Ten games into a season, the table already looks like an argument. A team is "in crisis". Another is "title contenders". Both labels come from a column of numbers that is roughly 26% signal and 74% luck.
How much of the table is noise
Points after ten matches correlate with final points, but weakly — the standard comparison is around 0.4, which means most of the variation disappears. The reasons are not mysterious:
- Fixture difficulty is not equal. One team has played three of the top four. Another has played three of the bottom five.
- Close games are coin flips. Most matches are decided by a single goal or a draw, and single-goal outcomes are close to random given the same performances.
- Red cards and penalties cluster. A team can lose four points in ten minutes for reasons that say nothing about their quality.
The two columns that stabilise faster
Goals for and goals against tell you more over a short sample than points do, because they aggregate more events. A team with a strong expected-goals margin and a mediocre points total is usually a team that will improve — the results are lagging the performance, not the other way round.
The reverse is more common than people think: a team with eight wins from ten but a negative goal difference is riding variance. The table says "good". The numbers say "not yet".
What to actually look at after ten games
- Goal difference, not points. It is less noisy and it points forward rather than backward.
- Points per game against top-half opponents. That is a rough proxy for how the season will feel in March.
- Whether the gap is bigger than the fixture difficulty gap. A six-point lead earned against the bottom six is not a six-point lead.
The arithmetic people get wrong
A common mistake is reading a gap as decisive when it is not. A six-point lead with 28 games left is not safe; it is one bad fortnight away from being two points. Our points calculator lets you check this directly: enter the gap and the games remaining and it tells you how many wins are needed, not how many points — because a draw does not close a gap.
Reading a table is a skill, and it is not the same skill as remembering where a team finished last year. The league table course covers the full set of things a table conceals.