Final scores, checked against the call.

Most tips sites go quiet about the ones they got wrong. We do the opposite on purpose. Every result on this page — the wins and the upsets alike — stays on the board with the same visibility, because a results archive that only shows correct calls isn't a results archive, it's a highlight reel. If FDVS is going to ask you to trust a pre-match consensus number, the honest thing to do afterward is show you exactly how that number actually held up once the final whistle went.

How the verdict is decided

Every completed fixture gets checked against the single clearest claim its pre-match consensus made — usually the match winner, or a draw when the consensus leaned that way. If Man City were favoured at 61% and they won, that's marked correct. If Inter were favoured at 68% and the match finished level, that's marked as an upset, plainly labelled, not softened into vague language. We don't retroactively pick a friendlier market to grade against once we know the score — the verdict is locked to whatever the consensus actually said before kickoff.

What "hit" and "miss" mean underneath a result

Below the final score, the market chips tell a more granular story than the single win/loss verdict does. A match can carry an incorrect win verdict while still showing several hits underneath it — Over 2.5 goals landing, for instance, even when the favoured side didn't win. Reading both layers together, the top-line verdict and the market-by-market hits, gives a fuller picture than either number alone. A team can lose a match everyone expected them to win while several of the secondary markets still played out exactly as read.

Final xG versus pre-match xG

The xG line on a completed match reflects what actually happened during the game — the quality of chances both sides created in real time — not the rolling pre-match average that fed the original consensus. Comparing the two tells you something the scoreline can't: a team can win 1-0 while being outcreated on xG, which is a very different kind of win than a 1-0 built on a 2.4 to 0.6 chance quality gap. When a correct verdict is backed by a lopsided final xG line, that's a genuinely convincing result. When a correct verdict comes with a close final xG line, the scoreline flattered the winner more than the performance did.

Why upsets matter more than another streak of hits

It would be easy to only surface results where the consensus was right, and it would make the page look sharper than it actually is. We'd rather you see the Bayern home defeat sitting right next to the Real Madrid landslide, because that contrast is the actual point: no consensus number, however well-built, converts to a guaranteed outcome. Upsets are where the model, the market, and recent form genuinely disagreed with what actually happened, and reviewing those cases — not hiding them — is exactly how the underlying process gets sharper over time.

How results feed back into the system

Nothing on this page is a dead end. Every graded result — hit or miss — becomes another data point folded back into how future fixtures get read: which markets tend to run hot for a given league, which teams' recent form has been over- or under-weighted, where the community consensus has been sharper than the statistical model or the other way around. A results archive that's actually reviewed, rather than just published and forgotten, is what keeps the fixtures board from calcifying around last season's patterns.

Reading accuracy over time, not one result at a time

A single upset doesn't mean the consensus was badly built any more than a single correct call means it was perfectly built — football has enough built-in variance that even a well-calibrated 70% read will be wrong three times in ten, by design, not by failure. The useful way to read this page isn't fixture-by-fixture in isolation, but as a running record you can check back against over weeks: does the 70%+ band actually land close to seven times in ten over a real sample, or not. That's the standard we'd want held to, and it's the same standard this page is built to make checkable.

From today's results to tomorrow's fixtures

This page and the fixtures board are two views of the same underlying process, just facing different directions in time — one showing what the consensus expects, the other showing how it actually turned out. If you're deciding how much weight to put on tomorrow's consensus numbers, this is the page that shows you what that same process has actually delivered recently, mistakes included. Free access covers the full results history shown here; Plus and Pro add the deeper source-by-source breakdown behind each verdict, for anyone who wants to see exactly where a call came from before deciding how much to trust the next one.