How We Score Laptops
Every score on Omnission is derived from structured spec data and benchmark indexes — not editorial opinion or advertiser relationships. This page explains the complete scoring pipeline so you can evaluate our rankings with full transparency.
The Scoring Pipeline
- 1
Structured Data
Every laptop is entered into a canonical schema with numeric values for all measurable specs — no free-form text for quantifiable data.
- 2
Dimension Scores
Six dimension scores are computed from raw specs: Performance, Display, Battery, Portability, Value, and Build. Each uses a different formula.
- 3
Value Normalization
The Value score requires the full dataset to compute. It is normalized so the highest value laptop scores 100 and the lowest scores 0, dataset-relative.
- 4
Category Weighting
A composite score is produced by applying a category-specific weight profile. A gaming laptop's composite score uses very different weights than an ultrabook's.
- 5
Cross-Profile Scoring
Every laptop is also scored under all six category profiles, enabling the "Use Case" filter to re-rank the entire dataset without page reloads.
Scoring Dimensions
Six dimensions are scored 0–100. Each is computed from objective, measurable spec data.
Performance
Measures compute capability using indexed CPU and GPU scores derived from benchmark data. The CPU/GPU blend ratio is category-specific — gaming weights the GPU heavily; business work weights the CPU.
Inputs
- CPU benchmark index (0–100)
- GPU benchmark index (0–100)
- Category-specific CPU/GPU blend
Display
Composite of panel technology, refresh rate, peak brightness, color gamut, and resolution. Panel type alone accounts for 40% — the OLED vs IPS gap is real and meaningful.
Inputs
- Panel type (OLED > Mini-LED > IPS > VA > TN)
- Refresh rate (Hz)
- Peak brightness (nits)
- DCI-P3 color coverage (%)
- Resolution (total pixels)
Battery
Primarily driven by real-world estimated battery life (75% weight), supported by raw capacity as a secondary signal (25%). Floor: 3 hours; ceiling: 20 hours.
Inputs
- Estimated real-world life (hours)
- Battery capacity (Wh)
Portability
Inverse of weight and chassis thickness. Lighter and thinner laptops score higher. Weight is weighted at 70% since it is the dominant portability pain point.
Inputs
- Weight (kg) — inverted
- Chassis thickness (mm) — inverted
Value
Performance score divided by price (per $1,000). Normalized dataset-wide so a $499 laptop with adequate performance legitimately outranks a $3,799 flagship. This is the only score that requires the full dataset to compute.
Inputs
- Performance score
- MSRP (USD)
- Dataset-wide min/max normalization
Build
Based on chassis material quality and MIL-STD-810 certification. Carbon fiber and magnesium alloy score highest. MIL-SPEC adds 15 points for real-world durability.
Inputs
- Chassis material (carbon-fiber > magnesium > aluminum > mixed > plastic)
- MIL-STD-810 certification (+15 pts)
Category Weight Profiles
Each category applies different weights when computing the composite score. These weights reflect what buyers in each segment actually care about — not arbitrary choices. All weights sum to 100%.
| Category | Perf | Display | Battery | Portability | Value | Build |
|---|---|---|---|---|---|---|
| Gaming | 40% | 20% | 8% | 7% | 15% | 10% |
| Ultrabook | 20% | 20% | 25% | 20% | 10% | 5% |
| Budget | 25% | 15% | 20% | 15% | 20% | 5% |
| Creator / Design | 35% | 30% | 12% | 8% | 10% | 5% |
| Business / Work | 20% | 20% | 22% | 20% | 13% | 5% |
| 2-in-1 / Convertible | 20% | 22% | 20% | 18% | 12% | 8% |
Blue = highest weight in each row.
Performance Blend: CPU vs GPU
The Performance score is a blend of the CPU and GPU indexes. The ratio is category-dependent because a gaming laptop primarily sells on GPU power, while a business workstation needs a strong CPU.
Gaming
CPU 35%
GPU 65%
Ultrabook
CPU 60%
GPU 40%
Budget
CPU 55%
GPU 45%
Creator / Design
CPU 45%
GPU 55%
Business / Work
CPU 65%
GPU 35%
2-in-1 / Convertible
CPU 60%
GPU 40%
Value Score
Value is the only score that cannot be computed in isolation. It requires the full dataset because it is normalized relative to all laptops in the database.
// Step 1: raw ratio
value_raw = performance_score / (msrp_usd / 1000)
// Step 2: normalize across dataset
value_score = (value_raw − min) / (max − min) × 100
Acer Aspire 15 ($549)
value_raw = 26 / 0.549 ≈ 47.4
Budget — strong value
ROG Zephyrus G16 ($3,779)
value_raw = 94 / 3.779 ≈ 24.9
Flagship — lower value ratio
Data Sources
Benchmark Indexes
CPU and GPU score indexes (0–100) are curated from public benchmark data including Cinebench, Geekbench, 3DMark, and gaming benchmarks. Updated as new generations launch.
Review Consensus
Review scores are sourced from major tech publications: Tom's Guide, NotebookCheck, PCMag, LaptopMag, and RTINGS. Normalized to 0–100 regardless of each publication's own scale.
Manufacturer Specs
All spec data (weight, battery, display, connectivity) comes directly from official manufacturer spec sheets and verified against retail listings.
Retail Pricing
MSRP prices are from manufacturer listings. Live prices are fetched from Amazon product pages via our price worker and cached for 30 minutes at the Cloudflare edge.
Data Freshness
Every laptop record includes a last_updated date visible on detail pages. When specs or prices change significantly, records are updated manually.
Benchmark indexes are reviewed when new CPU or GPU generations launch. Category weight profiles are reviewed when new product segments emerge (e.g., the Snapdragon X Elite generation prompted review of the business-work and 2-in-1 profiles).
Live prices are fetched from Amazon at page-load time and cached for 30 minutes per ASIN. MSRP is always shown as a fallback if live pricing is unavailable.