Normalized Power Explained

Average power is the most lied-to number in cycling. You finish a punchy three-hour ride with five climbs, a couple of group-ride surges, and a tailwind cruise home, and your computer tells you 187 watts average. Cool story. Your legs know it cost you a lot more than a steady 187-watt effort would have. Normalized Power is the metric that catches the lie.

This guide unpacks what NP actually computes, why it matters for prescribing training load, where it breaks down, and how the Movement Rebels coach uses your recent NP (not raw average watts) to scale the next interval session.

Why Average Power Is a Bad Estimate of Cost

Physiological cost scales nonlinearly with intensity. Two minutes at 350 watts followed by two minutes at 50 watts is not the same as four minutes at 200 watts, even though the average is identical. The hard surge taps glycolytic systems, dumps lactate, recruits Type IIa fibers, and leaves a metabolic hangover the easy bit cannot undo.

Average power treats every second as equal. Real-world rides almost never are. Stops at lights, descents at zero watts, climbs at threshold, attacks above FTP, recovery between efforts — average just smears them into one meaningless number. For a steady indoor trainer ride at a fixed wattage, average power equals NP. For any ride with variability (which is most outdoor riding), average power undersells the cost, often badly.

Andy Coggan introduced Normalized Power in the early 2000s precisely because riders and coaches needed a single number that approximated "what steady wattage would have produced the same physiological cost as this messy ride?" That single number is what NP gives you.

The Math (Quick, So You Trust It)

NP is computed in four steps. You do not need to do this by hand — Garmin, your head unit, Strava, and the MR coach all calculate it — but knowing the steps explains why it behaves the way it does.

1. Take a 30-second rolling average of power across the entire ride. This smooths out the spikes a power meter naturally produces from pedal-stroke variation.
2. Raise each smoothed value to the fourth power.
3. Take the average of all those fourth-power values.
4. Take the fourth root of that average.

The fourth-power weighting is the magic. It pushes the hard moments way up the curve while leaving the easy moments roughly where they are. A 30-second segment at 400 watts contributes 256× more to the NP calculation than a 30-second segment at 100 watts (4× the power, raised to the 4th, equals 256× the weight). That matches the rough shape of how lactate, ventilation, and substrate utilization scale with intensity.

The 30-second smoothing exists because metabolic responses lag the pedal stroke. A 5-second spike to 800 watts during an attack doesn't cost what 800 watts sustained would. The rolling window approximates how your body actually integrates load over time.

Variability Index — The Other Number You Should Read

Once you have NP, you get a free bonus metric: Variability Index (VI). VI = NP divided by average power. A perfectly steady ride has VI = 1.00. A surgy criterium might come in at 1.25 or higher. A solo time trial done right is usually 1.02–1.05.

VI tells you something average power and NP separately cannot: how cleanly you executed the ride. For a long aerobic ride aimed at building Zone 2 base, VI above 1.10 is a sign you spiked too much (probably climbing too hard) and the session drifted out of its intended adaptation zone. For a race or fartlek session, high VI is the whole point.

The MR coach reads VI on every cycling activity and flags it when the intent and execution drift. If you scheduled "endurance Z2, 2 hours" and came back with VI 1.18 and a third of the time above threshold, the coach knows the next day's plan needs to absorb more recovery, not piling another hard session on cooked legs.

NP, FTP, and Intensity Factor

NP also lets you compute Intensity Factor (IF) = NP divided by FTP. A two-hour ride at IF 0.75 was a solid endurance ride. The same duration at IF 0.92 was a brutal threshold day, even if average power suggested otherwise. IF is the input to Training Stress Score, which we cover in detail in our Training Stress Score guide.

Stack of relationships you should burn into memory:

• Average power = the raw mean, lies for variable rides.
• Normalized Power = physiological-cost-weighted mean, the honest number.
• VI = NP / avg power, tells you how steady the ride was.
• IF = NP / FTP, tells you how hard the ride was relative to your fitness.
• TSS = duration × IF² × 100, the load number that feeds periodization.

Every one of those numbers is downstream of NP. Get NP right and the rest behaves.

Where NP Breaks Down

NP is not gospel. Three honest limits:

Very short efforts. A 20-minute crit with massive surges will produce an NP that the four-hour aerobic system math wasn't really designed for. It still works directionally, but treat short, max-effort NP with a grain of salt.

Rides with long zero-power stretches. Long descents or stoplight-heavy commutes can pull avg power down enough that VI spikes artificially. The cost wasn't as high as VI suggests because you actually recovered during those zeros.

Trainer vs outdoor. Same NP indoors usually feels harder because there's no descent recovery, no cooling, no microbreaks. Don't mechanically equate them.

The MR coach factors these in: indoor sessions get a small intensity adjustment in the load model, and rides flagged as commutes (sub-45-min, low duration) don't drive the same forward prescription that a structured ride does.

How Movement Rebels Uses NP

The coach reads NP from your Garmin or Strava sync (Garmin native, Strava read+write native) and uses it as the primary load signal for cycling, not raw average power. Three places this shows up in your plan:

• Adaptive interval prescription. When the coach writes Wednesday's threshold session, it scales the target watts off your last 4 weeks of NP at threshold-ish efforts, not your stored FTP guess. If you've been holding higher NP at lower perceived effort, the next interval block bumps up. If recent NP is sagging while RPE is climbing, the coach pulls back and writes a lighter day — and tells you why.
• Cross-domain context. If your last three rides show declining NP at the same HR and your Rebel Fuel log shows you've been 600 kcal under target three days running, the coach connects those dots in the morning brief. You're not under-trained, you're under-fueled. That kind of cross-read is what an app with strength logging, Rebel Fuel, breathwork, NSDR, biohack history, and HRV-guided readiness in one timeline can do that a single-purpose power-analysis tool can't.
• Honest VI feedback. If your "endurance" ride keeps coming back at VI 1.15, the coach calls it. You're not riding endurance, you're riding stochastic threshold. Recovery debt accumulates either way.

Pricing

Movement Rebels is one app instead of five — strength logger, endurance planning, hybrid programming, snap meal, hydration, breathwork, NSDR, cold exposure, fasting timer, body comp, biohack history, and the always-on coach all under one subscription. Free 7-day trial, no card. Pro+ after that.

A note on the social side: MR replaces Strava's training value (planning, logging, AI coaching, recovery read, fueling), not its social feed. Segments, kudos, followers — keep Strava for those. The MR↔Strava integration is read+write, so your rides still land in your friends' feed via Strava with the coach's session summary written back into the description.