Slow starts, explained: who should worry and who can improve?

Aston Villa left Tottenham with their first league win of the season. Spurs finally scored their first league goals. Neither side could leave the ground thinking everything was fixed.

That is the difficulty with September football. A single result can transform the mood without resolving the problem. Equally, a team can spend weeks near the bottom while playing better than its points suggest.

Our reading of the opening phase is that Gladbach and Valencia have substantial performance problems to address. Tottenham's difficulties extend beyond finishing, while Villa still need to show that one encouraging result can become a pattern. Bournemouth have a stronger case for patience; Bologna have some evidence in their favour, but their recent attacking performances weaken it. Lens and Marseille show why even a reassuring statistic needs a closer look.

This is a selection of eight clubs across the Premier League, LaLiga, Serie A, Bundesliga and Ligue 1, using league matches through 20 September. It completes our review of the opening phase. It is an assessment of the evidence so far, not a forecast of the final table.

What makes a slow start worrying?

Start with three questions. Is the team creating enough? What is it allowing at the other end? And how much of that picture comes from the opponents it has faced?

Expected goals, or xG, helps with the first two: it estimates the likelihood of shots becoming goals from their characteristics. Creating chances worth more than the goals scored can suggest finishing has contributed to a poor return. It cannot tell us that better results are owed, or isolate a single cause such as bad luck or goalkeeping.

We also need the individual matches. One heavy defeat can dominate a total. One big win can hide several flat performances. And missing xG is missing evidence, not a zero.

With that in mind, the most useful distinction is between teams already producing something worth building on and teams that need to change the football itself.

Tottenham and Villa: the same game, two unfinished repairs

Tottenham: two points from five games. Aston Villa: four from five.

The 3–2 Villa win ended two uncomfortable waits. For Spurs, however, the timing of their goals matters: they were already three down before scoring twice late on. The goals ended the drought; they did not rescue a point.

There was real encouragement in Tottenham's early play. The club's match assessment records important saves by Zion Suzuki and a blocked Dominic Solanke header. Seven shots on target were a step up from the previous two league games combined, when Spurs managed just two.

But Roberto De Zerbi's own post-match explanation focused on defensive organisation and the balance lost while chasing the game. Those are problems that a striker rediscovering his touch cannot solve alone.

Across the opening four fixtures, Tottenham produced 3.26 xG and allowed 6.74. They scored none, so finishing clearly forms part of the discussion. Yet the chances they allowed were worth more than twice those they created. This was not simply a strong attack waiting for its reward.

Villa's equivalent figures were 3.37 for and 7.66 against, with one goal scored. Their three goals at Tottenham were therefore a welcome attacking improvement. The full five-game picture remains uncomfortable: Villa have been outshot 81–48, with a 26–12 deficit in shots on target.

There is no xG for their meeting in our saved dataset, so those four-game figures should not be read as an assessment of the 3–2 itself. The new evidence deserves weight, without erasing the games before it.

Our earlier Tottenham analysis examined the goal drought. The next question is broader: can Spurs create danger while keeping enough protection behind the attack? For Villa, it is whether they can produce another convincing attacking performance without needing their goalkeeper to keep the game alive first.

What to watch: Spurs' organisation when they lose the ball or fall behind; Villa's ability to sustain chance creation across the next few matches.

Gladbach: six goals cannot cover sixteen conceded

No points from four games. Six goals scored, sixteen conceded.

The headline attacking total is not the main concern at Borussia Mönchengladbach. They have scored three in each home game and lost both 4–3. Away, they have lost 3–0 and 5–0.

Opponents have put 38 shots on target across those four matches, against Gladbach's 17. The first three games with xG available show a 3.64–8.87 deficit. Both measures point towards repeated defensive exposure, even though they do not identify every tactical cause.

The Mainz defeat shows why attacking flashes are insufficient. Gladbach led, later recovered to 2–2, then conceded twice more before a late penalty narrowed the score. The Bundesliga's report also records an earlier missed Gladbach penalty. There were opportunities to change the result, but another four goals conceded remained the larger issue.

