The Stat Man's Premier League Matchweek Two

One round of football has been played and the table has already started telling stories: Hull above Manchester United, Brighton top on goal difference, Aston Villa bottom. Every one of those lines is a single match. This page treats the opening weekend as what it is, ten data points laid on top of last season's 380, and makes a call on every matchweek two fixture from the combination. Where a club's evidence comes from the Championship, I say so and attach the asterisk. Where a result and its underlying numbers disagree, I say that too, because that disagreement is usually where the value in the numbers lives.

A note on a new column. From this season the football-data.co.uk results file carries an expected goals figure for each match, so for the first time these pages can quote xG: in plain English, the number of goals a team's chances would produce on average, judged by where and how each shot was taken rather than by whether it went in. One provider, one season, used only for the games it covers. Last season's ledger has no xG in it, so last season's numbers are counting stats and are labelled as such.

Friday

Crystal Palace v Manchester City (20:00)

Palace lost the opening game at Everton by two goals to nil and, on the new column, deserved better: 1.96 expected goals to Everton's 1.12, from 11 shots to 16. That is the shape of a side that finished 15th last season with a shot ledger almost exactly level, 442 taken and 445 faced, and a home record of 4 wins, 9 draws and 6 defeats worth 21 points. Selhurst Park was not a fortress; it was a place where Palace did not lose much and did not win much either.

City's opening win over Bournemouth, 2-1 from 13 shots to 5 and an xG of 2.36 to 0.63, was a routine afternoon for a side that scored 77 last season and took 33 points on the road from a 9-6-4 away record. The pattern in this fixture is the thing to hold onto. City won both meetings by three goals to nil last season, and in the December game at Selhurst Park Palace actually out-shot them 16 to 7 and still lost by three. Volume without accuracy is a recurring Palace theme: 141 shots on target across 38 games last season against 162 faced.

The Stat Man's call: Manchester City. Palace's opening-day xG says they are better than one week's table, but the head-to-head says that has never yet been enough against this opponent.

Saturday

Liverpool v Nottingham Forest (12:30)

Liverpool produced the biggest away performance of the opening round and got a draw for it: 2-2 at Newcastle from 27 shots to 13, 7 on target to 4, and an xG of 2.98 against 1.57. They had the third-highest shot count in the division last season, 588, and Anfield gave them a 10-6-3 home record with 34 scored. The problem last season was elsewhere: 7-3-9 on the road, which is why 588 shots bought fifth place rather than a title challenge.

Forest lost at home to Leeds 1-0 in a game with 0.66 expected goals for them and 0.47 against, a match nobody deserved to win. Their season-long profile has not changed: 4-8-7 at home for 20 points, 7-3-9 away for 24, a club that plays better as the visitor. Then there is the head-to-head, which is the strangest pair of results on this page. Forest won three-nil at Anfield in November from 15 shots to Liverpool's 21, and Liverpool won 1-0 at the City Ground in February from 10 shots to Forest's 18. In both games the side with more shots lost.

The Stat Man's call: Liverpool. The opening-day xG and last season's home column both lean the same way; Forest's Anfield win from 15 shots is the reason the pencil stays sharp.

Bournemouth v Everton (15:00)

Bournemouth's opening day at the Etihad, 5 shots and an xG of 0.63, tells you about the opponent rather than about Bournemouth. Their real identity is last season's 18 draws in 38 games, 10 of them at home inside a 7-10-2 home record. Nobody drew more, and only 7 defeats all season put them sixth. A Bournemouth home game is, on the evidence, the most likely fixture in the division to finish level.

Everton won their opener two-nil against Palace while being out-xG'd 1.96 to 1.12, a result to enjoy rather than to read much into. Their season-long numbers still say a side that is better on the road: 26 away points to 23 at home last season, and 533 shots faced against 420 taken. The head-to-head sharpens the picture. Everton won 1-0 at the Vitality in December, limiting Bournemouth to 5 shots, and Bournemouth won 2-1 on Merseyside in February. The away team won both, which fits both clubs' road-leaning ledgers rather neatly.

The Stat Man's call: a draw. Ten home draws from 19 for the hosts, and two away wins from two in the head-to-head, pull in opposite directions and meet in the middle.

Coventry City v Hull City (15:00)

The first all-promoted meeting of the season, and both clubs arrive with their Championship ledgers doing the talking, second-tier asterisk attached to every figure. Coventry were champions on 95 points, and home was where the title was built: 17 wins, 4 draws and 2 defeats at home, 51 scored and 19 conceded, 55 home points. Their opening day at Arsenal, 4 shots and 1 on target for an xG of 0.2, was a first taste of what happens when the opposition quality jumps.

