Frequently asked questions

Altora Analytics is a sports analytics platform publishing calibrated win probabilities across eight sports (football, tennis, darts, snooker, rugby, esports, basketball and handball), alongside systematic trading research. Every prediction is published before the event, every result is settled publicly, and nothing is ever deleted. The records are the product.

No. We publish analysis: model probabilities, not instructions to bet. Some record pages also track fixed £10 level-stake performance so visitors can judge the models transparently, but nothing here is financial advice. If you bet, that's your decision and your responsibility. 18+ only, and please read BeGambleAware.org.

Most sports use an adaptive rating system combined with live market information and machine-learning calibration, validated on tens of thousands of matches the models never trained on. Football adds further features such as expected-goals metrics. The full technical write-up is on the methodology page, and the complete build method is in our handbooks.

Never. Every record page shows wins and losses, and offers the complete history as a CSV download: date, selection, odds, result. A deleted loser would leave a hole in the CSV and a mismatch with the posts we've already published. That verifiability is the whole point of the site.

Because that would make it impossible to verify when they were made. Every prediction is published before the event starts and remains online afterwards, so anyone can check both the timing and the result. Hidden-then-revealed picks are how this industry cheats; public-before-play is how you prove you didn't.

Because it's true. Betting markets at closing are extraordinarily efficient. Matching them is genuinely hard, and “we smash the bookies” is one of the oldest claims in this industry. Our models are calibrated (when we say 70%, it happens about 70% of the time) and statistically level with the market on most sports. We show the comparison either way.

We make no such promise, and you should distrust anyone who does. Bookmaker margins mean most bettors lose over time. What we offer is honest, calibrated information and fully transparent records. What you do with it is up to you.

Because predictions without evidence are worthless. Every model has good weeks and bad weeks. We'd rather publish every result permanently and let people judge the models over thousands of predictions than advertise isolated winning streaks.

Two one-off PDF guides that teach the complete method behind this site: one for building a sports prediction model, one for building an automated FX trading bot. Designed for beginners, with complete Python code provided throughout. £29.99 each (launch price, rising end of August), instant download. They're educational: they promise knowledge, not profits.

We believe trust has to be earned before anything is monetised. Every sport begins with a public record so anyone can evaluate the model. Premium features add earlier access and additional markets, but the public records remain available to everyone, permanently.

Coming with the football season: earlier delivery of the daily picks and extra markets, once the public record has proven the season's model. The free records stay free. That's permanent.

Email altoraanalytics@gmail.com and we'll sort it out. Delivery is instant via Whop; if something's wrong with your copy, that's on us to fix.

We use established third-party data providers including:
• API-Sports (football, rugby union, basketball, handball)
• Betfair Exchange (live market prices and results)
• football-data.co.uk (historical football odds)
• tennis-data.co.uk (historical tennis results and odds)
• OpenDota (Dota 2)
Individual record pages credit the data sources used for that sport.

The raw feeds aren't ours to sell. They belong to the sources above. Our own model outputs and published records are freely downloadable from every record page as CSV.