There has since been a significant change: Alexander Blessin was appointed on 22 September, after this article's performance cutoff. The club confirmed his arrival on a contract until 2028. These four results belong to the period before him; they are not a verdict on his work.

What to watch: whether opponents find it harder to get shots on target. Even an untidy, low-scoring game would tell us more about defensive progress than another dramatic 4–3.

Valencia: signs of life, with a low attacking baseline

Four points from seven games. Four goals scored, thirteen conceded.

Valencia went scoreless in four of their first five league matches. Their opening six produced 4.51 xG, only about 0.75 per game, against 8.85 conceded. A team creating at that rate has little room for a poor finish or an individual defensive mistake.

The 5–0 loss to Barcelona weighs heavily on those figures, and should not define every performance. Remove that match from the six with xG, though, and Valencia still trail 3.85–4.77. The balance improves considerably, but it does not turn into evidence of sustained attacking superiority.

The final two games brought something more hopeful: a 1–0 win at Alavés and two goals in the 3–2 defeat by Real Sociedad. The latter produced 22 Valencia shots. According to the club's report, they equalised twice before conceding the winner in added time.

That is a useful qualification to the bleak overall numbers. There was more attacking activity, and the match remained competitive. But 21 shots conceded in the same game underline how far a complete improvement still has to travel.

25 September data update: xG for that match is now available: 1.55 for Valencia and 2.65 for Sociedad. It reinforces the caution. The six-game totals above retain the original figures used across this series.

Óscar Sánchez was the coach giving the post-match assessment on 20 September. Valencia have since appointed Javier Aguirre, with the club announcing his arrival on 23 September. He inherits the record assessed here; these results do not judge his work. The latest evidence offers something to build on; it does not yet establish a recovery.

What to watch: whether Valencia can repeat that attacking involvement while reducing the chances at the other end. More action is encouraging only if they gain some control over it.

Bournemouth: a winless start that deserves some patience

Three points from five games, from three draws.

Bournemouth have lost to Manchester City and Liverpool, drawn away at Newcastle and shared the points at home with Everton and Brentford. That fixture list gives the lack of a win some context.

More encouragingly, they have outshot their opponents 68–54, with a 26–20 advantage in shots on target. Their four matches with xG show a small positive balance, 5.91–5.41. Those are signs of a competitive team, rather than one repeatedly unable to get into games.

The 2–2 with Brentford is the clearest example: Bournemouth led the shot count 20–8 and the saved xG 1.75–0.50. One game cannot prove that points will follow, but it helps explain why a winless record need not mean uniformly poor performances.

There are limits to the reassurance. The latest 1–0 loss to Liverpool brought only two Bournemouth shots on target. Liverpool's report describes late pressure that often ended in distant attempts or crosses the defence could handle. Our xG coverage does not include that game.

Bournemouth have earned a measured reading. They still need to turn pressure into chances that make a draw become a win.

What to watch: whether the positive shot balance continues, and whether their strongest attacking spells produce better openings before the match becomes a chase.

Bologna: some wastefulness, but less encouragement lately

Two points from five games. Three goals scored, six conceded.

Bologna are a useful contrast because xG covers all five of their league games. They have created 5.54 xG and allowed 5.08: a broadly even chance balance alongside a much poorer set of results.

The opening defeats give the clearest reason for sympathy. Bologna lost 1–0 to Lazio despite a 1.78–0.63 xG advantage, then 1–0 at Atalanta with the saved chances favouring them 1.05–0.15. Across those games they took 30 shots and scored none.

There is a plausible finishing problem in that evidence. There is also a warning against treating the aggregate as a clean bill of health.

The last two fixtures, at Napoli and at home to Torino, produced just 0.83 xG combined, against 2.68. Bologna's only shot on target in the Torino draw became their goal. Their latest problem was not a procession of clear opportunities going unfinished.

