How We Score
Every restaurant on SushiMap is evaluated across 6 axes, totaling 100 points. We call this the FitScore — it measures how well a shop fits our philosophy of "local advantage, intimate counter, honest craft."
Total Points
A. Local Advantage
30 ptsHow much of the neta comes from local ports and nearby seas? Direct relationships with local fishermen and markets score highest. A sushi shop in Toyama using Toyama Bay fish earns more than one importing from Tsukiji.
B. Intimate Counter
20 pts6-10 counter seats is ideal. 11-14 is acceptable with deductions. 15+ seats significantly reduces the score. Single seating (no rotation) scores highest; double is acceptable; triple or more incurs heavy penalties.
C. Price Sweet Spot
20 ptsThe sweet spot is ¥22,000-27,000 (tax included) for dinner omakase. This is not about cheapness — it is about the zone where quality, generosity, and value align. Below ¥15,000 or approaching ¥30,000 gets lower marks, though composition and satisfaction can adjust scores.
D. Honest Craft
15 ptsProper knife work, aging (shime, kobujime), careful nikiri application, and a high nigiri ratio. Multi-course extravaganzas that bury the sushi under appetizers lose points. We want sushi-forward omakase.
E. Photo Friendly
10 ptsCan you photograph your food without tension? "Photos welcome" scores highest. "Please ask first" is fine. Strict no-photo policies reduce the score — not because photography is a right, but because it signals the overall atmosphere.
F. Calm Atmosphere
5 ptsIs the focus on eating rather than entertaining? Corporate entertainment crowds, Instagram showmanship, and excessive theatrical presentations lose points. Quiet concentration on the food is what we value.
GTrust — Google Reliability Score
What It Is
A Bayesian-adjusted reliability score (0-10) derived from Google Maps ratings and review counts. It corrects for shops with very few reviews and provides a reliability check — not a quality judgment.
How It Affects Ranking
GTrust only acts as a tiebreaker when two restaurants are within 4 FitScore points of each other. It never overrides a significant FitScore gap. Think of it as a secondary signal, not a primary one.
The Formula
Where R = Google rating, v = review count, C = global mean rating (prior), and m = minimum reviews for reliability. This is a standard Bayesian average that pulls low-review-count shops toward the mean.
What We Don't Do
Rank "The Best"
We help you find sushi that fits your trip, not a single #1.
Inflate for Advertisers
Zero conflict of interest. Affiliate links never touch scores.
Fake Visits
Unvisited shops are always labeled "Database Recommended."