Privacy Policy

The short version: we collect the minimum data needed to run the experiment. We use it to score predictions, show aggregate results, improve the project, and understand how confidence behaves before reality grades it.

What we collect

When you make a prediction, we may collect:

• the match you predicted
• the outcome you selected
• your confidence level
• the time the prediction was submitted
• whether the prediction came from the main site or embedded version
• an anonymous participant ID stored in your browser
• your score after the match is final, including Brier Score and Calibration Score

The anonymous participant ID helps us connect your predictions across matches without requiring an account.

If you use the feedback form, we may also collect:

• your message
• feedback type
• page path
• optional display name
• optional email address, only if you choose to provide one

What we do not collect by default

• We do not require an account.
• We do not require your real name.
• We do not require your email to make predictions.
• We do not collect payment information.
• We do not sell your data.
• We do not use your predictions for advertising.

How we use the data

We use the data to:

• save and score predictions
• calculate Brier Scores and Calibration Scores
• show aggregate public results
• understand where people were confident, cautious, right, or wrong
• improve the experiment
• write public analysis about calibration and confidence

Public reporting is aggregate. For example, we may publish that a certain percentage of participants picked a team, or that the crowd was overconfident on a match. We do not publish raw participant IDs.

Local browser storage

The site uses local browser storage to remember your predictions and anonymous participant ID.

You can clear this data by clearing your browser storage for this site, or by using the controls below.

Public experiment data

If you submit a prediction to the public experiment, it is stored in our database so it can be included in aggregate results.

This data is pseudonymous, not account-based. That means we do not ask who you are, but your submissions may still be linked together by the anonymous participant ID stored in your browser.

Feedback

If you send feedback, we use it to improve the experiment. If you include your email address, we may use it to reply to you. Email is optional.

Research and summaries

Human Actually may analyze aggregate results from the experiment and publish summaries, charts, or essays about what the data shows.

These summaries are about patterns, not individual people.

Service providers

We use a small number of standard service providers to host the site, store experiment data, fetch official match results, and send the occasional email. They process data on our behalf to operate the experiment. They do not receive prediction data for any other purpose, and they are not used to sell your data.

Data retention

We keep raw experiment submissions for as long as needed to run the World Cup experiment, score predictions, debug issues, and complete analysis.

After the experiment and analysis period, we may delete raw submissions or keep only aggregate, de-identified results.

Children

The Calibration Cup is not designed for children under 13. Please do not submit information if you are under 13.

Contact

For privacy questions or deletion requests, contact privacy@humanactually.com.

Please include enough detail for us to understand the request. Because the experiment does not use accounts, we may not always be able to identify a specific submission unless you provide the anonymous participant ID from your browser. You can copy your ID from the controls above.