Forkin' Worth It

Forkin methodology · version 1.0

How the Forkin Score Works

How the score is calculated, in full — what counts as evidence, what each factor moves, and what we refuse to use.

What is the Forkin Score?

The Forkin Score is a 0–10 desk-evidence calibration, not a personal dining opinion or a crowd-rating average. It is built entirely from published, independent and attributable food journalism, then rounded to one decimal place.

We begin each calculation within its assigned evidence band, assess the quality and shape of the supporting evidence, and keep the final result within that band. The score is designed to make our evidence legible — not to impersonate a universal truth about a meal.

The Algorithm

Start with the editorial band’s base score, apply five deliberately small evidence adjustments, then clamp the result to that band. No single factor can move a venue into another tier.

01

Source Consensus

How many independent, attributable sources agree.

  • Strong +0.15
  • Moderate +0.05
  • Limited −0.15
02

Evidence Recency

How current the published assessment or verification is.

  • Current +0.05
  • Recent +0.00
  • Dated −0.10
03

Praise Specificity

Concrete dish and experience signals, rather than generic praise.

  • High +0.10
  • Medium +0.00
  • Low −0.10
04

Critic Calibre

The editorial standing of the publication providing the evidence.

  • Major +0.15
  • Established +0.05
  • Independent +0.00
05

Consistency

The degree of agreement across the attributable sources.

  • High +0.10
  • Moderate +0.00
  • Low −0.15

Base band score + evidence adjustments → band clamp → rounded to one decimal place

Score Bands

8.5–9.5

Forkin Worth It

Strong consensus across substantive independent coverage.

7.0–8.4

Worth a Look

Provisional evidence — positive but thinner independent coverage.

5.5–6.9

Mixed Forks

Material disagreement between substantive sources.

—

Unrated

Below the evidence bar, so no verdict is published at all.

What We Exclude

  • Google reviews and other crowd-rating data
  • Sponsored or paid-for content
  • AI-generated sentiment analysis
  • Our own editorial opinions — stamps are separate
  • Personal dining experiences

Independence & Transparency

The numeric score is calculated independently from editorial stamp copy, even where the two may naturally correlate; its evidence band is a published safeguard, not a paid placement. No venue can pay for a higher score, and no Google data is used. Every source used is attributable and disclosed.

Where independent evidence is not sufficient, we publish no verdict at all rather than invent precision. We do not publish an UNRATED verdict — a venue below the bar simply carries no score and no stamp, which is why many listings show neither.

Last algorithm update: August 2026 · Version 1.0