Plain-English guide · No jargon
Data Dictionary
Every chart, metric and model in this dashboard explained in everyday language. If a term ever feels unclear, look it up here.
How to read confidence levels
| Confidence | What it means | How to act on it |
|---|---|---|
| High | Strong evidence | Trustworthy. Good for planning and acting on. |
| Medium | Useful guide | Take it as a directional hint, not a verdict. |
| Low | Early signal | Treat as a hypothesis worth verifying. |
Dataset reconciliation (audit)
Every Executive Overview KPI has been reconciled against the source export activities_1225.xlsx. The numbers below are the frozen ground-truth used by the in-app integrity checker.
| KPI | Ground truth | Definition |
|---|---|---|
| Activities | 1,925 | Row count from the Strava export. |
| Total distance | 52,699 km | Σ distance ÷ 1,000. |
| Total elapsed time | 3,051.75 hrs | Σ elapsed_time ÷ 3,600. |
| Avg distance / activity | 27.4 km | Total distance ÷ activities. |
| HR-tracked activities | 1,783 | Activities where avg HR > 0. |
| Avg HR | 138 bpm | Mean of avg HR across HR-tracked activities. |
| Max HR observed | 211 bpm | Single highest avg HR in dataset. |
| Active ISO weeks | 467 | DISTINCTCOUNT of ISO-8601 year+week (Mon-start, week 1 contains Jan 4). |
| Avg activities / month | 15.9 | Activities ÷ calendar months in active window (≠ active months). |
Date parsing
Strava CSV dates are parsed exactly as %b %d, %Y, %I:%M:%S %p in UTC, then truncated to a YYYY-MM-DD day key.
ISO week rule
ISO-8601: Monday is day 1; week 1 of any year is the week containing 4 January. Years can have 52 or 53 weeks — never capped.
CTL · ATL · TSB
CTL = 42-day exponentially-weighted load (λ = 1/42). ATL = 7-day EWMA. TSB = CTL − ATL. Load proxy = elapsed minutes × intensity factor.
ACWR (injury proxy)
Acute (7d km) ÷ Chronic (28d km ÷ 4). Flagged > 1.5 (overload) or < 0.8 (under-load). Capped to avoid divide-by-zero on returning weeks.
Core Metrics
Distance
- What it is
- Total kilometres covered in a workout.
- How it's worked out
- Comes straight from your GPS-recorded activity.
- How to read it
- More kilometres = more volume. Stack lots of weeks of distance to build endurance.
Moving Time vs Elapsed Time
- What it is
- How long you were actually moving vs total time the watch was running.
- How it's worked out
- Moving time pauses when you stop; elapsed time does not.
- How to read it
- Big gap = lots of stops (lights, café, photos). Use moving time to compare effort fairly.
Average & Max Heart Rate
- What it is
- Your typical and peak heart rate during a workout.
- How it's worked out
- Recorded by your chest strap or watch every few seconds, then averaged.
- How to read it
- Higher avg HR = harder effort. Max HR rarely sustained for long.
Heart Rate Zones (Z1–Z5)
- What it is
- Five effort levels based on a percentage of your maximum heart rate.
- How it's worked out
- Z1 < 60% · Z2 60–70% · Z3 70–80% · Z4 80–90% · Z5 ≥ 90% of estimated Max HR.
- How to read it
- Z1–Z2 builds endurance. Z4–Z5 builds speed. Spending most time in Z3 (the 'grey zone') is usually inefficient.
Predicted Max HR (Tanaka formula)
- What it is
- An age-based estimate of the highest your heart can beat.
- How it's worked out
- 208 − (0.7 × your age). For example, age 39 → 181 bpm.
- How to read it
- It's a rough guide for setting zones. Your actual max can be ±10–15 bpm different.
Personal Record (PR)
- What it is
- Your best-ever time or pace for a given distance or sport.
- How it's worked out
- Tracked across all your activities for each common distance (5 km, 10 km, etc.).
- How to read it
- PRs cluster when you're peaking. Long PR droughts can mean fatigue or under-training.
Volume & Activity
Monthly Training Volume
- What it is
- How many activities and kilometres you logged each month.
- How it's worked out
- We bucket every recorded activity by calendar month and sum the count and distance.
