League tables in the opening weeks look like information and mostly are not. A handful of matches cannot separate a strong side from a lucky one, and forecasts built on them wobble accordingly.
Four matches cannot rank twenty teams
After a month, the gap between top and bottom of the table is often a few results wide. Almost every position in the middle is within one weekend of every other.
The ordering therefore encodes very little about relative quality. It records who has played whom and who has had the run of the ball so far.
Presenting it as a standing is nonetheless unavoidable, since a league has to be displayed somehow. The mistake is reading it as though it were a season's worth of evidence.
Fixture order does most of the early work
Two teams of identical strength can sit ten places apart in September purely because one has played the weakest opponents available and the other the strongest.
Adjusting for opponent quality usually collapses most of the apparent spread. What remains is a much smaller and much less dramatic set of differences.
Because the fixture list balances out over a full season, this distortion is temporary. It is also the largest single source of early-season nonsense.
Why last season still matters
Sensible models start each season with an estimate carried over from the previous one, then update it gradually as new matches arrive.
That carried-over view outperforms the current table for a surprisingly long stretch, often well past the point where supporters have stopped consulting it.
Squad turnover weakens the link, and a heavily rebuilt team deserves a wider margin of uncertainty. It rarely deserves to have its history discarded entirely.
Promoted sides are the genuine unknown
A newly promoted club has no recent record at the level it is now playing, so the carried-over estimate has to be translated across divisions.
That translation relies on how previously promoted teams have fared, which is a coarse instrument. Individual clubs vary enormously around the average.
Forecasts for these teams should be visibly less certain than for the rest of the division, and the honest ones are.
How the noise fades
Roughly a third of the way through a season, results start to carry more signal than the prior estimate, and the table becomes a defensible summary.
Before that point, underlying performance measures give a faster read because they accumulate from many more events per match than goals do.
The practical rule is to treat early tables as a record of what has happened rather than a guide to what will. They are accurate history and poor prophecy.