Frequently Asked Questions

A quantitative analysis platform focused on sports. We publish probabilities generated by mathematical models, along with the historical record of their performance. We are not a bookmaker and we don't sell picks.

We currently publish probabilities for football matches (starting with La Liga). We cover markets like final result (1X2), total goals (Over/Under), both teams to score, and shots on target per player. We plan to expand to more sports progressively.

We apply mathematical models calibrated on historical data from thousands of matches. The system processes the information and generates a quantitative estimate for each market. We don't reveal specific details of our methodology due to intellectual property.

It means that, according to the model, that outcome has an estimated 70% probability of occurring. It's not a certainty. A model with 70% estimated accuracy will be wrong 30% of the time. That's the statistical reality.

No. We are not a bookmaker and we don't give betting recommendations. We publish quantitative probabilities so people can interpret sports information objectively and transparently.

No. We don't sell picks or betting advice. Our product is the published probabilities together with the historical record of their performance.

We publish the full track record in our Track Record section. We show metrics like Brier Score and hit rate by confidence range, always indicating how many observations back each metric. Historical results don't guarantee future results.

We use standard quantitative metrics like Brier Score (measures calibration) and hit rate by probability range. We compare every published prediction against the actual outcome and accumulate evidence over time.

It stays recorded in the history. We don't delete failed predictions. We don't rewrite the track record. Every published prediction is kept as part of the accumulated evidence, correct or incorrect.

Yes. All published predictions remain available in the history. Transparency about our real performance is a central principle of the platform.

Models evolve when we accumulate enough historical evidence to justify a new version. Each version is identified for traceability. A model is not updated on intuition, but on accumulated evidence.

Each new iteration of the system has a version (for example v7.4). We publish which version generated each prediction to maintain full traceability. When we launch a new version, the previous one is kept historically recorded.

During our launch phase, access is 100% free. We only ask for registration with email and password. No card required, no cost.

Click 'Sign Up' from any page. You just need an email and to create a password. Access to all probabilities is unlocked immediately.

You can write to us from our Contact section or send suggestions directly from the Feedback section. We read every message.