About

Public Methodology · v2.0

How the ClarityPoint score is built.

The pipeline in four steps

  1. Source acquisition — automated discovery surfaces reviews from a curated panel (RTINGS, The Verge, Wirecutter, Tom’s Guide, CNET, Engadget, SoundGuys, What Hi-Fi?, MKBHD, DC Rainmaker, DPReview, others by category). Only reviews from the panel count. Minimum 3 sources per consensus page; Amazon Buyers aggregate is allowed as one source when ≥50 ratings exist.
  2. Score normalization — every native scale converted to 0–100. Star ratings, /10 scores, /100 scores, percentage scores, Editor’s Choice / Top Pick designations all mapped via a published lookup table (see below). Missing sources are excluded from the average, never treated as zero.
  3. Aggregation — equal-weight mean across the panel. Subscores (ANC / battery / display / etc.) aggregated the same way per-dimension. Confidence flag attached based on panel breadth and inter-rater spread.
  4. Editorial verification — every takeaway on a page can be traced back to the source review’s URL. No editorial inference is added between the source and the page.

Normalization table

Native scale Mapping rule Example
0–10 numeric × 10 9.2 / 10 → 92
0–5 stars × 20 4.5 / 5 → 90
0–100 numeric passthrough 87 / 100 → 87
Percentage passthrough 88 % → 88
“Top Pick” / “Editor’s Choice” → 95 Wirecutter Top Pick → 95
“Recommend” → 88 MKBHD Recommend → 88
“Strong recommend” → 92 DC Rainmaker → 92
“Runner-up” → 85 Wirecutter Runner-up → 85
DxOMark sensor score (score/110)×100 102 → 93
Amazon Buyers aggregate native_stars × 20, gated ≥50 ratings 4.7 / 5 (8,712) → 94

Editorial team

No anonymous byline. Every page is signed by a category editor responsible for source selection, normalization audit, and verdict.

Marcus Cho
Headphones & Audio editor
Priya Patel
Phones & Mobile editor
Sam Lambert
Wearables & Health editor
Alex Chen
Computing & Tablets editor
Jordan Reese
Cameras & Imaging editor
Dani Kovac
Home Theater & TV editor

API access

Programmatic access to consensus scores, comparison matrices, and generation deltas. Endpoints:

GET /api/v1/consensus/{slug} # 0–100 score, reviewer panel, subscores
GET /api/v1/comparison/{a-vs-b} # head-to-head matrix
GET /api/v1/contrast/{family_key} # generation timeline + upgrade matrix
GET /api/v1/upgrade/{from}/{to} # specific upgrade verdict

Rate-limited free tier for personal use; commercial use from $500/mo. Contact for API key →

Are you sure want to unlock this post?
Unlock left : 0
Are you sure want to cancel subscription?