Data Lab: Deep analysis for every ride

Most cycling apps give you a ride summary. Data Lab gives you a training analysis. Select a date range, generate a report, and get 17 interconnected modules covering power output, fitness trends, energy systems, recovery signals, pacing, efficiency, and activity patterns. You can also import rides from FIT files or CSV exports if you want to analyze data from outside Ride Cave.

Generating a report

Open Data Lab from the main menu and click New Report. You have three ways to generate one.

Ride Cave history is the default. Pick a date range and Ride Cave analyzes all your completed rides in that window. This path gives you the most complete data because every metric—power, heart rate, cadence, strain, zone times, decoupling—was captured natively during your session.

FIT files let you upload activity files from Garmin, Wahoo, or any device that records in the FIT format. Ride Cave reads both the session summary and the raw second-by-second data stream, so the resulting report is nearly as rich as a native Ride Cave report. Peak powers, zone times, normalized power, aerobic decoupling, and energy system strain are all computed from the stream data. The more rides you upload, the better the analysis. A single FIT file gives you a snapshot. Dozens give you trend lines worth acting on.

CSV files work with exports from platforms like Strava, TrainingPeaks, Intervals.icu, and others. CSV imports are useful for summary-level analysis—training load trends, volume distribution, activity patterns. However, CSVs only contain per-ride totals, not second-by-second streams. That means peak powers, power zone times, heart rate zone times, aerobic decoupling, energy system breakdown, and the efficiency metrics cannot be computed from CSV data alone. Modules that depend on stream data will show an empty state with a note explaining what is missing. If you have FIT files available, use those instead.

Free accounts can import up to 20 FIT files or 20 rides from a CSV per import. Cave Crew members have no limit.

What is in a report

Every report starts with a summary strip at the top. Seven headline numbers give you the shape of the date range at a glance: total Rides, total Time, total Training Load (TSS), total Energy in kilojoules and kilocalories, current Fitness (CTL — your 42-day average training load), current Fatigue (ATL — your 7-day average), and current Form (TSB — fitness minus fatigue). Below that, 17 modules break down everything in detail.

Overview

Summary Stats aggregates your totals across the date range: rides, hours, training load, kilojoules, distance, elevation gain, and calories. It also shows average intensity factor, average TSS per ride, the number of Everests you have climbed, your CO2 offset equivalent, and a breakdown of rides by workout type.

Power Profile grades your best power outputs at five key durations — 5 seconds, 1 minute, 5 minutes, 20 minutes, and 60 minutes — against population benchmarks. Each duration gets an individual grade from Untrained through World Class, plus a watts and watts-per-kg figure. An overall performance score (0-100) rolls these into a single number.

Energy Systems shows how your training stress distributes across the three physiological systems: aerobic (low-intensity work below your aerobic threshold), glycolytic (hard efforts between threshold and VO2max), and neuromuscular (peak efforts above CP). You see absolute strain values and percentage splits. This module is unique to Ride Cave and requires power data — it is not available from CSV imports.

Power

Power Curve plots your best average power at every duration from a few seconds to multiple hours, using all rides in the date range. The curve also fits a Critical Power model to your 2-minute through 20-minute data points, giving you an estimated CP (your sustainable power ceiling) and W' (your anaerobic reserve in joules), along with the 95% confidence interval and goodness of fit. A phenotype radar displays normalized scores across five dimensions — sprint, anaerobic, VO2max, threshold, and endurance — and classifies you as a Sprinter, Pursuiter, Time Trialist, Rouleur, All-Rounder, or Climber based on your curve shape. The module also tracks which specific ride set each duration record.

Sprint and Neuromuscular tracks your peak power trends over time at sprint durations (5 seconds, 60 seconds). Each point on the chart is a ride. You can see whether your top-end power is improving, declining, or stable across the range. The module also classifies your sprint type as explosive (5s dominates), sustained (60s is relatively stronger), or balanced.

Effort Index measures how demanding each ride was based on pacing consistency and anaerobic contribution. A ride that was steady and mostly aerobic scores low. A ride with deep anaerobic efforts, hard accelerations, or significant power fade scores higher. The chart shows your effort index per ride over time, and the module shows your average across the range.

