Week 2 — Coaching Identity, Controllables, and Measurement

Outcome

Separate coaching value from athlete outcomes and classify performance information by control, influence, context, and uncertainty.

Required preparation

Bring the baseline snapshot from Week 1 and access to the same personal data source.

Part 1 — 5–7 minute instructional script

Consider this scenario: your athletes trained consistently, understood the race plan, supported one another, and arrived healthy. At the championship they performed below expectation. Did you fail as a coach? The honest answer cannot come from the result alone. Results matter, but they are created by many contributors: athlete ability, execution, competition, weather, illness, task difficulty, luck, officiating, and the decisions of other people. Coaching is one contributor, not the entire equation.

This is difficult because coaches are publicly associated with results. A win can make weak preparation look wise, while a loss can make sound preparation look foolish. If we anchor our identity entirely to outcomes, we become vulnerable to outcome bias: judging the quality of a decision only by what happened afterward. A risky choice can succeed. A thoughtful choice can fail. The coach’s job is to evaluate both the process and the result.

A useful model has four categories. Control includes your preparation, attention, communication, session design, and follow-up. Influence includes athlete understanding, team culture, and execution—you can shape these, but you cannot command them. Context includes weather, schedule, equipment, opponents, and life demands. Uncertainty includes what is missing, noisy, or not yet knowable. Mature coaching means owning the first category, working intelligently in the second, adapting to the third, and speaking honestly about the fourth.

Data can help, but measurement does not remove these categories. GPS pace may be affected by buildings, tree cover, sharp turns, or a misplaced activity type. Wrist heart rate can lag or produce cadence-like errors. Video angles can change an apparent joint position. A training summary can add numbers correctly while still missing the purpose of the workout. Measurement is not the same as truth; it is an observation created by a device, method, and context.

Precision can increase confidence without increasing accuracy. A watch may display 7:03 per mile, but the coaching conclusion should not automatically be more precise than the evidence. If the route is hilly, the day is hot, and heart-rate coverage is poor, the responsible conclusion may be “approximately the intended aerobic effort, but check perceived exertion and recovery.” Data-driven coaching is not about using more decimal places. It is about matching the strength of the statement to the strength of the evidence.

Return to your Week 1 question. Identify the metrics you selected. For each one, ask four questions: What produced this value? What could distort it? What decision could it change? What other context must I ask for? Weekly mileage may be reliable when activities are complete, yet it cannot tell you whether the athlete was healthy. Heart-rate zones may reveal intensity distribution, yet estimated zones may not fit the individual. A predicted performance may be useful for planning, yet course, tactics, and current readiness can make the actual result different.

Now apply the control model. Suppose your data suggests an easy run became moderate. You cannot change yesterday’s run. You can control how you interpret it, influence the next decision, account for terrain and weather, and name uncertainty about sensor quality or fatigue. This keeps data in its proper role: evidence that improves a decision, not a verdict about the athlete or coach.

A healthy coaching identity is built on intentionality, preparation, learning, athlete development, and relationships. That does not excuse poor results. It creates a better way to learn from them. Ask: Was the decision reasonable given what we knew? Was the athlete prepared to execute it? What did the evidence miss? What will we keep, change, and watch? This evaluation is more demanding than simply celebrating a win or apologizing for a loss.

Your challenge is to audit one metric you trust. Classify what you control, influence, and cannot control, then write one uncertainty statement. If you can do that without becoming defensive or falsely precise, you are beginning to use data as a coach rather than letting data define you.

Three takeaways

• Judge coaching decisions by preparation, reasoning, and learning as well as outcomes.

• Separate what you control, influence, must adapt to, and cannot yet know.

• Match the confidence of a conclusion to the quality and context of the measurement.

Part 2 — 30–45 minute application lab

1. Outcome-bias case — 5 minutes: Evaluate the championship scenario before and after receiving the result.

2. Metric audit — 12 minutes: Select pace, GPS distance, heart rate, form angle, prediction, or another personal metric. Record its source, likely errors, and missing context.

3. Control map — 10 minutes: Sort the decision factors into control, influence, context, and uncertainty.

4. Confidence rewrite — 8 minutes: Turn one overly certain claim into a proportionate conclusion.

5. Peer challenge — 5 minutes: A partner asks, “What else could explain this?” and “What decision changes?”

Data-driven participation assignment

Submit for discussion a metric-audit card containing the metric, data source, confidence level, one limitation, one contextual question, and the next controllable coaching action.

Literature + Generative-AI discussion preparation

Search coaching effectiveness, attribution, outcome bias, or decision quality in sport. Use the course AI-summary workflow and verify the source. Prepare to discuss: Athletes trained well but performed poorly at the championship. Which measures of coaching success should be examined before declaring success or failure?

Multi-sport adaptation

Audit the instrument your sport actually uses: stopwatch splits, jump height, shot chart, velocity reading, stroke count, force estimate, or video angle. Ask whether the number is repeatable, relevant to the decision, and comparable across sessions.

Protect athlete data and interpret cautiously

Do not convert a questionable measurement into a label about character, effort, or health. Keep personal examples de-identified when discussing them outside the course group.

Resources and next week

• rslts Development Metrics: https://rslts.run/docs/performance/data-analysis/development-metrics/

• Next: Week 3 — Culture, Self-Determination Theory, and Better Tools