Your score de sommeil says 72. Again. You went to bed on time, slept through the night, and feel mostly fine -- yet the number says otherwise.
The problem: most people optimize the wrong variables. They chase total sleep duration when the algorithm is penalizing their HRV. They fixate on bedtime when sommeil profond percentage is the bottleneck. This guide shows you how to read the data your device already collects and fix what actually matters.
5
Core metrics
That drive 80% of your score de sommeil
15–23%
Typical sommeil profond
Target range for adults 25–55
42 ms
Average HRV
Population median (age-adjusted)
3–4 weeks
Improvement timeline
For measurable score increases
Verdict rapide
HRV Optimization
Improving overnight HRV produces the single largest impact on both Oura and WHOOP scores.
How Your Score Is Actually Calculated
Oura's Score de sommeil (0-100) weights seven contributors: total sleep (~25%), sommeil profond (~20%), REM (~15%), efficiency (~10%), restfulness (~10%), latency (~10%), and timing (~10%).
WHOOP's Recovery (0-100%) weights differently: HRV (~40%), sleep performance (~25%), RHR (~20%), fréquence respiratoire (~10%), consistency (~5%).
The Critical Insight
HRV and sommeil profond are disproportionately influential on both platforms. Someone sleeping 7 hours with great HRV and 20% sommeil profond will consistently outscore someone sleeping 9 hours with poor HRV and 12% sommeil profond. Duration matters, but it is not king.
The Five Metrics That Move the Needle
| Rank | Metric | Why It Matters |
|---|---|---|
| 1 | Overnight HRV | Direct measure of parasympathetic recovery; most volatile and most improvable |
| 2 | Sommeil profond % | Physical recovery gatekeeper -- below 15%, you'll never break 90 |
| 3 | Sleep Efficiency | Time in bed is not time asleep; low efficiency drags everything |
| 4 | Resting Fréquence cardiaque | Elevated RHR = incomplete recovery; alcohol, stress, illness spike it |
| 5 | Timing Consistency | Irregular timing destabilizes circadian rhythm, suppressing sommeil profond and HRV |
Metric 1: Overnight HRV
High HRV = flexible, recovered nervous system. Low HRV = something is suppressing your recovery. On WHOOP, HRV is ~40% of your Recovery. On Oura, it feeds into Readiness through multiple channels.
The biggest HRV killers:
- Alcohol (2+ drinks near bed): 15-25% reduction for 12-36 hours
- Late eating (within 2h of bed): 5-12% reduction
- Late intense exercise: 8-15% reduction
- Room temp above 21C / 70F: 8-18% reduction
The 10-3-2-1 Rule
10h before bed: no more caffeine. 3h: no more food or alcohol. 2h: no more work. 1h: no more screens. Even partial adherence measurably improves overnight HRV within two weeks.
The fix: Week 1-2, eliminate your biggest suppressor (check WHOOP Journal or Oura Trends). Week 3-4, add resonance breathing before bed (5.5 breaths/min for 5-10 min) -- shown to increase overnight HRV by 10-18%. Week 5+, optimize bedroom environment: 16-19C, total darkness, below 30 dB.
Metric 2: Sommeil profond
Sommeil profond (N3) is when growth hormone peaks, the immune system activates, and the glymphatic system flushes brain waste at 60% higher rates. If your sommeil profond is consistently below 15%, your score will never break 90.
Three things that destroy sommeil profond:
- Alcohol -- even one drink within 3h of bed reduces N3 by 9-12%
- Late meals -- digestion elevates core temp and RHR, suppressing N3 by 7-15%
- Warm bedrooms -- N3 onset requires a 1-2F core temp drop; optimal is 16-18C (61-65F)
The Melatonin Misconception
Melatonin helps you fall asleep faster but has no proven effect on sommeil profond duration. The behaviors that actually increase N3 -- earlier alcohol cutoff, cooler rooms, consistent timing, regular exercise -- are less convenient but far more effective.
sleep-score-improvement
Unify all your wearable data and get personalized AI health insights in one place.