Training

Fitness Model is the Performance Management Chart: Fitness (CTL), Fatigue (ATL), and Form (TSB) plotted daily across the date range. It also shows ramp rate (how quickly your CTL is changing), monotony (how varied your training stress is day to day), and Bannister strain (weekly TSS multiplied by monotony). A second set of time-series tracks your energy system CTL separately — aerobic fitness, glycolytic fitness, and neuromuscular fitness each get their own trend line.

Training Distribution breaks your training into weekly summaries. Each week shows hours, TSS, kilojoules, ride count, and average intensity. A stacked bar chart shows how much time you spent in each of the seven power zones. Two derived metrics per week help identify your training style: polarization index (the share of time spent in Z1-Z2 vs. everything harder) and sweet spot ratio (time in Z3-Z4 as a fraction of total training). Volume ramp percentage shows how each week compared to the previous. The module also detects structured training phases — base, build, peak, recovery, and taper — from the weekly volume and intensity patterns.

Interval Analysis has two parts. The first is a W' balance depth chart: for each ride, it shows how far into your anaerobic reserve you went. Rides that barely touched your W' sit near zero. Rides with hard repeated efforts dip deep. The second part shows your best efforts across seven interval categories — sprint, anaerobic, VO2max, threshold, sweet spot, tempo, and endurance — with power, duration, and watts-per-kg for each.

Workload Ratio is the acute-to-chronic workload ratio (ACWR), plotted daily. A value between 0.8 and 1.3 is generally considered the optimal training zone. Above 1.5 indicates a significant spike relative to your fitness base. The module shows your current ACWR value and a status: undertraining, optimal, caution, or danger.

Pacing Consistency tracks variability index (VI) per ride — the ratio of normalized power to average power. A VI close to 1.0 means steady, controlled output. Higher values indicate more surging and variable pacing. The module plots VI across all rides in the range and shows your average.

Body

Recovery Status tracks two heart rate signals over time: warmup HR and max HR. A rising warmup HR or declining max HR can indicate accumulated fatigue. The module synthesizes these into an overall recovery status — fresh, monitor, caution, or overreaching — and surfaces individual recovery signals with a green, yellow, or red indicator explaining what the signal means.

Heart Rate Zones shows your weekly time distribution across five HR zones. Each week is a stacked bar showing how much time you spent at each intensity from a cardiovascular perspective. This complements the power zone distribution in Training Distribution.

Efficiency and Economy has three parts. Efficiency Factor (EF) is your average power divided by average heart rate for aerobic rides — a higher EF means more power for the same cardiac cost. The module plots EF over time and fits a linear trend to show whether your aerobic efficiency is improving. Aerobic decoupling is plotted separately: it measures how much your power-to-HR ratio drifted between the first and second halves of each ride, where lower is better. Cadence efficiency bins show your average EF across different cadence ranges, which can reveal whether you are riding at an optimal pedaling rate for your aerobic output.

Activity

Activity Calendar renders a daily heatmap of training load across the date range. Darker squares mean higher TSS. Below the heatmap, you get your current consecutive-day riding streak, your longest streak in the range, and a day-of-week breakdown showing both ride frequency and average TSS by day. Monday through Sunday at a glance tells you a lot about your training habits.

Time of Day buckets your rides by the hour they started (0-23) and shows your ride frequency distribution. A second chart shows average power by hour, so you can see whether you tend to ride harder in the morning, afternoon, or evening. The module also surfaces your most active time window.

Route and Climbing shows your weekly elevation gain as a bar chart, giving you a simple view of how much climbing volume you have accumulated and how it trends across the date range.

Saving and sharing reports

Once a report is generated, you can save it with a name. Saved reports appear on the Data Lab home screen and can be loaded instantly — the full analysis is stored so there is no re-computation wait. You can also export any report to PDF from the report header.

Free vs. Cave Crew

Free accounts can generate reports using any date range preset up to 30 days and can save up to 2 reports. FIT and CSV imports are limited to 20 activities per upload. All 17 analysis modules are available.

Cave Crew members get longer presets — 90 days, 6 months, 1 year, and all-time — custom date ranges, unlimited saved reports, and unlimited imports. If you have years of training data and want to see long-term trends in fitness, power curves, and training distribution, the extended ranges make a significant difference. More data means more accurate CP and W' models, more stable trend lines, and clearer training phase detection.