Williams

Pre-season baseline — Mercedes customer, underwhelming despite mileage

Despite completing the joint-most laps in the first Bahrain test (422, tied with McLaren), 'all available evidence suggests the Williams is nothing to write home about' (The Race). Long runs were 'uninspiring' and performance runs 'caught for wrong reasons.' 898 total laps provides decent reliability data (rated 4.8). Mercedes PU provides some energyRecovery (4.5) and straightLineSpeed (4.8) uplift relative to the weak chassis. Sainz's +0.5 offset (effective 4.6) and Albon's +0.3 (effective 4.4) place them in the lower midfield, consistent with expectations. SANDBAGGING/PU CONTEXT: As a Mercedes customer, Williams ran the same de-tuned PU mappings as McLaren and Alpine during testing. Williams was 'surprised by what Mercedes can extract from the 2026 engine' (The Race), highlighting the works-vs-customer knowledge gap. energyRecovery (4.5) and straightLineSpeed (4.8) are artificially depressed relative to the actual Mercedes PU hardware capability. However, Williams appears to have the weakest chassis of the four Mercedes-powered teams, and the PU uplift alone cannot compensate for the fundamental aero/mechanical deficits.

Round 1 baseline — overweight, low downforce, energy deployment confusion

Albon: 'overweight and not producing enough downforce.' Albon P15 in quali (+2.423), P12 in race. Sainz didn't qualify (engine/gearbox issues), started P21, finished P15. Race pace 85.662 (+2.164). Graining 'especially on hard tyre.' Team 'caught off guard by what Mercedes can extract.' Albon hydraulics failure FP1, Sainz multiple technical issues. Sources: The Race, RaceFans, FIA speed trap PDFs.

Round 2 — Albon DNS, Sainz recovery drive, growing reliability concerns

Albon DNS from hydraulic failure adds to Williams' growing reliability headaches (Sainz missed sessions in earlier rounds). Sainz recovered from P17 to P9 — impressive driver extraction from a poor car. Race pace 2.531s off Mercedes confirms lower-midfield standing. Team acknowledged weight issues and aero deficit. Sources: Formula1.com, The Race, Williams.

Round 3 — speed trap 3rd fastest (Mercedes PU), race pace ahead of Alpine/Racing Bulls

Speed trap 332 km/h (3rd fastest on grid, Mercedes PU advantage). Race pace +1.236s — actually faster than Alpine and Racing Bulls at Suzuka. Sainz P15, Albon P20 (+2 laps, unclear cause). Front suspension cladding upgrade. Albon's memorable quote: 'there isn't really any high-speed anymore' about 2026 energy clipping. Sources: Formula1.com, ReadMotorsport upgrades, FastF1 data.

Round 6 (Miami) — both Q2 first time, double points, fastest in speed traps with 7 upgrades deployed

Breakthrough weekend. Both cars Q2 for first time in 2026; SAI P9 + ALB P10 = double points. Speed trap 351 km/h (fastest on grid, ahead of Russell's 350 km/h). 7-piece upgrade package deployed: refined floor contours, reshaped sidepod inlet lips, revised engine cover + rear cooling exits, repositioned mirror bodies, new bracket behind tailpipe, revised fairings around rear impact structure. Race pace +1.724s — better than Audi, Haas, Racing Bulls, Aston Martin, Cadillac. Williams chose low-DF setup for Miami's straights — part of the speed-trap gain is circuit-specific setup, but the new floor + aero efficiency parts are real chassis improvements. straightLineSpeed +0.2 is back-to-back same direction (R3 was +0.2 from 4.0→4.2) but each step has had two-source confirmation. Sources: F1.com, The Race, RaceFans, F1i speed trap analysis.

Round 7 (Canada) — modest flow-conditioning package, SAI P9 recovery, ALB data contaminated

Williams deployed FBD scoop geometry with reprofiled exit, repositioned exhaust tailpipe, revised suspension cladding, flow-conditioning mods. SAI recovered P15 grid → P9 race (good Mercedes-PU race pace, VSC pit timing onto mediums). ALB's weekend was contaminated: hit a groundhog/marmot in FP1 T6-7 chicane (chassis damage, gearbox + PU replaced), then DNF lap 13 of the race after Piastri's lock-up into him at the hairpin. Discount all ALB pace data. straightLineSpeed +0.1 — modest continuation of the prior trend (R6 was also +0.2 on speed trap evidence). No other car-DNA moves warranted with so much ALB-side contamination.