Hull are the round's headline act after beating Manchester United two-nil, and the numbers say enjoy it carefully: 8 shots to United's 21, an xG of 1.01 to 1.83, a win earned by the goalkeeper and the finishing rather than by control. That is consistent with their promotion season, in which they scored 70 from 515 shots while facing 682 and conceding 66. One specific fact matters here more than any of that. These clubs met twice last season and neither scored: 0-0 at Coventry in August from 18 shots to 7, and 0-0 at Hull in April from 13 shots to 6. Two games, 44 shots, no goals.

The Stat Man's call: Coventry, on a 17-4-2 home record, and a low-scoring one; the two goalless head-to-heads are the loudest number in this fixture.

Tottenham Hotspur v Newcastle United (17:30)

Spurs opened with a three-nil defeat at Brentford in which they faced 26 shots and an xG of 3.88, the worst expected-goals-against figure anyone posted in the round. That extends last season's home problem rather than starting a new one: 3 wins, 6 draws and 10 defeats at home for 15 points, 31 conceded in 19 home games. Spurs were a different club on the road, 7-5-7 for 26 points, but Saturday is at home.

Newcastle drew 2-2 with Liverpool while being out-shot 27 to 13, a point earned against the run of the numbers. Their season-long split is as lopsided as Spurs' but the other way round: 10-2-7 at home with 36 scored, 4-5-10 away for 17 points, one of the weakest away hauls in the division. The head-to-head, though, belongs to Newcastle. They drew 2-2 at home in December from 19 shots to 8, then won 2-1 in north London in February from 21 shots to 16. Unbeaten in both, out-shooting Spurs in both.

The Stat Man's call: Newcastle, in soft pencil. The third-weakest home column in the division meets one of the weakest away columns, and the head-to-head is the only tiebreaker on offer.

Sunday

Chelsea v Brighton (14:00)

Chelsea won at Fulham 3-2 with the better numbers, 18 shots to 14 and an xG of 1.7 to 1.35, which is what their season-long ledger usually delivers: 511 shots taken against 415 faced, the profile of a top-six side that finished tenth. Stamford Bridge was worth exactly what the road was worth last season, 7-5-7 in each, and I have yet to find another club in the division that indifferent to venue.

Brighton posted the second-biggest xG of the opening weekend, 3.77 to Villa's 0.28, from 21 shots to 6, in a four-nil win. Their away column last season was poor, 5-5-9 for 20 points, but this specific fixture is where the poor away column stops mattering. Brighton did the double over Chelsea: 3-1 at Stamford Bridge in September and three-nil at the Amex in April, that second game finishing with 9 shots on target to Chelsea's none. Six goals scored, one conceded, across the two meetings.

The Stat Man's call: Brighton, narrowly. Chelsea's shot ledger is the better long-run number, but the head-to-head and the opening-day xG both point the other way.

Leeds United v Brentford (14:00)

Leeds won 1-0 at Forest with 0.47 expected goals, the lowest xG of any winning side in the round, so the three points say more about the opposition than about Leeds. Elland Road is a different matter: 9-5-5 at home for 32 points last season, against a road record of 2 wins in 19. Leeds' season lived and died at home, and Sunday is at home.

Brentford's three-nil win over Spurs came with the highest xG of the weekend, 3.88, from 26 shots and 8 on target. Set that against last season's ledger, 405 shots taken and 504 faced across 38 games, and the opening day looks like an outlier rather than a new normal. The head-to-head is two draws, a goal apiece at Brentford in December and goalless at Leeds in March, and in both games Leeds had far more of the ball's business: 18 shots to 7, then 14 to 6, for a single goal across the two games.

The Stat Man's call: a draw, leaning Leeds. Two draws, a home side that out-shoots this opponent, and an away side that traditionally faces more shots than it takes.

Sunderland v Fulham (14:00)

Sunderland were last week's regression case and the opening day did nothing to change the file: a 2-1 defeat at Ipswich with an xG of 0.66 against 1.8. Seventh place last season was built on a shooting ledger from the bottom half, 395 shots taken and 542 faced, 126 on target for and 167 against, and a goal difference of minus 6. As I put it in the book, the league table ranks results, not performances, and Sunderland's results ran a long way ahead of their performances.