- How to read it
- Look for seasonal build-ups (spring/summer peaks) and recovery dips. Consistent month-on-month volume is a strong base.
Active Weeks per Year
- What it is
- Number of calendar weeks each year where you logged at least one workout.
- How it's worked out
- Out of 52 possible weeks, we count the ones with ≥1 activity.
- How to read it
- 40+ active weeks signals a strong, sustained habit. Big drops can flag injury, burnout or life events.
Longest Streak
- What it is
- The most consecutive days in a row that you trained.
- How it's worked out
- We scan your daily activity log and find the longest unbroken run of training days.
- How to read it
- Streaks are motivating but not always healthy — rest days matter for adaptation.
Sport Mix / Activity Mix
- What it is
- The share of your training time or distance that comes from each sport.
- How it's worked out
- We group activities by sport (Ride, Run, Swim, Other) and divide each by the total.
- How to read it
- A balanced mix reduces injury risk; a heavy single-sport mix shows specialisation.
Season Drop-off
- What it is
- How much your training volume falls after the main season ends.
- How it's worked out
- We compare distance covered Jan–Oct against Nov–Dec each year.
- How to read it
- A big drop is a normal off-season taper. A small drop suggests year-round consistency.
Time-of-Day Pattern
- What it is
- When during the day you typically train.
- How it's worked out
- Each activity's start time is bucketed into Early Morning, Morning, Afternoon, Evening or Night.
- How to read it
- Reveals your training rhythm — useful for spotting routines or scheduling conflicts.
Day-of-Week Pattern
- What it is
- Which weekdays you train most often.
- How it's worked out
- We count activities by the day of the week they happened on.
- How to read it
- Highlights your weekly rhythm — e.g. weekend long rides, midweek runs.
Performance Detail
HR Histogram (Time in Zone)
- What it is
- How many activities fall into each 10-bpm heart-rate bucket, coloured by training zone.
- How it's worked out
- We round each activity's average HR to the nearest 10 bpm and colour it by Z1–Z5 against your age-predicted max.
- How to read it
- A wide spread = varied intensity. A spike in Z3 is the 'grey zone' to avoid.
HR Box Plot by Sport
- What it is
- How variable your effort is in each sport.
- How it's worked out
- For each sport, the box covers the middle 50% of efforts (Q1–Q3); whiskers show min/max; line in the middle is the median.
- How to read it
- Wider boxes = more variable intensity. Narrow boxes = very consistent effort.
HR vs Duration Scatter
- What it is
- Each dot is one workout — duration on the X-axis, average HR on the Y-axis.
- How it's worked out
- Plotted directly from your activity log, coloured by sport.
- How to read it
- Top-right dots = long efforts at high intensity (impressive but taxing). Bottom-left = easy short sessions.
Elevation Climbed
- What it is
- Total vertical metres gained across all your activities.
- How it's worked out
- Summed from each activity's GPS elevation gain.
- How to read it
- Hilly sessions add stress beyond what distance alone shows. Big climbing weeks deserve extra recovery.
Calories Burned
- What it is
- Estimated energy expenditure across your training.
- How it's worked out
- Strava's estimate based on duration, HR, sport and (if available) power.
- How to read it
- Useful as a rough guide for fuelling — not precise enough for strict diet tracking.
Training Load
Training Load (TRIMP)
- What it is
- A single number that captures how hard a workout was.
- How it's worked out
- Combines duration with average heart rate — long + hard sessions score the highest.
- How to read it
- Higher score = more stress on the body. Use it to plan recovery days.
CTL — Fitness
- What it is
- Your long-term fitness, built up slowly.
- How it's worked out
- A 28-day rolling average of daily training load.
- How to read it
- Higher CTL = fitter. It rises slowly with consistent training and falls with rest.
ATL — Fatigue
- What it is
- Your short-term tiredness from recent workouts.
- How it's worked out
- A 7-day rolling average of daily training load.
- How to read it
- ATL spikes after hard weeks. Give it time to drop before key efforts.
TSB — Form
- What it is
- How fresh you feel right now.
- How it's worked out
- Fitness − Fatigue (CTL − ATL).
- How to read it
- Positive TSB = fresh & race-ready. Negative TSB = tired but building. Strongly negative = at risk of overtraining.