Bologna have already changed coach. Raffaele Palladino replaced Domenico Tedesco on 16 September, making Torino his first league game in charge. The five-match record describes the team's start, not his tenure. His immediate task is to restore the chance creation that made the opening defeats look so frustrating.

What to watch: the return of chance creation. Better finishing helps most when there is enough to finish.

Lens and Marseille: check what is inside the reassuring number

Lens: four points from five. Marseille: three from five.

Lens have several encouraging totals: 74 shots against 56, 32 on target against 25, and a positive 10.18–7.86 xG balance in four covered games.

Much of the xG advantage, however, came from their opening 5–2 win over Auxerre. In the next three covered matches, Lens trailed 5.77–6.02. That is still competitive, but it is a different picture from a team consistently creating far better opportunities than its opponents.

The 2–1 defeat at Monaco, which has no xG in the snapshot, offers another reason to avoid a sweeping verdict. Monaco's report records a key Lukáš Hradecký save before Jonathan Gradit's own goal decided the match. Lens were in the contest; that alone does not establish how strongly they should be expected to recover.

Marseille's apparently positive xG balance is less reassuring still. The 4.60–3.78 total covers only three games. It excludes defeats by Paris FC and PSG, while including the opening 4–0 win over Strasbourg.

Across all five matches, Marseille trail 66–85 in shots and 24–31 on target. Against PSG, the shot count was 12–31. Leaving that performance outside the xG total is a coverage limitation, not evidence that it went well.

What to watch: for Lens, whether they can sustain their attacking threat beyond the opening-day high; for Marseille, whether they can reduce the pressure opponents put on them. Neither team should be labelled unlucky from one partial total.

What would change the assessment after the break?

An early table tells us what teams have collected. The next few games need to tell us whether the way they play is changing.

• For Spurs and Villa: repeated chances, with fewer moments of defensive exposure.
• For Gladbach and Valencia: better protection of their own goal alongside any attacking improvement.
• For Bournemouth: a sustained positive balance of chances, turned into wins more often.
• For Bologna: an attack that gives improved finishing something to work with.
• For Lens and Marseille: performances that make the fuller picture more encouraging than the selective one.

A win would lift every one of these clubs. The more convincing sign of progress would be several matches in which the problem identified here becomes harder to see.

How we assessed the opening phase

This article uses the same 250 completed league matches as the earlier pieces in this series: 50 in the Premier League, 69 in LaLiga, 50 in Serie A, 36 in the Bundesliga and 45 in Ligue 1. The cutoff is the scheduled fixture date of 20 September 2026 inclusive in each league's local timezone. The snapshot was saved on 21 September at 13:59 UTC, or 16:59 in Bucharest. Levante–Athletic Club was postponed and excluded. Cups, European competition, friendlies and internationals are excluded throughout.

The clubs are selected examples of different slow-start patterns, not a ranking of all struggling teams. Gladbach had played four matches; Valencia had played seven. We show the number of games rather than comparing their points totals directly.

Shots and results cover every match for the selected teams. In the original snapshot, paired xG covers four of five games for Tottenham, Villa, Bournemouth and Lens; six of seven for Valencia; three of four for Gladbach; three of five for Marseille; and all five for Bologna. Figures are rounded to two decimals. Any comparison of goals with xG uses the same matches.

The saved xG was supplied through API-Football; its upstream model is unspecified. We have not combined it with another provider's estimates or treated missing values as zero. Other match statistics in Oddsivio may come from SportMonks or legacy API-Football data. These are observations and editorial interpretations, not causal proof or promises of a rebound. Club and league reports are linked beside the relevant passages; comments are attributed rather than treated as independent verdicts. The selected results and statistics were checked against the public Oddsivio API on 25 September. Newly available Valencia–Real Sociedad xG is reported separately, leaving the original series totals intact. Blessin's and Aguirre's appointments are explicitly dated after the shared results cutoff.