Fulham lost their opener 2-3 at home to Chelsea, and away from Craven Cottage they were one of the weaker travellers last season, 4-5-10 for 17 points. This opponent, though, brought out something else. Fulham won 1-0 at home in November from 24 shots to Sunderland's 4, then won 3-1 at the Stadium of Light in February. Twenty-four shots to four is the most lopsided count in any head-to-head on this page.

The Stat Man's call: Fulham. A weak away column against a weaker underlying home side, and a double from last season that was earned rather than stolen.

Manchester United v Ipswich Town (16:30)

United lost at Hull from 21 shots without scoring, with an xG of 1.83 to Hull's 1.01. One afternoon of finishing is not a trend, and the season-long ledger is unambiguous: nobody took more shots last season (596) and nobody put more on target (216), inside a 13-3-3 home record with 39 scored. Regression to the mean is the relevant chapter; it works in both directions, and a side that out-shoots an opponent 21 to 8 and loses is usually owed something.

Ipswich are the third promoted club, second in the Championship on 84 points. Their away record of 9-7-7 for 34 points was the weakest road haul of the three clubs who came up, but the shot ledger underneath it was the best of the three by ratio: 693 shots taken against 486 faced, and 47 conceded across 46 games, with only Coventry meaner. Their opening win over Sunderland was earned on the numbers, an xG of 1.8 to 0.66, which is more than Hull could say about theirs. Asterisk applied: those Championship figures came against Championship defences, and there was no top-flight meeting between these clubs last season to lean on.

The Stat Man's call: Manchester United. The opening-day loss reads as variance on top of the strongest shooting ledger in the league; Ipswich's shot ledger makes it a proper contest rather than a procession.

Monday

Aston Villa v Arsenal (20:00)

Villa's opening weekend produced a number I did not expect to write: no shots on target. Six shots, none on target, 0.28 expected goals, a four-nil defeat at Brighton. Last season's fourth place was built on a home record of 12-2-5 for 38 points, and on last season's evidence Villa Park is where the recovery should start. The shot ledger was always the worry, 483 taken and 490 faced across the season, a mid-table profile wearing a Champions League points total.

Arsenal beat Coventry three-nil from 20 shots to 4 and an xG of 1.88 to 0.2, which is how champions who conceded 27 all last season are supposed to open. Their away column was as good as anyone's home column: 11-5-3 for 38 points, 16 conceded in 19 road games. The head-to-head split last season, and it split in a way that matters for Monday. Villa won 2-1 at Villa Park in December with shots level at 15 apiece, then Arsenal won 4-1 at the Emirates later the same month from 22 shots to 11. Home advantage has shrunk over the years, but Villa Park against Arsenal is one place it showed up.

The Stat Man's call: Arsenal. The champions' road record and the opening-day gap between the two clubs' numbers both point one way; the December result and a 12-2-5 home column are why the call is written in pencil.

The Stat Man's takeaway

• Three opening-day winners were out-xG'd: Everton, Hull and Leeds. Those are the results to trust least.
• Manchester United's 21 shots without a goal at Hull is variance sitting on the league's strongest shooting ledger, not a new fact about United.
• Coventry against Hull produced 44 shots and no goals across two Championship meetings; a low score is the numbers' expectation.
• Brighton have scored six and conceded one in their last two meetings with Chelsea, and opened with an xG of 3.77; the away column is the only thing arguing against them.
• Villa's zero shots on target at Brighton is the single worst attacking line of the round; a 12-2-5 home record is what they have to set against it.
• Ten calls, all in pencil. One round of xG is a hint, not a verdict.

Sources and method. Last season's figures are full-season counting stats from the football-data.co.uk 2025-26 results files for the Premier League and Championship, aggregated on our transparency pages (Premier League 2025-26, Championship 2025-26); the opening-round figures come from the 2026-27 file, aggregated at Premier League 2026-27 after matchweek one. Expected-goals figures are the per-match xG columns published in that 2026-27 file, one provider, one season; no xG appears for 2025-26 because that file carries none. Data pulled 25 August 2026. Championship figures carry the lower-division asterisk throughout. The calls above are readings of the numbers, not betting advice; the numbers describe what has happened and do not promise what will.

Sources: BBC Sport (fixtures); football-data.co.uk (2025-26 Premier League and Championship results data; 2026-27 Premier League results and xG data).

18+. This page contains no betting advice and no selections. If gambling stops being fun: GamCare 0808 8020 133, or self-exclude via GamStop.