How to analyze your fatigue curve using power data

Your power-duration curve shows what you can do from a few seconds to hours. The fatigue curve is the part that describes how your sustainable watts fall as the duration gets longer. Watching how that long-duration end changes week to week is one of the fastest ways to spot overreaching before it turns into a slump.

This guide explains how to build and interpret your fatigue curve, what patterns flag non-functional overreaching, and how to adjust training and recovery when you see the signs.

What is a fatigue curve?

Think of the fatigue curve as the slope from your shorter efforts to your longer efforts. If the curve gets steeper, your long-duration power (30–120 minutes) is falling faster than expected relative to your shorter durations. That often reflects reduced durability, poor fueling, or accumulating fatigue.

Common reference points along the curve:

• Short power: 1–5 minutes (anaerobic capacity, VO2max).
• Threshold range: 20–60 minutes (around FTP; time to exhaustion, TTE).
• Endurance range: 90–180 minutes (aerobic durability and stamina).

Several practical metrics capture the curve’s shape:

• P20, P60, P120: Best 20-, 60-, and 120-minute power over a rolling 28–56-day window.
• TTE at FTP: How long you can actually hold your tested FTP.
• Decoupling (power–heart rate drift): How much power drops or HR rises during steady endurance at a fixed intensity (e.g., Zone 2).

Fatigue slope (20→60) % = 100 × (P60 − P20) / P20
Durability ratio (60/20) = P60 / P20
Decoupling % (steady ride) ≈ 100 × (second-half power − first-half power) / first-half power

Track these values over time rather than obsessing over a single test day.

Build and read your curve step by step

1. Collect clean efforts. Over 3–6 weeks, include a few steady long efforts (30–120 minutes) and at least one controlled threshold effort. Log RPE, sleep, fueling, heat, and altitude.
2. Extract MMP points. From your power files, record the best 20-, 60-, and 120-minute powers (P20, P60, P120). Also note TTE at FTP, if you test it.
3. Compute ratios. Calculate P60/P20 and P120/P60. These remove day-to-day noise and highlight changes in the curve’s slope.
4. Check decoupling. Do a 2–3 hour Zone 2 ride at a steady target and calculate drift between halves. Keep conditions comparable.
5. Compare to your baseline. Look at 4–8 weeks of history. Normal week-to-week variation is small; bigger, persistent drops matter.

In this example, short-to-mid power holds, but the 60–120 minute end slips. The steeper curve suggests reduced durability rather than a true drop in aerobic capacity.

How long-duration power drops reveal overreaching

Functional overreaching is a short, planned dip that rebounds with recovery. Non-functional overreaching is when you keep pushing and performance declines persist.

Red flags in your fatigue curve:

• Disproportionate long-duration drop: P60 or P120 down 5–8% or more vs baseline for 7–10+ days while P5–P20 is relatively stable.
• TTE at FTP shrinks: e.g., from 50–60 minutes to 30–35 minutes at the same tested FTP.
• Decoupling increases: Drift >7–10% on steady Zone 2 rides of 2–3 hours in similar conditions.
• Subjective load is high: Elevated RPE at endurance watts, poor sleep, low mood, sore legs on easy days.

Check confounders before you call it overreaching:

• Heat, altitude, wind, and route changes.
• Dehydration or low carbohydrate availability (under-fueling).
• Bike fit or position changes (aero vs upright).
• Illness coming on or high life stress.
• Power meter calibration differences.

What to do when the curve tilts

If confounders are ruled out and the long end stays suppressed, back off early. A small course correction is faster than digging out of a hole.

• Deload 3–7 days: Reduce volume by 30–50% and limit intensity to low Zone 2 with short strides only.
• Fuel properly: 6–8 g/kg/day carbohydrate on training days; 60–90 g carbs per hour on long rides; 20–30 g protein in meals and post-ride.
• Hydrate: 500–750 ml per hour in temperate conditions, 700–1000 ml in heat; add sodium 500–1000 mg per hour based on sweat rate.
• Sleep: Target 7.5–9 hours, plus a short nap if needed.
• Retest simply: After the deload, repeat one steady 60-minute effort or a TTE at FTP. Your P60, P120, and decoupling should rebound if it was functional overreaching.

Simple field protocols to monitor durability

1) Endurance decoupling check

Ride 2–3 hours at mid Zone 2. Split the file in half and compute drift.

Decoupling % = 100 × (Power₂ − Power₁) / Power₁
Goal (temperate, fueled): ≤5% drift

If drift climbs above 7–10% several weeks in a row, suspect accumulating fatigue or under-fueling.

2) FTP and TTE snapshot

After an easy day, ride at your current FTP and time how long you can maintain it evenly (TTE). If TTE falls by 20%+ from your recent best while your short efforts are unchanged, you are likely overreached.

3) Durability under load

Do 2–3 hours Zone 2, then attempt 1×20–30 minutes at or just below FTP. Compare the watts to a fresh-day effort:

• If the post-ride interval drops >10–15 W vs fresh, and RPE is much higher, durability is compromised.
• Repeat every 2–3 weeks to track progress.

Putting it together in your training plan

• Weekly: Log P20, P60, P120, TTE, and one decoupling test.
• Block-level: Expect small, temporary dips late in a build. Plan a recovery week every 3–5 weeks.
• Return to intensity: When decoupling is ≤5% and P60/P20 returns within ~2–3% of baseline, reintroduce threshold and VO2max work.

Durability is trainable. If you consistently fuel endurance rides, progress long intervals (e.g., 2×20 → 1×40–60 minutes near FTP), and manage recovery, the long end of your curve will flatten. That means steadier watts deep into rides and races—and fewer surprises from overreaching.