We got tired of watching
smart people bet blind.
SharpCut started because the maths that professional bettors use has been public for fifty years — and almost nobody in the largest betting market in Africa has ever been shown it in a language they think in.
The Saturday that started it
A friend of ours — call him what everyone calls him, Chidi — lost ₦3,000 on a five-game accumulator. Nothing unusual there. What stuck with us was the conversation afterwards. He could name every player in the Arsenal squad. He knew Chelsea hadn't kept an away clean sheet since August. He had genuinely good instincts about football.
He also had no idea that his five-game accumulator at odds of 4.50 had a true win probability of about seven percent, or that the bookmaker was paying him as though it were twenty-two. Nobody had ever shown him. Not because the information is secret — it's arithmetic — but because every tool that explains it does so in a dialect of statistics that assumes a university education and a comfort with the word “stochastic.”
The gap in this market was never intelligence. It was translation.
What we actually built
SharpCut computes true probability, expected value and a Kelly-sized stake for every match across 150+ leagues. So do several other products, and some of them do it well. The part we spent our time on is what happens after the number exists:
- An explanation of the pick that a person can read in Pidgin, plain English, Yorùbá, Hausa, French, or full analyst notation — same model output, seven mouths.
- A record of your accuracy, not just ours, so the product can tell you which markets you're genuinely good at and which ones are quietly costing you money.
- A virtual bankroll that runs from day one, so after thirty days you have evidence instead of marketing.
- A public track record with the losing months left in.
Where our incentives sit
SharpCut is paid by subscribers, not by your losses. We hold no stake and take no cut of a bet, which means a user who blows a bankroll in six weeks is worth nothing to us and a user who bets carefully for years is worth a great deal.
You shouldn't take our word for that, though — which is exactly why the track record is public, permanent, and still carries the months we'd rather you skipped.
What we won't do
- We will never hold your money. SharpCut is not a bookmaker, has no wallet, and cannot accept a stake.
- We will never sell a “sure game.” There is no such thing. Anyone selling one is selling you something else.
- We will never delete a bad month. The track record is append-only.
- We will never market to anyone under 18.
How we work
Five principles that decide arguments when we can't agree on something.
Show the maths or don't ship it
Every number a user sees must be traceable to a data source and a formula. If we can't explain where a probability came from, it doesn't go on a card. A Match Story must always quote the same probability the card does, in every language, with no room for drift.
Write for the betting shop, not the boardroom
If a first-time bettor in Kano can't read it, it isn't finished. Jargon is a failure of effort, not a signal of rigour. The analyst voice exists for people who want it — it is never the default.
Publish the losses
Any product in this category can look brilliant by showing you a good week. Our track record is append-only and filterable by month, market and confidence band, including the months we'd rather you skipped.
Protect the bankroll before growing the number
When a growth idea and a user's financial safety conflict, the bankroll wins. We cap stake recommendations, we flag chasing behaviour, and we do not send win-back messages to users who have just lost.
Naira first
Prices, stakes, profits and examples are all denominated in the currency our users actually hold. A dollar price tag on a Nigerian product is a signal that the product wasn't built for Nigeria.
Come and check our workings.
Every prediction we've ever published, with the results attached. Filter it however you like.