How can I tell if a horse racing tipster is fake?

Apply five tests: picks must be published BEFORE races with timestamped odds; losses must remain visible in the record; returns must use prices that were actually available; the sample must be large (hundreds of picks, not a hot month); and results should be compared against simply backing favourites. Fakes fail on deleted losers, 'advised at' prices nobody could get, and screenshots instead of ledgers.

How many tips does it take to prove a tipster has skill?

Statisticians put meaningful evidence at 500+ settled picks minimum, and real confidence needs thousands. Any record shorter than a few hundred picks — however profitable — is indistinguishable from luck.

What is a proofed betting record?

A record where every selection is logged publicly before the event with the odds at that moment, settled from official results, and never edited afterwards — so anyone can verify both the wins and the losses independently.

OddsLens

Data desk · Consumer defence

How to check if a horse racing tipster is legit

Somewhere right now, a Twitter account with a Lamborghini header photo is posting a betslip screenshot captioned "TOLD YOU 🔥". Here is the uncomfortable truth about that screenshot: it proves nothing, and the person who posted it knows it. Tipping is an industry where the product is a record, and most records you'll see have been quietly edited into fiction. Fortunately, fakes are easy to catch. Five tests, all verifiable in minutes:

Test 1

Were the picks published before the race?

A genuine record shows every selection timestamped before the off. If tips arrive as after-the-fact winner screenshots, or the "record" is a spreadsheet the tipster controls, walk away. Anyone can win yesterday's races.

Test 2

Are the losers still there?

Scroll their history. A tipster with a 40%+ claimed strike rate and no visible losing runs has deleted them — losing runs are mathematically unavoidable, even for genuinely skilled operators. The record that shows its worst weeks is the only kind worth reading.

Test 3

Could you actually have got the prices?

"Advised at 16/1" means little if the horse was 8/1 everywhere within a minute of the tip. Honest records settle at starting price or at odds demonstrably available when published — and say which.

Test 4

Is the sample big enough to mean anything?

A profitable month is luck's favourite disguise. Meaningful evidence starts around 500 settled picks; real statistical confidence needs thousands. Anyone selling a subscription off six good weeks is selling variance.

Test 5

Does it beat just backing favourites?

The default strategy — back every SP favourite — wins about 35% of races and loses about 10p per £1 (our data here). Any record must be judged against that baseline over the same races, not against zero. Losing less than favourites is progress; most paid tips do worse.

Red flags that end the conversation

"DM for VIP group" · betslip screenshots as proof · guaranteed daily profits · pressure to "get in before today's banker" · a record hosted only where the tipster can edit it · strike rates above 40% claimed over any serious sample · affiliate links to sign up at specific bookmakers as the actual business model.

Full disclosure: we publish picks too — from a machine-learning model, not a person — and this article is the standard we invite you to hold us to. Every OddsLens pick since April 2026 (3,000+) is logged pre-race with timestamped odds, settled from official results, benchmarked against favourites, and displayed losses-first-class on a page we cannot edit after the fact. It shows losing runs right now. That's what a real record looks like.

Apply the 5 tests to our record →

Quick answers

How can I tell if a tipster is fake?
Five tests: pre-race timestamped picks · losses still visible · prices that were actually available · 500+ pick sample · beats the favourite baseline. Fakes fail at least one, usually all five.

How many tips prove skill?
500+ settled picks for meaningful evidence; thousands for confidence. Anything less is indistinguishable from luck.

What is a proofed record?
Selections logged publicly before the event with odds at that moment, settled from official results, never edited afterwards — independently checkable, wins and losses alike.