Clever Chicken

Meet Clever Chicken, the independent pairing engine for USCF tournaments, now in public beta.

CLEVER CHICKEN / PUBLIC BETA

Clever Chicken

An independent pairing engine for USCF tournaments, built for decisions that can be reproduced, examined and explained.

Smart pairings. Verifiable decisions.

Create a test tournament → See how we test

Beta notice: use Clever Chicken for testing and independently verify its pairings before relying on them for official play.

INDEPENDENT ENGINE

Rigorous does not have to sound serious.

Clever Chicken is named after the favourite plush toy of the creator's son: personal, memorable and a reminder that careful engineering can still have character.

WHY WE ARE BUILDING IT

A pairing engine for USCF tournament workflows.

Clever Chicken is an independent project created to help clubs and Tournament Directors run USCF events with a pairing process that is deterministic, inspectable and supported by reproducible evidence.

The engine is being developed as part of ChessPairings.org, a free platform built without commercial intent to support organisers and grow chess communities.

TESTING WITH EVIDENCE

Differences are evidence, not verdicts.

Swiss pairings can admit more than one legitimate solution. That is why we do not label every different output as a defect: we classify it against the rules, preserve the inputs and investigate the reason.

9,795

Reproducible cases

Common benchmark scenarios can be run again with the same tournament state and the same inputs.

0

Hash mismatches

The current corpus preserved input identity across the one-to-one comparisons.

0

Strong defect candidates

No divergence in the current corpus crossed the evidence threshold for assigning a confirmed defect to Clever Chicken.

WHAT THE COMPARISON SHOWED

One corpus, three observable patterns.

Across 9,795 identical inputs, the two independent reference outputs did not always make the same legitimate choice.

The reference outputs chose different opponents8,776 89.6%
Same opponents, colours and board order538 5.5%
Same opponents, but colours or order differed481 4.9%

This distribution helps classify the corpus; by itself, it does not determine whether a Clever Chicken pairing is right or wrong.

01

Reproduce

Keep tournament state, player data and configuration available for an exact rerun.

02

Classify

Evaluate a difference against the pairing rules instead of treating a distinct solution as automatically wrong.

03

Retain

Turn meaningful findings into regression cases so later changes can be checked against earlier evidence.

These automated results are encouraging, not a claim of perfection or certification. The corpus continues to grow and every new real-world case can make it stronger.

ENGINEERING PRINCIPLES

Built to make the decision traceable.

Deterministic

The same verified inputs should produce the same decision.

Rule-aware

Pairing choices are evaluated in the context of the constraints that produced them.

Auditable

Testing retains enough evidence to reproduce and investigate a result.

Regression-driven

Confirmed edge cases become permanent tests rather than one-off fixes.

EXPERIENCED TDS WANTED

Automated tests cannot replace tournament experience.

Create a sample USCF tournament, review the generated pairings and tell us what you see. The test area lets you leave a rating and notes about the engine's decision.

Help test Clever Chicken →

Clever Chicken is an independent project and is not affiliated with or endorsed by US Chess. It is currently in public beta.