Methodology v2.1
How We Test Mattresses
9 weighted scoring factors. 43 individual measurements per mattress. Real equipment, real numbers, and the exact formulas we use to turn raw data into the 1–10 scores you see in every review.
Why we publish the raw measurements
Most mattress review sites give you a single number — say, "9.4 for cooling" — and ask you to trust it. We don't think that's good enough.
On every review where we have lab data, we publish the actual measurement: the surface temperature delta in degrees Fahrenheit, the peak g-force in motion-transfer units, the inches of edge sinkage. You can verify our scoring math, compare mattresses on the metric you care about most, and re-run the calculation under your own weights if you disagree with ours.
Our scoring formulas are public. Every factor below shows the equation that converts a raw measurement into the 1–10 score. If a mattress scores 8.4 for cooling, you can plug the published delta back into the formula and get 8.4.
The 9 Scoring Factors
Each factor has a fixed weight (totalling 100%) that contributes to the overall 1–10 score. Tap any factor for the equipment, protocol, formula, and dataset benchmarks.
What we measure but don't score
Some metrics are useful as context but don't deserve a weighted slot in the overall score. We publish them on every lab page so you can use them however you want.
- Sinkage. Total compression at the lumbar under a 230 lb back sleeper. Useful for comparing 'hug' but doesn't translate cleanly to a single sleep outcome.
- Bounce. Steel-ball rebound percentage. Predicts feel (latex-bouncy vs. memory-foam-dead) but isn't itself good or bad.
- Cover composition. Fiber content and weight (oz/yd²). Reported, but our durability simulation captures the practical impact.
Methodology changelog
Our scoring system evolves as we learn. Major version bumps re-score the entire test catalog so comparisons stay apples-to-apples.
v2.1
January 2026Added durability simulation (30,000-cycle roller test). Adjusted Cooling weight from 16% to 18% based on user-survey signal. Re-scored top 50 mattresses.
v2.0
August 2025Replaced subjective firmness vote with calibrated indentation force deflection (IFD). Added pressure-mapping at 4 body points instead of just shoulder. Full catalog re-test.
v1.4
March 2024Added off-gassing quantitative VOC reading alongside the trained sniff panel.
v1.0
March 2018Initial 7-factor rubric with 22 raw measurements.
Errors & corrections policy
We make mistakes. When we do, we fix them publicly. If a manufacturer changes the construction of a mattress mid-cycle, we re-test and update the published numbers — and we annotate the review with the date and what changed.
Spotted something that looks wrong? Tell us. We respond to every correction request within 5 business days.
Submit a correctionHow Sleep Sage researches sleep
Sleep Sage publishes independent research and product guidance using hands-on product evaluation, manufacturer specifications, peer-reviewed research, published safety guidance, owner feedback, and other reputable sources. Health-related information is provided for educational purposes and should not be considered medical advice. We do not employ clinicians, and our pages are not medically reviewed.