Fin-setting causality

Skiers tweak fin geometry constantly — DFT, depth, length, wing angle. The question is rarely should I tweak it? and almost always did the last tweak actually help, or was that lake just glassier?

This recipe groups your sets by fin DFT and surfaces your performance at each setting.

Prompt

Group my sets by `skiSettings.finDftIn` rounded to 3 decimals.

For each distinct DFT value, compute my best pass at:
- `speedMph === 34` (substitute your typical top speed)
- `lineOffUs >= 32`

Best pass = longest `lineOffUs`, tiebreaker highest `buoyCount`.
Exclude draft passes and `autoGenerated: true` passes.

Output a markdown table with columns:
- Fin DFT (in inches)
- Number of sets at that DFT
- Date range (earliest → latest `setDate` at that DFT)
- Best pass at -32 or shorter
- Average buoy count at -32 / 34 mph

Flag with a ⚠️ any DFT change that coincided with a clear jump (or drop) in best pass.

Example output

Fin DFT (in)	Sets	Date range	Best pass at -32 or shorter	Avg buoys at -32 / 34 mph
0.755	6	2024-06-12 → 2024-07-30	2.0 @ -32	1.4
0.760	18	2024-08-02 → 2025-04-19	4.5 @ -32	3.1 ⚠️
0.765	14	2025-04-26 → 2025-08-14	5.5 @ -32	4.6 ⚠️
0.770	4	2025-08-21 → 2025-09-09	3.0 @ -32	2.5

⚠️ Bumping DFT 0.760 → 0.765 lines up with a clear jump in average buoy count at -32 / 34 mph.

Follow-ups

“Plot finDftIn on the X-axis and average buoyCount at -32 off / 34 mph on the Y-axis. Use Mermaid xychart-beta.”

“Now do the same analysis but for finDepthIn instead of finDftIn. Then wingDeg. Then finLengthIn.”

“For the DFT setting where I performed best, list every set that contributed to that average so I can sanity-check it.”

Variations

Pair fin changes with results. Order sets chronologically; for each setDate where finDftIn changes from the previous set, show the best pass before and after. This is closer to a real before/after analysis.

Multi-variable. Ask the AI to find the combination of finDftIn × finDepthIn that produces your best pass. Sample size will be small — caveat the answer.

Causality caveat. Correlations like these are suggestive, not causal. A DFT change usually rides with a coaching session, a different lake, or a fitness trend. Use the recipe to find candidates worth retrying, not to conclude the fin alone made the difference.