The record
When a model disagrees with the market, who is right?
We measured it, on every settled selection we have, for two different forecasts. The answer is uncomfortable and it is the most useful thing on this site: the further a forecast strays from the closing price, the worse it does. Wherever the gap is wide enough to act on, the price is the better forecast — on the count and on the scoring rule that punishes confident mistakes alike.
Two forecasts, measured the same way against the same closing prices. They do not have the same record and they do not have the same threshold — which is the point of showing both.
Check a disagreement
Put in what a forecast says and what the market says. The record answers with what happened the last time a gap that size was settled — and refuses to answer when the gap is too small to mean anything.
Your model, a tipster, your own read — any probability that did not come from the price.
The price with the margin taken out. Take the sharpest book you can see and strip its cut.
Disagreement 8.0pp
Reading the record…
The record is read from the calibration table at load; its sample and threshold appear here once it answers.
Every band, with its sample
Each bar runs from the coin flip: right when the market finished closer to the result, left when the forecast did. Read them downward — a forecast holds its own while it agrees with the price, and loses ground as it argues.
Each row is a band of disagreement between the forecast and the sharp closing price, and the figure is how often the market turned out to be closer to the result. Bars run from the coin flip: right when the market finished closer, left when the forecast did. The further the forecast strays, the more reliably it is the one that was wrong. That is why a model holds a veto here and never a price.
Two scorekeepers, and they disagree
There are two ways to ask who was right, and they do not always answer the same. One counts how often each side finished closer to the result. The other scores every forecast by how much probability it put on what actually happened — a Brier score, which punishes being confidently wrong rather than merely being on the wrong side.
Reading the record…
Where they split, the count is the one doing the flattering. It rewards a forecast for winning near-ties by a hair and never notices it losing the occasional band by a mile — which is what the last column measures. A model with crude inputs hedges toward the middle, wins more hairs, and looks better than it is.
This is why the threshold is a property of the forecast and not of the site. Each one is set above every band where its own two scorekeepers contradict each other. Setting a single number for everything would have meant mislabelling whichever model it was not written for.
The limits, stated
It describes these forecasts. A Dixon-Coles goals sheet fitted per season, and an Elo ladder built from results alone. A different forecast — yours, a tipster’s, a neural net’s — has its own record, and neither of these transfers to it. What almost certainly does transfer is the shape: a forecast fighting a liquid closing price is fighting the aggregate of everyone who has already done the work.
Under its own bar each forecast says nothing, deliberately. In the narrow bands a “model right 51%” headline drawn from a tie would state, in the only way a reader parses it, that the model beats the market. It does not — and where the closer-count says otherwise, the Brier score says it plainly.
At the narrowest gaps there is nothing to separate them. Under a single point of disagreement the forecast and the price score within a rounding error of each other, and for the goals model that band tips marginally to the forecast. That is not an edge — it is what a tie looks like when you slice it thinly enough, and it is exactly the region each model’s bar exists to keep off the page.
It is about accuracy, not profit. Being closer to the truth and being a winning bet are different things — a price can be wrong in your favour on a match your model has no read on at all. Our own settled record, including the parts that are not yet significant enough to publish, is on the performance page, and the method behind all of it is on how it works.