Most places that publish a number about a sporting event publish one number. Chicago at 54 percent. Texas at 19.8 to win the championship. Tomorrow the number is different, and yesterday’s is gone.
We keep them all, and we show you the whole run. This is a short explanation of why, because it is the thing that makes this site different from the alternatives, and it is not obvious at first glance.
A number on its own tells you very little
Suppose you open a game page and see a team at 54 percent. What have you learned?
Not much. You know the model thinks the matchup is close to even, tilted slightly one way. That is genuinely useful, but it is a single frame from a film.
Now suppose you also know that the number opened at 61, drifted down through the low fifties across a fortnight, and has been climbing back since the weekend. That is a different piece of information entirely. Something happened. The model is responding to inputs that changed — form, availability, rest, whatever the schedule did — and you can see the shape of that response rather than just its current position.
Or suppose the number opened at 54 and has not moved a tenth of a point since. That is also information. It says the model has seen nothing in three weeks that changed its mind, which is a meaningfully different situation from a number that arrived at 54 this morning after a week of turbulence.
Same figure on the page. Three quite different stories underneath it.
Why almost nobody keeps it
Keeping a history is not technically difficult. The reason it is rare is more mundane: for most publishers, yesterday’s number has no commercial value. A bookmaker’s price exists to balance a book today. A tipster’s pick exists to be right or wrong once. Neither has much reason to maintain an archive, and a bookmaker in particular has some reason not to — a public record of where their line has been is a record of where they were wrong.
We keep it because the trajectory is the part we think is actually worth reading, and because it is the only honest way to let anyone check our work. If you are going to publish probabilities, you should be prepared to have them looked at afterwards.
What we show
Every event page carries the full recorded history of its number, from the first time we published one. Alongside it we plot a ten-day trailing average, which smooths out day-to-day noise so the underlying direction is visible. The rising or falling label is measured against that average rather than against the opening figure, because a number can sit above where it opened while clearly easing — and calling that “rising” would contradict what the chart plainly shows.
Where a series starts part-way through, it is because that is when we first priced the event. We do not backfill estimates for periods in which none existed. A short line means a new event, not missing data.
Team and participant pages do the same across a whole season, and across every futures market that team is entered in. So you can look at one game, or at how a club’s championship number has travelled since August.
What it is not
A moving number is not a signal to act on, and we are not going to tell you it is. We do not publish selections, rate one side above another, or flag anything as good value. Both sides of every event are shown with the same fields and the same emphasis. Where a page has a favourite, that is arithmetic rather than endorsement.
What the history gives you is context for your own judgement. A number that has been stable for a month and a number that moved four points overnight deserve to be read differently, and you cannot tell them apart from the figure alone.
Early days
The site is new and we are adding to it steadily. Some sports carry more detail than others, some series are only a few weeks deep, and parts of the model are still being validated. We would rather say that plainly than present an early-stage product as a finished one.
If you find something that looks wrong, we would genuinely like to hear about it. The methodology page explains how the numbers are built, what we check, and what we are not yet claiming.