RankquantRQ

Browse the complete cruise ships index →

Live · 298 cruises

Cruises

Every ship with at least 3 z-qualifying Cruise Critic reviewers, ranked on the mean per-reviewer z-score and dual-scored (z-normalized + raw average).

298
Cruise ships
171.7K
Cruise Critic reviewers
214.1K
Reviews ingested
21.2K
Z-qualified pool

Top of the rankings

Highest z-normalized percentile

Browse all 298 →
Borealis
Fred. Olsen100
Fred. Olsen · n=3
Amadolce
AmaWaterways99.7
AmaWaterways · n=4
Silver Ray
Silversea99.3
Silversea · n=3
Viking Egil
Viking99
Viking · n=4
Amadara
AmaWaterways98.7
AmaWaterways · n=4
Amacerto
AmaWaterways98.3
AmaWaterways · n=3
Bolette
Fred. Olsen98
Fred. Olsen · n=4
Viking Eir
Viking97.6gap
Viking · n=5
Silver Muse
Silversea97.3
Silversea · n=4
Ms Antares
Other97
Other · n=3
Viking Hlin
Viking96.6
Viking · n=4
Star Breeze
Star Clippers96.3
Star Clippers · n=13

Where the methodology pays off

Biggest inflation gaps

Ships where z-normalized and raw-average percentiles diverge most.

Azamara Onward
Azamara82.5+56
Azamara · z=82.5 raw=26.3
Carnival Radiance
Carnival14.5-50
Carnival · z=14.5 raw=64
Queen Elizabeth
Cunard79.8+46
Cunard · z=79.8 raw=34
Carnival Jubilee
Carnival35-40
Carnival · z=35 raw=75.4
Seven Seas Navigator
Regent Seven Seas60.6+39
Regent Seven Seas · z=60.6 raw=21.5
Aurora
P&O33.7-37
P&O · z=33.7 raw=71

Score distribution

0255075100

z-normalized · all 298 cruises

Source coverage at launch

One source at launch — Cruise Critic — with every qualifying reviewer counted once. Cruise Critic is the deepest crowd-review pool in the category (Trip Advisor-owned since 2007) and has the widest ship coverage. Roadmap: TripAdvisor cruise reviews, Cruise Hive editorial, Berlitz Guide.

Cruise Critic is the live source; the rest are roadmapped. None of them carries a weight — when a new source comes online, its reviewers are z-normalized and counted exactly like Cruise Critic reviewers.
Cruise Critic (live)171.7K reviewers, 214.1K reviews ingested. 21.2K qualified for z-normalization (DB1, stddev>0). Catalog filter: ≥3 z-normalizable reviewers per ship.
TripAdvisor cruises (planned)Wider crowd pool, and more gaming pressure to screen for.
Cruise Hive (planned)Editorial, news-driven coverage of new ships.
Berlitz Cruising (planned)Annual professional rating, scored from in-person sailings.
Cruise Critic is the live source; the rest are roadmapped. None of them carries a weight — when a new source comes online, its reviewers are z-normalized and counted exactly like Cruise Critic reviewers.

Global and cohort percentiles

Every ship gets two percentiles: global (its mean reviewer z-score ranked against all ships in our database) and cohort (the same mean, re-ranked within its cruise line). Both report the measurement as made — the sample-size correction lives in the AI-adjusted figure, not in these two.

Frequently asked questions

Why normalize cruise reviews?+
Cruise Critic's long-run rating average is 4.07 / 5 — virtually every ship clusters between 3.8 and 4.6. The crowd grade-inflates and rates regression-to-the-mean. Z-normalization re-centers each reviewer on their own scale, so a 4.6 from a tough rater carries more signal than a 5.0 from someone who 5-stars everything.
How is the line cohort defined?+
When the source DB doesn't list a line, we infer it from the ship name (Disney → Disney Cruise Line, "Of the Seas" → Royal Caribbean, etc.) for ~93% coverage. The remaining ships fall under "Other" until editorial cleanup.
Why only 298 ships when there are 600+ cruise ships?+
The catalog filter requires ≥3 reviewers with varied rating habits (stddev > 0). Ships in the source DB that carry no reviews, or that carry reviews but miss the qualifying-reviewer bar, are held out of the ranking. Thinly-reviewed ships (river cruisers, niche luxury yachts) are added as their review counts grow.
What about review-bombing or paid placements?+
Thin samples and coordinated pushes are handled by the AI-adjusted percentile, not by the headline. The global and in-cohort percentiles rank the mean reviewer z-score as measured; the AI-adjusted percentile ranks that same mean after pulling it toward the corpus average in proportion to sample size — a ship keeps the fraction n/(n + 53) of its measured distance from that average, so ten coordinated accounts surrender most of their effect while a ship with 200 calibrated reviewers keeps nearly all of its own. Per-reviewer normalization also removes the accounts that rate everything identically: a reviewer with no spread in their own ratings has nothing to normalize against and never enters the z-normalized track at all.

Status: cruises is live with 298 dual-scored ships across 29 cruise lines. Browse the catalog at /cruises/catalog, or view the full methodology for how each score is computed.