How to spot a fake betting record

Faking a betting record takes about ten minutes and no technical skill. Verifying one takes about two minutes once you know what to look for. That asymmetry is the entire reason this category works the way it does. None of what follows is obscure. It's just rarely written down by anyone in the industry, because most of the industry benefits from you not knowing it.

Post ten different selections to ten different groups. One group gets a perfect week. Show that group's results to everyone. The modern version is multiple accounts, or a private channel where losing tips quietly disappear. You are shown the surviving branch and told it's the whole tree. The check: does one source, one channel, one account carry every selection? If picks arrive from several places, or a channel's history has holes in it, you're looking at a survivor.

The selection was real. The price wasn't. Claiming 2.10 when the widely available price was 1.85 turns a losing record into a winning one without a single fabricated result. Best-price shopping is legitimate, but the price has to be one that actually existed, at a book you could actually use, at the time you could actually have bet it. The check: are odds recorded with the bookmaker and the timestamp? "2.10" is unverifiable. "2.10, Betfair, 14:32" can be checked against historical data.

Straightforward: bad weeks get deleted. On platforms without edit history this is undetectable after the fact. The subtler version is reclassification. A losing pick becomes "not an official selection", a "fun bet", or "advisory only" — after it loses. The check: is the record append-only somewhere with independent timestamps? A third-party proofing service, a public commit history, or posts that can be shown to predate kick-off. Self-hosted, self-editable results are worth exactly what it costs to edit them.

Every strategy has a good stretch. Pick it, start the chart there, and the line goes up. Watch for records that begin at a suspiciously convenient date, or that quietly switch method partway. "We refined the model in March" is often "the results before March were bad". The check: does the record start when the service started? Does it cover the losing periods too? A record that begins mid-season is a chart with the first chapter torn out.

You don't need to detect each technique individually. Four questions defeat all of them. 1. Was every pick published before kick-off, with an independent timestamp? 2. Is the record complete — every selection, no gaps, no reclassification? 3. Is each selection priced, with the bookmaker named? 4. Is it long enough to mean anything — several hundred bets, not a hot month? Any record satisfying all four is trustworthy whether it's up or down. Any record failing one is unverifiable, which is not the same as fake, but should be treated the same way — because you have no way to tell the difference.

The fair response to an article like this is to point it at whoever wrote it. So: from the Premier League opening weekend, every pick we make is logged before kick-off and settled afterwards, with the price and the bookmaker recorded, in one place, complete. We published the method in advance — market, league, confidence threshold, stake — before placing a bet, so it can't be adjusted retroactively. Run the four checks on us. If we ever fail one, we've earned exactly the scepticism this page describes.