15 here as the driver we had just sold.
Racing Bulls	23	Lindblad P7 and Lawson P12, with both cars out in Q2 (P12, P15). Drivers 17 + 3 = 20, plus both-in-Q2 +3 = 23 exactly. The R14 discrepancy did not recur: the platform’s team view and its season-points movement agree this round. The returning Lindblad–Lawson pairing scored a third of R11’s 47, and almost all of it was Lindblad’s.

Key Intel

• The platform scale is now confirmed on all seven slots. Every score above reconciles to it, with non-negative overtake counts. Both constructors match the standard qualifying bonus (both in Q3 +10, both in Q2 +3) with no residual. The field-wide check in content/prices/R15-Prices.md gets the same result for all 22 drivers and 10 of 11 constructors. The exception is Red Bull, which is 8 below the formula, possibly because Driver of the Day doesn’t count for constructors. The Game Rules table in CLAUDE.md is still the old one and should now be rewritten, not just flagged.
• The boost doubles the full weekend, confirmed four times: R12 “57 × 2x”, R13 “26 X 2x”, R14 “35 X 2x” and R15 “28 X 2x”. The CLAUDE.md line saying it applies to “the Grand Prix race score only” is wrong.
• The bank reconciled exactly. The platform reads a $0.1M cost cap, as the picks file projected (Hamilton released $24.9M + $2.2M bank − $27.0M for Norris). Budget R16 from $0.1M, plus whatever sells release at R15 prices.
• ⚠ Leclerc’s engine penalty has not been taken. The R14 research expected it at Baku, and Vasseur called Baku “quite suitable”. It did not happen because Ferrari saved his Monza unit (see content/research/R15-Azerbaijan.md). No penalty is scheduled now, but both Ferraris have no spare power-unit parts, so any new part means a penalty with little warning.
• Russell won at Baku on the end-of-life power unit the picks file flagged as a risk at a power circuit. Mercedes won the race and took pole and the fastest lap, but their second car went out in Q1. That makes Antonelli’s Q1 exit the more important unknown. If it was mechanical, it is a reliability flag on the best asset in the game.
• Red Bull was the second-strongest car: Verstappen P2 from P8 and Hadjar P3 on his return from P4 on the grid. Red Bull scored +71, Hadjar’s price rose to $15.1M and Verstappen’s rose for the first time since R05.
• McLaren’s weekend fell apart. Both cars qualified in the top five (P3, P5), then Norris retired and Piastri dropped to P13 from P3. The constructor scored +1 and still rose +$0.3M.
• Championship after R15 (R14’s confirmed standings plus the Baku classification. Not yet checked against the official table): Antonelli 302 · Russell 236 · Hamilton 199 · Norris 186 · Leclerc 179 · Verstappen 163 · Piastri 120 · Hadjar 86 · Lawson 59. Antonelli’s lead over Russell shrinks from 81 to 66 with eight rounds left. Hamilton moves 13 clear of Norris for P3.

Result Notes

• The boost and the holds worked. The transfer didn’t. Keeping Hamilton instead of buying Norris would have scored 191, 26 more. Both prices then fell −$0.3M, so the swap gained nothing on value either. This round does not show the Norris thesis was wrong. One DNF of unrecorded cause is not a pace verdict, and the R14 evidence for him (pole, fastest on track) still stands. But the cost of chasing form instead of holding season value was 26 points this time, and that belongs in the record next to R14’s win for the same approach.
• The Leclerc hold paid off, though not for the reason given. The picks file defended holding him through a back-of-grid start, citing the R13 Antonelli recovery. He started on the front row instead and scored 20. The bigger test of the “don’t sell for a penalty” rule is still to come.
• No Negative was held, correctly by its own rule, and would have been worth +11. The chip was tied to a confirmed Leclerc penalty, which didn’t happen. Zeroing Norris’s −11 would have made this round 176. Singapore remains the planned window, but it is now R17, not R16. R16 is the Bahrain GP at Sepang (Oct 2–4), a race relocated in July that the project calendar missed. Singapore is still a street circuit and sprint weekend with the highest safety-car rate on the calendar.
• The cheap tier recovered: 32 points from $13.0M (Lindblad 17, Hülkenberg 15), after 7 in Madrid. Both scores came almost entirely from places gained and overtakes from the back. The platform scale pays little for P9–P10 from a top-ten grid slot and a lot for climbing from P15–P18. That makes cheap drivers who qualify badly but race well worth more than their qualifying suggests.
• Ferrari exposure is no longer out of proportion. Leclerc plus the constructor were 39% of squad value and returned 65 of 165 (39%), compared with 59% of value and 21% of points in Madrid.
• Season total: 3,441 pts (R01 190 · R02 437 · R03 202 · R04 260 · R05 261 · R06 187 · R07 158 · R08 223 · R09 298 · R10 217 · R11 231 · R12 354 · R13 144 · R14 114 · R15 165).
• Chip status: unchanged. Wildcard and No Negative remain. Wildcard is held for the Americas triple-header, and No Negative is lined up for R17 Singapore.
• Transfers: 1 of 3 used, 1 carried → 3 free at R16, $0.1M in bank.
• Price changes heading into R16 Bahrain GP (Sepang) (full detail in content/prices/R15-Prices.md): Antonelli ↑$0.3M to $26.9M · Norris ↓$0.3M to $26.7M · Leclerc ↓$0.3M to $23.7M, a third straight fall · Lindblad ↑$0.6M to $8.8M · Hülkenberg ↑$0.6M to $5.4M, a fourth straight max rise and +50% on the $3.6M we paid · Ferrari ↑$0.1M to $27.6M, its first rise this season smaller than +$0.3M · Racing Bulls ↑$0.6M to $15.3M, a max rise at every update this season. Our squad is now worth $134.4M, up $1.6M.