This version uses a direct peer-to-peer connection. No account or game server is required. The connection codes can be long because they contain the temporary connection information.
Create an invitation, send it to Player B, then paste their answer code back here.
Paste Player A's invitation, create an answer, then send the answer code back.
Box A allowed investigation. Each test added information, and the player could revise a conclusion before answering. Box B supplied one confident statement and offered no way to check it. Because the statement never changed when reality disagreed, Box B failed.
The game is a simplified comparison between an evidence-based method and a fixed-authority method. Science is powerful because its explanations are provisional: they can be tested, challenged, refined, and replaced when better evidence appears.
Both players were asked the same basic question: what is inside the box? The difference was not intelligence, effort, or honesty. The difference was the method available to each player.
Player A could interact with the box. Motion suggested shape. Light revealed color. A probe suggested whether the object was hard or soft. No single experiment supplied the whole answer, but several observations together produced a model that could be checked against reality.
Player B received a declaration. The declaration sounded certain, but certainty is not evidence. When the declaration was wrong, Box B provided no procedure for detecting the mistake, updating the claim, or learning from the failure.
Science is not a list of final facts. It is a method for building increasingly reliable explanations. Observations are collected, hypotheses are proposed, predictions are tested, results are shared, and conclusions remain open to revision. A scientific explanation earns confidence by surviving attempts to disprove it, not by being protected from questions.
This is why scientific conclusions can change. Revision is sometimes presented as a weakness, but it is one of the method's central strengths. A system that can admit error can become less wrong over time. New instruments, new experiments, and better data can expose limitations in older models.
The game uses Box B to represent religious claims when they are treated as fixed answers about physical reality that must remain true regardless of observation. Under that condition, the conclusion comes first and evidence is accepted, rejected, or reinterpreted according to whether it supports the conclusion.
Religious traditions are diverse, and many religious people accept scientific findings or understand religion as addressing meaning, ethics, community, ritual, or spiritual experience rather than laboratory questions. The game does not claim that every believer thinks in the same way. The demonstration here is that authority and faith are not reliable substitutes for repeatable evidence, especially when the question concerns the measurable contents of our physical reality.
A statement can be sincere, ancient, popular, comforting, or confidently delivered and still be wrong. Good methods give us a way to discover that. They also tell us how much confidence is justified by the available evidence.
Box A did not begin with perfect knowledge. A player could initially skip an experiment, misunderstand a result, or choose the wrong answer. The important difference is that the available tests allow the model to be checked and revised. That also reflects real science: people can make mistakes, but the process is inspectable and repeatable. Another person can perform the same tests, compare results, and identify the error.
The lesson is not simply “science is always right.” The lesson is that science contains a mechanism for becoming more accurate, while a claim that cannot be questioned has no dependable mechanism for correcting itself.
In Side by Side mode, both boxes are visible at once. In Hot Seat mode, pass the computer when prompted. In Online mode, Player A hosts the connection and Player B joins from another browser.
Make sure the next player is ready before continuing.