Round 9 (Barcelona) - double Q1 exit, SAI P12; deg narrative conflicts with raw data; grid re-center only

No performance-attribute changes: Sainz's 'degging two times more' quote and The Race's extreme-deg narrative conflict with the skill's primary source (FastF1 stint deg: Williams 0.103 avg = 3rd-lowest, flattered by VSC-era stints) - when sources conflict, no move; monitoring. Note the model has under-predicted Williams race pace four rounds running (R3/R6/R7/R9 backtest) - tracked as a derived-baseline calibration issue (round-6 preseason fallback), not a car-DNA authoring issue. ALB unclassified (90% rule) with handling/setup trouble - no incident logged (no mechanical cause reported). Reliability +0.2 = grid re-center only. Grid reliability re-center +0.2 applied to all 11 teams (incident penalties pushed the grid reliability average to 4.33, breaching the +/-0.5 band; re-centered to 4.53, relative gaps preserved). Pipeline re-derives qualiPace/racePace from committed raw round-09; no manual pace override (dry, representative weekend). Sources: FastF1 round-09; The Race; Wikipedia.

Round 10 (Austria) - poor weekend: worst race deg in the field, draggy, SAI electrical DNF

Q17/Q18; ALB P17 two laps down; SAI suspected-electrical DNF L23 (triggered the first VSC). tyreDegradation -0.2: worst race deg in the field (0.0942 team avg) - resolves R9's monitoring item where the extreme-deg narrative conflicted with stint data; data and narrative now agree (FP2 long-run figure excluded as unreliable fuel-uncorrected class). straightLineSpeed -0.1: 'underdeveloped and draggy' (The Race) despite the Mercedes PU at a power circuit. No upgrades registered (one of two teams). Reliability -0.3 pre-re-center: SAI electrical L23 = first Williams race mechanical since R2, first-failure cap, monitoring (Mercedes-customer attribution unconfirmed - manufacturer thread not extended). Net -0.1 after re-center. Overall -0.1 only, given the violent R9-R10 swing (Barcelona weekend P6 vs Austria P17/DNF) and the model's tracked under-prediction of Williams race pace. Grid reliability re-center +0.2 applied to all 11 teams (R10 incident penalties pushed the grid reliability average to 4.36, breaching the +/-0.5 band; re-centered to 4.56, relative gaps preserved). Pipeline re-derives qualiPace/racePace from committed raw round-10; no manual pace override (dry, representative weekend). Sources: FastF1 round-10; The Race; PlanetF1; Wikipedia.

Round 11 (Silverstone) - scrappy home weekend of team errors; no mechanical fault

Poor home weekend: double Q2 exit (P15/P16); SAI P12 on the road -> classified P17 (+1 lap) after the unprecedented 'penalty lap' (team wrongly instructed an unlap under the late SC - Silverstone SC-line geometry quirk); ALB retired lap 43 with damage from his own Brooklands lap-1 collision (10s penalty) after a sprint pit-lane start for an unapproved parc-ferme suspension change. New front wing (profiles + endplate surfaces) deployed at the home race - no measurable result. No mechanical fault this weekend: ALB's DNF is contact damage, and the raw 'dnf' flags on SAI (and ALO/STR) are lapped-finisher artifacts - no reliability decrement; +0.1 grid re-center only (monitoring flag from R10 SAI electrical stays). tyreDeg monitoring: team race deg +0.084 again worst-tier, but ALB's damaged 6-stint race pollutes the sample - no move on top of R10's -0.2. Vowles: 2026-rules pain lasts 'until next year'. Grid reliability re-center +0.1 applied to all 11 teams (R11 incident penalties pushed the grid reliability average to 4.45, breaching the +/-0.5 band; re-centered to 4.55, relative gaps preserved). Raw round-11 caveats for publish-time verification: race status mislabels lapped finishers SAI/ALO/STR as dnf at lap 51 (only VER/ALB/HUL genuinely retired; sprint lap-16 'dnf' flags for STR/PER are the same artifact); gapToWinner field broken; FP1 long runs not fuel-corrected. Sources: FastF1 round-11; F1.com; Motorsport.com.