Metric 3: Sleep Efficiency
Sleep efficiency = time asleep / time in bed. If you're in bed 8 hours but sleeping 6.5, your efficiency is 81% -- and both platforms will penalize you. Target: above 90%.
Fix it: Don't get into bed until you're genuinely sleepy. If you're not asleep in 20 minutes, get up. Add white noise or earplugs (pink noise reduces wake-after-sleep-onset by 38%). Keep your bedroom pitch dark.
Metrics 4 & 5: RHR and Timing
Resting fréquence cardiaque mirrors your HRV protocol -- reduce sympathetic activation before bed, stay hydrated (dehydration raises RHR by 5-10 bpm), and maintain cardiovascular fitness.
Sleep timing consistency is the least glamorous but surprisingly impactful metric. A 90-minute bedtime shift on weekends can suppress Monday's Recovery by 15-25 points. Keep your wake time within 30 minutes every single day, including weekends.
Skip the Weekend Sleep-In
Sleeping in to "repay" dette de sommeil shifts your circadian phase later, delays Sunday night sleep onset, and tanks your Monday score. Instead, keep your wake time fixed and add a 20-minute afternoon nap if needed.
The 30-Day Protocol
| Week | Focus | Actions | Expected Impact |
|---|---|---|---|
| 1 | Baseline + remove #1 suppressor | Track all 5 metrics; eliminate biggest HRV suppressor | 3-8 point HRV improvement |
| 2 | Sommeil profond | Room to 18C; last meal 3h before bed; warm shower 90 min before bed | 5-15% sommeil profond increase |
| 3 | Efficiency + timing | Fixed bedtime +/-15 min; white noise; no bed until sleepy | Efficiency +3-8% |
| 4 | HRV optimization | Add resonance breathing; ensure hydration; review remaining suppressors | HRV +8-15%; total gain 12-25 points |
Real Results
Most users see 3-8 points in week one (behavioral changes hit fast), compounding gains through weeks 2-3 as circadian rhythm stabilizes, and plateau at 85-92 by week 4. Consistently above 92 requires near-perfect compliance and favorable genetics. The goal is not perfection -- it's moving from compromised recovery (65-75) to solid recovery (85-92).
✓Pros
- Oura excels at stadification du sommeil accuracy and temperature-based insights
- WHOOP excels at recovery analytics and behavioral correlation via Journal
- Both provide sufficient data to follow this protocol
- Combining both devices gives the most complete picture
✗Cons
- Scores are relative to your baseline -- two people with identical sleep can get different scores
- Consumer stadification du sommeil is ~78% accurate (not perfect)
- Reaching 90+ consistently requires significant lifestyle changes
- Melatonin and most supplements won't move the needle on sommeil profond
Frequently Asked Questions
Neither is 'right' absolutely. Oura emphasizes sleep architecture (stages, duration, efficiency). WHOOP emphasizes autonomic recovery (HRV, RHR). You can have great architecture with poor recovery (after heavy training) or vice versa. Use both for a complete picture.
Short naps (15-25 min) before 2 PM generally don't affect nighttime scores. Naps over 45 min or after 3 PM can reduce sleep pressure, making it harder to fall asleep and cutting sommeil profond in the first cycle.
Magnesium glycinate (200-400 mg, 60 min before bed) has the most evidence for improving HRV and sommeil profond. L-theanine has moderate evidence. Everything else (CBD, valerian, tart cherry) lacks controlled evidence. Introduce one at a time and run a 14-day test.
Alcohol hits all five score drivers at once: suppresses HRV (15-25%), reduces sommeil profond (9-25%), elevates RHR, decreases efficiency through rebound waking, and disrupts timing. No other single behavior has this combined impact.
Technically possible on Oura, extremely rare. On WHOOP, 99-100% occurs on ~1-2% of nights. The clinically meaningful target is sustained scores above 85. The difference between 88 and 95 is insignificant for next-day performance.