AI Insights
Changepoint Detection
- What it is
- Finds the moments when your training habits shifted significantly.
- How it's worked out
- Scans your weekly distance history and flags weeks where the average level jumped or dropped.
- How to read it
- Changepoints often line up with life events (new job, new sport, injury comeback).
Era Clustering
- What it is
- Groups your training years into 'chapters' that look similar.
- How it's worked out
- Compares each year by volume, intensity and sport-mix, then clusters years that resemble each other.
- How to read it
- Useful for telling the story of your athletic career — e.g. 'cycling era' vs 'multi-sport era'.
Trend & Seasonality (STL)
- What it is
- Splits your training history into the long-term trend, the yearly cycle, and the noise.
- How it's worked out
- A statistical recipe (STL) separates a monthly time series into three clean layers.
- How to read it
- The trend tells you if you're building, the seasonal layer shows your annual rhythm.
Anomaly Detection
- What it is
- Spots workouts that look unusually different from your normal pattern.
- How it's worked out
- Compares each session to your last 28 days. If pace, HR, distance or elevation are way off the average, it's flagged.
- How to read it
- Anomalies are often races, breakthroughs, or bad days — worth a quick look.
HR Age Benchmarking
- What it is
- Compares your observed peak heart rate to the age-predicted Tanaka norm.
- How it's worked out
- A linear regression on your monthly max-HR series is plotted against the Tanaka band (208 − 0.7·age, ±5 bpm). The slope shows year-over-year drift; the gap shows where you sit vs the population.
- How to read it
- Above-band peaks suggest a genetically high ceiling; below-band peaks plus a steep negative slope can hint at fatigue, heat or detraining.
Top Discoveries (auto-surfaced)
- What it is
- The three most newsworthy findings the engine pulls from the patterns above.
- How it's worked out
- Each model emits a candidate insight with a tone and confidence; the engine ranks them and surfaces the top three at the page header.
- How to read it
- Use them as a quick read of what changed — every card links to the detailed model below.
Environment & Gear
Optimal Temperature
- What it is
- The temperature range where you tend to perform best.
- How it's worked out
- We fit a curve through your pace vs. recorded temperature data and pick the peak.
- How to read it
- Most endurance athletes peak between 8–15°C. Hot/cold sessions cost a few % of pace.
Current Conditions Score
- What it is
- A 0–100 score for how favourable today's weather is for training.
- How it's worked out
- Compares current temperature and humidity against your personal optimum.
- How to read it
- 80+ = ideal, 50–80 = workable, below 50 = expect slower paces or higher HR for the same effort.
Activity Detail
Activity Drill-through Table
- What it is
- The raw, row-by-row record of every workout in the current filter.
- How it's worked out
- Each row is one Strava activity with sortable columns (date, time-of-day, sport, name, distance, elapsed/moving time, avg & max HR). The current filter set can be exported as CSV.
- How to read it
- Use it to drill into a specific session, sort by any metric, search by name, or export a filtered CSV for offline analysis.
Year in Review
Peak Year / Peak Month
- What it is
- Your single highest-volume year and month.
- How it's worked out
- We rank every year (and every month) by total distance and pick the top one.
- How to read it
- Useful as a personal benchmark — the bar to beat.
Year-over-Year (YoY) Distance
- What it is
- How this year's distance compares to last year's at the same point.
- How it's worked out
- We compare cumulative distance to date against the equivalent period in the prior year.
- How to read it
- Positive % = ahead of last year. Negative = behind. Useful for staying honest.
Personal Records of the Year
- What it is
- Your single longest, fastest and biggest-climbing session of the selected year.
- How it's worked out
- Within the year filter we pick the activity with the maximum distance, max average speed and max elevation gain respectively.
- How to read it
- A quick highlight reel — the three sessions worth remembering from that year.
Active Days & Longest Streak
- What it is
- How many calendar days you trained in the year, and the longest unbroken run of training days.
- How it's worked out
- We count distinct activity dates within the year and scan for the longest consecutive sequence.
- How to read it
- More active days = stronger habit; long streaks are motivating but rest days still matter for adaptation.
Everything runs on your device.
No data leaves your browser. These insights are designed for personal training reflection — not medical advice.