2026-06-03 — First baselines for 60°/PW/3W/4i: the shallow-AoA pattern holds across the bag

Sessions: four shot-analysis sessions tonight — 60° 21:06 (17 shots, 15 clean), PW 21:16 (23, 15 clean), 3W 21:37 (28, 23 clean), 4i 21:51 (26, 26 clean). First stored data for all four clubs. Worked on: attack angle (continuing the AoA/release thread from 2026-06-02 and the 6i re-read) — but the primary goal was getting these clubs into the dataset at all, since none had data to calibrate on. Treat these numbers as baselines, not a verdict on a swing change.

What the data showed

Clean strikes only, published bands where they exist:

• PW: the deficit is systemic, not chunks. Clean filtering moved smash 1.096 → 1.101. CHS is above band (84.4 all-shots vs 78–82) yet carry is ~15–20 yd short — swinging harder to buy back efficiency. corr(smash, carry) +0.78. The lean is there (8.1° deloft, low point +1.2″ ball-first); the AoA is ~4° too shallow and launch/descent run HIGH.
• 4i: best contact session in the dataset — 26/26 clean, smash CoV 4.7%, 21 direct center-face reads averaging −1/+1 mm (dead center) — but the delivery is upside-down. AoA +0.8° ascending on a 24° iron, low point 1.0″ before the ball: a sweep, not compression. Launch/apex/descent all HIGH, carry 15–25 yd short of band. corr(smash, carry) +0.91.
  • Honesty flag: smash 1.317 with spin loft below band (20.1 vs 23–26) and measured-center contact doesn't fit the guide's off-center explanation. Either the rising strike costs efficiency in a way the impact reads don't show, or the 1.42–1.44 band is optimistic for this club/proxy. Watch it once a descending baseline exists.
• 3W: adding 4.0° of loft through impact (DL 17.0 vs 13° static) — the flip signature, with low point at/behind the ball (−0.3″). Clean carry 201 but sd 23 yd; the 5 mishits include outright tops (21-yd carry, 0.2-ft apex in the raw set).
• 60°: delivery is structurally fine (9.7° deloft, low point +1.6″) but distance control is strike-limited: carry CoV 13.9%, corr(smash, carry) +0.97 and corr(smash, launch) −0.69 — the short ones are the scooped, higher-launching strikes. No bag.json entry yet (analyze.py is normalizing against the sim proxy's 60° spec) — add the real club to reference/bag.json.
• Path, all four clubs: +4 to +6° in-to-out with face closed to path (F2P −1.3 to −3.7) — the same push-draw geometry the 6i/8i showed at +7–9°. Slightly tighter here, still the secondary item.
• Spin measured directly on only 4/94 shots (indoors); spin means are interpolated.

Mechanism

Same story as the 6i re-read, now visible across the rest of the bag: the release/lean move survives — the bottom of the arc doesn't move forward with it. Deloft scales correctly down the bag (60° −9.7°, PW −8.1°, 4i −3.5°) and low point is ball-first on the wedges, but AoA never gets near band anywhere, and by the long clubs it inverts (4i ascending, 3W flipping with low point behind the ball). The longer and faster the club, the worse the arc-bottom position — consistent with the 6i finding that the in-to-out path (here +4–6°) is also a swing up. The AoA intent tonight did not arrive at the long end.

Highest-leverage next focus

Low point forward on the long clubs — 4i and 3W. Same prescription family as the 6i: not "hit down harder," but bottom-of-arc in front of the ball with the lean that's already there. If it works: 4i AoA goes negative (target −3.5–−2.5) with low point positive, launch drops into 13–15°, spin loft rises into 23–26, and the missing ~15–25 yd of carry returns at the same CHS. 3W: DL down toward static (≤ ~15°) instead of +4° added. PW is the same fix at smaller magnitude (AoA −1.4 → −5ish).

Housekeeping that was the actual point of tonight (done same session): calibrated personal bands for all four clubs (calibrate.py --all --clubs "60°,PW,3w,4i", n=15–31 clean each — essentially one session, so provisional), and aliased 60° → LW in reference/bag.json (the Cleveland RTZ was already there). Per the 6i note's caveat, these bands encode tonight's faults — the 4i band literally blesses an ascending AoA (−0.2–1.2) and the PW smash band tops out at 1.133 vs 1.28 published. Use them for solid/mishit classification only; judge AoA/low-point/smash against published targets until the low-point fix lands, then recalibrate.

Trend

None — first data for all four clubs. Baselines to beat next session: PW smash 1.101 / AoA −1.4; 4i AoA +0.8 / low point −1.0″ / carry 180; 3W DL-vs-static +4.0°; 60° carry CoV 13.9%.