How OddsAura actually works.

1. Editorial process for match pages

Every match page on this site goes through the same five-stage pipeline. One of the three of us is the lead writer; the other two review before publication. There is no AI-generated text in our editorial copy — the AI Match Analysis tool is a separate, clearly-marked interactive feature, described in section 6 below.

Stage 1 — Story identification

Before we write anything technical, we identify the editorial story of the match. Why this fixture matters. What each manager will try to do. Where the personal or historical stakes are. We ask: if we removed odds and lineups entirely, what would still be interesting to read? If we cannot answer that question, we delay the page. We have never published a match preview that started with "Both sides will be looking for three points."

Stage 2 — Squad and lineup research

Predicted XIs are cross-referenced against three independent sources: the official federation announcement, the most reliable national-team beat reporter in each country, and the manager's last-three-matches starting eleven. Where a manager has not announced his XI, we say so explicitly in the page text and provide our best prediction. Injury and suspension notes come from official club bulletins, not from aggregator sites.

Stage 3 — Odds sourcing

Match odds across 1X2, total goals, BTTS and Asian handicap markets are collected from our three partner books (Mostbet, Fonbet, 1xBet) plus 5–6 independent comparison sources. Where prices diverge meaningfully across the wider market, we note the divergence in the editorial copy. We refresh odds at least once before publication and rerun the check 24–48 hours before kickoff for any timely match.

Stage 4 — Value-bet picks

The inline "value bet" picks on each match page are Mike's personal calls. Each pick is based on a specific analytical reason that is written out in the editorial copy — never a one-line recommendation without context. See section 3 for the detailed criteria.

Stage 5 — Cross-language review

Vasco coordinates the German and Spanish editions. The three editions are not machine-translated. Each one is written, edited and reviewed by a native speaker. Editorial decisions (which pick to lead with, which narrative angle to centre) are identical across all three languages; the phrasing is adapted to each.

2. How we source and compare odds

Every match page lists six core betting markets: 1X2 (home / draw / away), total goals over 2.5, both teams to score yes, and a contextually-chosen Asian handicap line. The "best price" highlighted in lime on each row is the highest decimal odds among our three partner books for that specific market at the time of publication.

Why three partner books

Our commercial partners are Mostbet, Fonbet and 1xBet. We chose three because that is the minimum useful set for a comparison site — fewer would force readers to a single operator, more would clutter mobile layouts without adding meaningful price diversity. All three accept customers from most European and Latin American jurisdictions and offer the full football match-market depth this tournament requires.

What "best price" actually means

It is the highest decimal odds we observed at our last scheduled refresh. It is not necessarily the highest price in the wider market. If you have access to other books with sharper lines for a specific market, you should bet there. We are not pretending that three operators represent the entire universe of available prices. We are saying: of the three options we link to, this is the one currently paying you the most for that market.

How often we refresh

Match pages are refreshed when we re-edit them. Odds-only refreshes happen on a 24–48-hour cadence in the week leading up to a match. We do not run live odds. If you are seeing this page within 6 hours of kickoff and the listed odds look stale, the operator's own page is the authoritative source — every "Bet now" CTA links there.

What we do with significant odds movements

If one of our partner books moves meaningfully between editorial publication and kickoff (typically 10%+ on the lead market), we add a note to the editorial copy explaining what the market is reacting to. We do not silently update odds without acknowledging the change.

3. How we choose value bets

The lime-highlighted "value bet" picks inside the editorial copy on each match page are Mike's personal calls. They follow a fixed analytical process and a fixed selection of market types. They are not certainties. We lose plenty.

The shape of a pick

A pick on this site has three required elements: a specific market (e.g. "Both teams to score yes at 1.95"), a specific operator (always one of our three partner books), and a specific analytical reason written out in the surrounding prose. We do not publish picks without all three. We do not publish "lock of the day" picks. We do not publish guaranteed-win picks. Those things do not exist in this market.

What we look for

• Identifiable mispricing. A market where the implied probability we'd assign differs from the bookmaker's by more than the typical bookmaker margin (usually 5–7%). If we'd price the market 50/50 and the book is offering 2.10, we'd consider that pick. If the book is offering 1.95, we'd not.
• Structural reasons, not gut feel. Every pick has to be explainable in two sentences. "Both teams scored in five of their last six meetings, both defences have rotated, the game is at altitude" is structural. "I think this side is due a win" is not. We do not publish the second kind.
• Defined exit conditions. A 90-minute Asian handicap pick has clear outcomes. A "Player X anytime scorer" pick does too. We do not pick markets with vague resolution criteria.

Market types we use

The picks you'll see on the site are concentrated in five market types, in roughly this order of frequency: Asian handicap (quarter, half and full lines), goals totals (over/under 2.5 or 3.5), both teams to score, anytime goalscorer, and double chance. We use these because they are the markets where our editorial reasoning is most testable and where book pricing is most consistent across operators.

Market types we avoid

We do not publish picks on accumulators, parlays, first-goalscorer, half-time/full-time, exact correct score, or any market with margins above 12%. Those markets are designed to look attractive — high payouts, easy to imagine — but the book margins make them structurally unprofitable for the bettor. If you see a tipping site leading with exact-score picks, that is a tell.

4. Track record & honest accounting

We publish a running track record of every value pick we have made on this site. The methodology, the result, the operator, and the closing price are all recorded. We do not retroactively delete losing picks. We do not selectively highlight winners in marketing copy.

What the track record measures

Profit and loss against the price we recommended at publication time, assuming a flat 1-unit stake. We use closing-line value (CLV) as the secondary measure — a pick that beats the closing line is, in the long run, more likely to be profitable regardless of whether the individual bet won. Honest CLV is the single best indicator of whether a tipster knows what they're doing.

