Tank Royale Rumble

Writing a bot is one thing. Finding out how good it really is takes a lot of battles against a lot of opponents.

Tank Royale Rumble is the community-run ranked competition for Robocode Tank Royale. It follows the tradition of RoboRumble and LiteRumble from classic Robocode: submit your bot, let it fight across the shared ladder, study the results, and come back with a better version. If you are serious about Robocoding, this is where you put your code to the test.

You do not steer a tank during a Rumble battle. You write the program that controls it. Rumble clients run the battles automatically and publish the results, so rankings come from the same bot code fighting many different opponents.

The Rumble loop

1. Build a bot with one of the official Tank Royale Bot APIs.
2. Practice locally in the Tank Royale GUI until your bot behaves the way you want.
3. Submit its source code to the reviewed Rumble bot catalog.
4. Community members run Rumble clients on their own computers. The clients choose useful matchups, run ranked battles, and submit the results.
5. The dashboard combines the accepted results into rankings.
6. Learn from the results, improve your bot, and submit a new version.

The cycle never really ends. A bot that dominates today may meet a smarter opponent tomorrow.

Choose how you want to participate

I want to...	Start here
Enter my bot in the rankings	Submit a bot
Donate computer time and run ranked battles	Run a Rumble client
See which bots are winning	Open the live dashboard
Help review submissions and keep the competition fair	Moderate the Rumble

You can submit a bot without running a client, and you can run a client without owning a bot. Many competitors do both because the client gives matchups involving their own bots priority when more samples are needed.

Ranked game types

Each game type has its own leaderboard.

Game type	What fights	Rounds	Battlefield
1v1	Two individual bots	35	800 x 600
TwinDuel	Two teams with two bots on each team	75	800 x 800
Melee	Ten individual bots at once	35	1000 x 1000

TwinDuel is the classic twin-team format, not a general team category. Mini, micro, nano, and other code-size classes from classic Robocode are not part of the Tank Royale Rumble because source-size limits do not compare cleanly across Java, C#, Python, and TypeScript.

How the system works

The Rumble has three public GitHub repositories:

• rumble-bots is the reviewed, source-only bot catalog.
• rumble-client runs battles on contributors' computers.
• rumble-data accepts results, calculates the rankings, and publishes the dashboard.

There is no central battle server. Clients download the same pinned bot catalog and game settings, run battles locally, and send completed results through GitHub issues. Automation validates each result before it becomes part of the ranking data. The raw accepted results remain in Git, so the leaderboard can be rebuilt and checked by anyone.

Catalog

The catalog is the list of bots and TwinDuel teams allowed in ranked battles. Every entry has an owner, a version, and a hash of its reviewed source. A new source version is a new ranked identity; only the latest active version appears on the current leaderboard.

Matchups and samples

A matchup is a set of opponents that fought each other. One completed battle is one sample of that matchup. The client prefers new and under-sampled matchups, which helps new bots receive a meaningful rank and keeps older rankings fresh.

Rankings and APS

The main ranking number is APS, or Average Percentage Score. For each matchup, Rumble calculates the percentage of the total score earned by a bot or team. It averages repeated battles of that matchup, then averages across all of the entry's matchups. Higher APS is better.

A new bot or version starts with no battle samples, so its first position can move sharply. The ranking settles as it fights more opponents. When a game-observable engine change starts a new behavior version, the current leaderboard uses only results from that new epoch. Earlier results stay in the public data history, but they are not mixed with battles played under different rules.

When rankings update

An incoming result submission starts the ingestion workflow as soon as GitHub applies its result-submission label. A scheduled sweep also runs at 17 and 47 minutes past every UTC hour in case an event was delayed. Accepted results regenerate the leaderboard, and GitHub Pages publishes the changed dashboard after the data commit. In normal operation, a result appears within minutes; the scheduled sweep is the fallback, not a guaranteed deadline.

The bot catalog synchronizes at 23 minutes past every UTC hour. A newly merged bot normally reaches the dashboard after that synchronization, then waits for clients to produce its first ranked battles.

Current availability

Bot submission, ranked battle execution, result ingestion, and the dashboard work end to end. The Rumble client does not yet have a published production container image, so battle contributors currently build it from source and use the native command line. The client guide shows the exact supported path and calls out the parts of the Docker workflow that are still under development.