The Championship play-offs have a habit of defying expectation, and BBC Sport's Ask Me Anything team has turned to statistics in an attempt to find out whether the numbers can cut through the chaos and identify a likely winner.
The premise, as BBC Sport frames it, is a reasonable one. Modern football analysis has grown sophisticated enough to surface patterns that the eye misses, and a tournament of this size and commercial significance — the play-off final is routinely described as the most valuable single fixture in club football — generates enough data across a full season to make statistical modelling at least plausible.
The difficulty, as anyone who has watched the play-offs with any regularity will appreciate, is that the format itself works against prediction. Five fixtures across two legs and a Wembley final compress an entire season's worth of variance into a very short window. A goalkeeper's afternoon, a deflection, a sending-off in the first leg — any of these can override whatever the underlying numbers suggest about which side is the better team.
That structural volatility is part of what makes the play-offs so watchable and so commercially potent. It is also what makes them resistant to the kind of modelling that works reasonably well over a 46-match season, where sample sizes are large enough for quality to assert itself. In a knockout format this brief, the gap between expected outcomes and actual results tends to be wide.
BBC Sport's analysis may well identify correlates — possession metrics, expected goals tallies, defensive solidity across the campaign — that tilt the odds towards one contender or another. Whether those tilts are enough to constitute reliable prediction is the more interesting question, and one the sport's analysts have not yet answered to general satisfaction. The play-offs, for now, remain as open as they have ever been.