What we are honest about

• We are aiming for a small positive expected value across the tournament — not a guaranteed profit on any individual match. Over a tournament of 104 matches with 200+ picks, our target is +3% to +6% ROI. Anyone claiming to consistently hit +20% across an event of this size is either lying or has been lucky for a short period.
• We will publish picks that lose by clear margins. That is what happens. The honest comparison is whether our picks are systematically beating closing lines, not whether every individual bet won.
• The track record page is updated within 24 hours of each match's final whistle. If a pick is rendered void by a postponement or rescheduling, we note that and remove it from the count.

Track record page launches with the tournament. Live URL: /track-record/.

5. How we rank bookmakers

Where you see bookmaker comparisons or "best for X" lists on this site, they follow a fixed scoring framework that we share in full here. The framework was designed by Mike based on twelve years of working inside operators and looking at them from the outside. Each criterion is scored from 0 to 10, and the overall ranking is a weighted average across them.

What we explicitly do not weight

• Commission rate paid to OddsAura. The bookmakers we partner with do not all pay us the same commission. Our rankings do not reflect that.
• Bonus headline number alone. A "300% welcome bonus" is meaningless without the wagering terms attached. We score net expected value, not advertised maximum.
• Operator brand recognition. A well-known operator with poor payout history scores lower than a smaller, lesser-known operator with clean books.

6. What the AI Match Analysis actually does

Every match page on this site has an "AI Match Analysis" card you can expand. It is currently a beta feature, clearly marked as such, and we want to be transparent about what it is and isn't.

What it does

When you click "Generate analysis," the card runs a Monte Carlo simulation of the match: a probabilistic model that plays the fixture 10,000 times, distributes the outcomes, and identifies the most-common scorelines, the implied 1X2 probabilities, and where the market may be mispricing the line. The output is a paragraph of natural-language summary plus a value-bet flag if the simulation suggests one.

The model uses publicly-available match data: recent results, scoring rates, expected goals (xG) per match, home/away splits, lineup strength scores derived from minutes played at top-five-league level, and a few tournament-specific adjustments (altitude, heat, travel distance, rest days).

What it doesn't do

It does not write the editorial. The prose paragraphs above and below the AI card are 100% human-written. The AI tool sits between them as a self-contained interactive element.

It does not have insider information. It cannot predict the future. It does not account for things we humans can see — squad morale, manager press-conference signals, last-minute training-ground intelligence — that frequently move match outcomes. Treat its output as one more data point, not a prediction.

It does not replace our value-bet picks. The picks inside the editorial copy are Mike's calls, supported by his analytical reasoning. If the AI flags a different pick, the editorial pick wins. We say that explicitly.

Why we built it

Honestly, because it was an interesting thing to build and because readers seem to like having a quantitative anchor on top of our editorial prose. It is currently in beta and we are still tuning it. We will retire it if it stops adding value.

7. Corrections, updates & reader feedback

We make mistakes. When we do, we fix them visibly. Our policy on corrections:

• Factual error (wrong score, wrong date, wrong squad member, wrong stat): corrected in the article body and a "Corrected on [date]" note added at the bottom of the page if the change is material.
• Editorial judgment (our prediction was wrong, our reasoning was poor): we don't retroactively edit. The original article stays. If we wrote something we later disagree with, we publish a follow-up or update note rather than rewriting history.
• Odds movement: handled per section 2 above.
• Reader-flagged corrections: email [email protected]. We aim to respond within 48 hours and act within 5 working days if the correction is warranted.

8. Sources we use

We try to source primary where possible. The list below covers the bulk of our research diet — it is not exhaustive but it is honest.

Match data & statistics

• Opta (via FotMob and Sofascore) — match statistics, xG, lineups
• Transfermarkt — squad valuations, transfer history, contract data
• WhoScored — match-by-match player ratings
• FBref / StatsBomb-derived data — advanced metrics where available
• Official FIFA, UEFA, CONMEBOL, CONCACAF, CAF, AFC, OFC sources for federation data and qualifying records

Editorial & reporting

• National federation websites and press releases for squad announcements, injury bulletins, manager statements
• The most reliable national-team beat reporters for each country (we won't name them by name — these are individuals whose access we don't want to compromise — but they are who you'd expect)
• The Athletic, The Guardian, BBC, ESPN, Bild, Marca, AS, La Gazzetta dello Sport, Lance!, Globo Esporte, and other established outlets for tactical context and breaking news

Odds & betting markets

• Our three partner books for primary partner-market odds
• Pinnacle (as a sharp-line reference, not a partner), Betfair Exchange (for market-implied probability)
• 5+ independent comparison sources as price-check secondary references

Responsible gambling & regulation

• UK Gambling Commission, Spain's DGOJ, Portugal's SRIJ, Italy's ADM, Germany's GGL, France's ANJ for regulatory data
• GamCare, NCPGambling, BeGambleAware, Gambling Therapy for problem-gambling support content

9. Editorial independence

Bookmaker partners do not see our editorial copy before publication. Our partners cannot request changes to our content. They cannot pay for a higher ranking inside our bookmaker comparison framework. They cannot suggest topics for us to cover or which picks to make.

Our partnership agreements include a one-line standard term that confirms editorial independence; the rest of the contract is the same revenue-share boilerplate every affiliate program uses. If any of our partners were to attempt to influence editorial content — which to date none has — we would terminate the partnership and write about it.

We do not accept gifts, hospitality, paid trips, sponsored events, or free accounts beyond what is needed for testing during the bookmaker review process. Test accounts are funded with our own money. Test withdrawals are processed through our own bank channels.

If you ever have reason to suspect we have failed to follow these standards, please email [email protected]. We will look into it, and if we got something wrong, we will say so on this page.