The TWT Ships data tool

World of Warships ship stats, with the context to use them.

A useful shortlist starts with a question: which ships have the range, concealment or handling you want? Our searchable ship database brings Wargaming's published base values into one place so you can investigate that question.

Explore the ship database →

What does the ship database include?

Our scheduled catalogue imports the complete default-profile object and published research collections from the official Ships endpoint. A saved copy serves complete profiles without a new Wargaming request for every visitor. The source-field section lets you inspect published values beyond the main charts.

AreaAvailable when publishedLimit to keep in mind
Survivability and handlingHit points, armour ranges, speed, turning circle and rudder shiftArmour ranges are not a 3D armour model or a complete plate layout.
Weapons and detectionMain and secondary batteries, shells, torpedoes, AA, depth charges and detection rangesMissing values are unknown. Range alone does not establish a weapon's effectiveness.
Aircraft and submarinesPublished aircraft profiles, battery, underwater mobility and sonarThe API uses legacy aircraft categories and may omit modern systems.
Research and equipmentModule trees, compatible upgrade IDs and next-ship research linksA list of compatible modules does not apply their effects to the displayed profile.

This is not every value in the game client. Consumable loadouts, alternate firing mechanics, individual armour plates and newer aircraft systems are not comprehensively described by this endpoint. An empty section or em dash means the source did not provide a usable value; it does not prove the ship lacks that system. Wargaming can also change the API independently of the game.

Source: Wargaming's official Ships reference. TWT Ships is an independent fan tool and is not endorsed by Wargaming.

Why do these stats differ from my ship in Port?

Your ship may have researched modules, commander skills, upgrades, signals or other modifiers applied. The database displays the default configuration returned by the API. It does not calculate your selected build or simulate battle conditions.

When a number differs, first compare the module configuration, then the modifiers, then the data date. The commander planner helps organise skill points; it does not recalculate this ship profile. A mismatch after those checks may be an API gap or stale upstream record. Report a data issue with the ship name, field and source so it can be investigated.

Where can I find World of Warships ship win rates?

You do not need to wait for TWT Ships to build its own sample to research ship performance. These independent tools publish their own ship statistics:

Those links open external services. Their coverage and methodology belong to each provider; their numbers are not imported into our database. Match the server, battle mode, time period and player group before comparing results between tools.

A win rate needs a denominator

For a specified sample, ship win rate is wins divided by battles, multiplied by 100. If one profile has 6 wins in 10 battles and another has 40 wins in 100 battles, the combined result is 46 out of 110: about 41.8%, not the 50% average of their individual percentages.

A high percentage from a few battles is less informative than a large, clearly defined sample. Lifetime results can mix several versions of a ship. Skilled players selecting a particular ship can also affect its observed performance. A win rate describes results in that sample; it does not isolate ship strength.

Will searching player names build a server-wide win rate?

No. Wargaming's Ships statistics method returns statistics for specified player accounts. Visitors searching names are a self-selected group, and private profiles can be unavailable. More searches would not turn that group into a representative server population.

TWT Ships does not currently publish an aggregate ship win-rate dataset. Any future contribution programme must clearly explain participation, minimum sample sizes, server coverage and retention. It would be labelled as a community sample, with its own limitations. Until then, missing win rates stay missing.