A week ago we opened something like a public ledger: nine teams whose results had drifted away from what they were producing on the pitch, five underperforming and four overperforming, and a commitment to report back on how it went without picking the convenient pieces.
Now they have all played, and there is an international break, which is the right moment to settle up.
The nine, one by one
Monza were the most underrated team in Italy, 1 point from 7.74 expected. They beat Sassuolo 2-1. Went as it should.
Bologna were the biggest gap in the league, 1 point from 6.94. They drew 1-1 with Torino, moving their points total but playing their poorest match, 0.70 expected goals. Half a point in their favour.
Villarreal were the extreme case in La Liga, 2 points from 7.47. They won 3-1 at Malaga producing 4.25 expected goals, then won 3-1 against Levante producing 4.29. Two wins, two performances worthy of a top side.
Lecce were among the inflated ones, 6 points from 2.06. They lost 3-0 in Milan, finishing the match with 0.22 expected goals. Went as it should.
Atalanta were on the same list, 6 points from 2.33. They lost 2-0 in Turin against Juventus, with 0.49 expected goals. That one too.
Lille, 10 points from 4.69, lost 2-1 at Nice in a match that expected goals call even, 1.93 to 1.93. The direction is the right one.
And then the three that did not go the way the numbers said.
Brest, whom we flagged as the team producing the most in France without collecting, lost 2-1 at Auxerre creating 0.30 expected goals: not only did they fail to cash in the credit, they also stopped producing for an evening.
Betis, whom we called inflated, won 1-0 against Getafe and drew 1-1 at La Coruña. Four points from six, and above all a performance worth 2.61 expected goals against Getafe: here the model was right on the points and wrong on the direction, because Betis in the meantime has started producing for real.
Cagliari won again, 1-0 at Udine, entering the opposition box once in ninety minutes. It is the third result in a row they have got that way, and they now have 12 points from 5.77 expected.
Six out of nine in the right direction, one halfway, and two clearly wrong.
What we learn, and it is not what it looks like
The temptation, after a reckoning like this, is to say the method works six times out of nine. That would be wrong for two reasons.
The first is that a single matchday is not a test. Over ninety minutes the noise is bigger than the signal, and a team that deserves 1.2 points per game always gets three or zero: if we had done the same count the week before or the week after, the result would have been different without anything changing in the method.
The second is more useful. The Betis case says that the gap between actual points and expected points, on its own, is a crude indicator. Betis were collecting too much relative to what they had produced in the early rounds, but their production was rising week after week: they now have 13.15 expected goals produced in seven matches. A gap that closes because the team improves is not the same as a gap that closes because the luck runs out.
The right way to read it, then, is to always cross two things: how much a team collects relative to what it produces, and in which direction its production is moving. Villarreal was the perfect case precisely because the two things pointed the same way.
The open positions, updated at the break
In Italy the biggest gap remains Cagliari's, 12 points from 5.77 expected, followed by Lazio, 13 from 8.00, and Frosinone, 10 from 5.19. On the other side Bologna, 2 points from 8.02, Venezia, zero points from 5.48, and Monza, 4 from 8.68.
In France Monaco have 13 points from 6.82 expected, which is almost exactly the Cagliari case: first place and mid-table production. On the other side Lens have 4 points from 8.82 and Brest 5 from 9.75.
In England Manchester City lead with 15 points from 9.59 expected, and that is the biggest gap in the Premier League. Underperforming are Bournemouth, 3 points from 7.76, Nottingham Forest, 5 from 9.17, and Sunderland, 4 from 7.69.
In Spain Betis remain top of the lucky list, 16 points from 9.83, but with the rising production mentioned above. Villarreal, despite two wins, are still on 8 points from 13.33 expected, and have carried 109 balls into the opposition box in seven matches: that is the highest figure across the five leagues.
In Germany Borussia Dortmund have 12 points from 8.10 and Bayer Leverkusen 7 from 9.64.
The honest part
First, what we are not doing. We do not give betting tips and we do not attach any of these names to a price: the odds we collect stop at last spring, so we do not even know how much the market is already pricing these divergences. Without the price, an idea is not a bet.
Second, the margin, when there is one, is thin. Our test on 26,726 simulated bets says that betting blind on the simple markets loses 6.14% purely because of the bookmaker's margin, and that the market price gets probabilities right to within two percentage points. Any edge built above that level requires hundreds of bets to show itself.
Third, the most important thing. Betting remains gambling. This column exists to tell the story of how to read a season through numbers and to keep the ledger in public even when it goes badly, not to promise earnings.
The next reckoning after the restart, with today's nine and the new names.
T.L.
Want to see how much each team really deserves its league position? The Meritometro compares actual points and expected points. For those who want the models and not the headlines, there is the Nerd Zone.