One prices the grid before a ball is kicked. The other sits inside the match and waits for the moment the numbers turn. Different edges, same rule: the signal goes public before it settles.
The model prices all 25 endings from confirmed lineups and trades a portfolio of five or six scorelines where the market pays more than it should.
Fixed rules read the state of a live match and fire when it turns. No discretion, no overrides, no tuning after a bad week.
Confirmed starting lineups, attacking output, defensive setups, tempo of the fixture. No gut feel, no narratives — inputs only.
Every plausible scoreline gets a model probability. Where the market pays more than the model says it should, that is the edge.
A single score is a lottery ticket. A priced portfolio is a position: one hit covers the book and prints the profit.
Every signal timestamped before it can settle, wins and losses alike. Nothing gets edited afterwards. That is the whole product.
Each one watches for a different kind of match. What they look for stays in house; what they produce is public, and it updates here every day.
Pressure without a breakthrough. One side owns the match and still hasn't scored.
The strictest layer of the family. It only speaks when everything lines up at once.
Built for chaos rather than for the scoreline — matches that swing end to end.
Both sides still pushing, one clearly on top. A duel that usually ends in a goal.
You get one scoreline. The model prices the whole grid and backs five. Kick off and see who is up at full time.
Where it runs, when it fires, and what it looks like in practice.
On Betfair Exchange, exclusively — and only on high-liquidity fixtures: the main first divisions. A correct score grid is only worth trading where the money is deep enough to get filled at the price the model wants.
Those run with soft bookmakers. Different market, different edge — the exchange is not where those prices live. The prices we quote are taken from bet365.
The late-goal models read the match strictly at the 70th minute and speak within seconds of it. Nothing fires before, nothing fires after.
The position goes on at a minimum price of 1.50. If the market sits under it, the bet waits until it gets there.
From there the staking is down to you. Some put the whole stake on that line. Others split it: half on the Asian handicap x.0, half on the next line up once it reaches 1.50. Both are fine — the signal gives you the match and the moment, the rest is each trader\'s own strategy.
It happens now and then — the signal goes out and the ball is already in the net. Worth knowing before it happens to you.
Across the record it is worth about one point of hit rate, so it is a nuisance rather than a hole in the numbers.
Four of them, and they are the reason the record is worth reading.
A track record with no red in it is fiction. Ours has losing days on the wall like everything else — including the model currently running under its own standard.
Every signal goes public before it can settle, where edit history doesn't exist.
If the price sits under it, the bet waits until it gets there. Short prices eat a hit rate alive.
A model is frozen before it goes live and never re-fitted after a bad week. When one stops working it gets retired in public.
The channel is free. Grids land after lineups, models fire inside the match, results land at full time, and the log never gets edited.
JOIN EXACT ON TELEGRAM