Trackman stats analysis

Trackman Stats Analysis

Pull the user's Trackman data via the MCP and produce an honest, specific diagnosis of where they're losing strokes. This skill does NOT prescribe drills — it hands a prioritized weakness list to the golf-coaching skill.

Inputs

Call the MCP tools (run auth first if needed):

• trackman(action="profile") → current handicap.
• trackman(action="sessions") → recent practice + rounds (default: last 60–90 days).
• trackman(action="rounds") → scorecards for scoring analysis.
• trackman(action="clubs") → per-club gapping and dispersion.
• trackman(action="session", activity_id=...) → shot-level detail where you need it.

Then call synthesize() for the cross-source, context-aware view — it aligns these Trackman findings against any GameBook on-course findings by skill area (Trackman is clean-room/flat-lie; GameBook is on-course, real conditions), so a gap that shows up in both isn't double-counted and one that only shows up on-course points at lies/pressure/course-management rather than pure mechanics. Diagnose from the aligned view where it's available, falling back to raw trackman output when synthesize has nothing (e.g. no GameBook rounds saved).

If a tool errors with auth expired, re-run auth. If the MCP isn't built yet (Phase 0/1 incomplete), say so and stop — don't invent numbers.

What to analyze

Work through these lenses and keep only what the data supports:

1. Club gapping. From trackman(action="clubs"), list avg carry per club. Flag overlapping clubs (gaps < ~8–10 m between adjacent clubs) and big holes in the set (gaps > ~18–20 m). Gaps cost approach shots.
2. Dispersion / consistency. Per club, look at carry spread and side/left-right scatter. Wide side dispersion on scoring clubs (wedges, short irons) is high-value to fix. Note the tightest and loosest clubs.
3. Launch quality. Spot inefficiencies: low smash factor (poor strike), spin too high/low for the club, launch + spin combos that kill carry. Driver especially: launch/spin vs an efficient window.
4. Scoring trends (from rounds): fairways hit %, greens in regulation, putts/round, and where doubles+ come from (driving? approach? short game?). This anchors practice to what actually costs scores.
5. Handicap movement. Is it trending up, flat, or down over the window? Tie it to the above.

Output: the diagnosis

Produce a short, ranked list — highest stroke-impact first — e.g.:

1. [Approach] 8-iron side dispersion ±18 m, carry varies 12 m → missing greens long/short.
2. [Gapping] 4-iron and 5-hybrid both carry ~178 m → a wasted slot; 188–196 m gap is open.
3. [Driver] launch 9.1° / spin 3400 rpm → launching too low with too much spin → weak, dropping carry, ~14 m left on the table.
4. [Short game] 3-putts mostly from >9 m → speed control.

For each item give: the club/area, the specific number that's off, and why it costs strokes — never vague ("be more consistent"). Cite the metric you read it from. If the data is too thin to judge something, say "not enough data" rather than padding the list.

Hand this ranked list to golf-coaching.

On-course rounds

Trackman covers practice and launch-monitor data. The user's real course rounds come from Golf GameBook via the gamebook tool (screenshot-ingested). When diagnosing scoring/course trends, include gamebook(action="compare") — or better, synthesize(), which already folds GameBook's scoring findings in alongside Trackman's — but trust only the scoring dimension unless a stat's coverage is not none.