Camel Up Cup — Barron Wasteland

Barron Wasteland

1 min read Original article ↗

The original 2018 bring-your-own-bot tournament never really ended.
Write a Python bot that plays Camel Up, upload it below, and it'll play a batch of games against every bot on this board.

Champion

tb-PortfolioPam

Elo 1558.4 ·

33.7

% win rate

Standings

RankBotUploadedElo ▼WinsGamesWin %CoinsMax ms
🥇

tb-PortfolioPam

by Tyler Barronuploaded

Jul 5, 20261558.4225226689433.7%25.63598.2
🥈

FH-FH

by Freduploaded

Jul 10, 20261557.9223116655533.5%26.73593.7
🥉

CurriedCamel13

by Matthewuploaded

Jul 4, 20261543.8210596668631.6%25.62597.1
4

curried_camel

by Matthewuploaded

Jul 3, 20261543.2208936632231.5%25.61262.6
5

tb-PairTestedPeter

by Tyler Barronuploaded

Jul 3, 20261528.2197226671629.6%24.43659.5
6

tb-FabelFelix

by Tyler Barron

Claude Fable 5

· 2026

—1514.8185536659527.9%242956
7

FH_ActualReplicant

by Freduploaded

Jul 4, 20261513.0183756682127.5%26.43304.9
8

FH_Thirstybot1

by Freduploaded

Jul 2, 20261478.7155606637823.4%24.53328.5
9

tb-ClaudeCamel

by Tyler Barron

Claude Sonnet

· 2026

—1447.6134376690120.1%22.92470.4
10

tb-OpusOmul

by Tyler Barron

Claude Opus

· 2026

—1388.899606662114.9%21.41688.7
11

tb-GeminiGerry

by Tyler Barron

Gemini

· 2026

—1377.493036696213.9%22.45008.8
12

tb-HandcodedHenry

by Tyler Barron

Hand-coded; won the original Cup as Sir_Humpfree_Bogart

· 2018

—1356.283076654912.5%20.81437.5

200000 games per tournament · random 4-player seatings · ties split wins · Elo counts only wins, adjusted for opponent strength · updated 8/29/2026, 10:58:36 AM

Enter your bot

One .py file, one class with a move(player, gamestate) method returning [0] (roll), [1, trap_type, pos], [2, camel] (round bet), [3, camel] (game winner) or [4, camel] (game loser). Standard-library compute only (math, random, itertools, …) — no network, files, or threads. Max 5 seconds per move: slower bots are disqualified, crashing moves become dice rolls. Grab the engine and full rules from the GitHub repo to test locally first.