Analisi della Premier League

Tottenham’s 0–0 draws with Nottingham Forest and Everton should not be treated as the same low-scoring matchup. At Forest, Spurs produced 1.57 xG, but 0.82 came from Micky van de Ven’s missed corner chance in the 55th minute. That single attempt supplied more than half their chance value; none of their 12 shots reached the target. Against Everton, the restriction was broader: eight of 14 attempts were blocked, the entire attack generated only 0.64 xG, and the two efforts on target carried approximately 0.08 xG between them. Spurs had more penalty-area touches against Everton than Forest, 37 versus 30, yet much less chance value. Reaching the box was therefore not the same as finding an unobstructed finish.

The 2–3 Villa match offers a different development, but not proof that the problem has disappeared. Tottenham’s two goals came in the 86th minute and at 90+8, from Conor Gallagher’s assisted finish and Jan Paul van Hecke’s set-piece chance, each worth about 0.35 xG. Villa added six shots on target and three goals from 1.16 xG, making the opposing contribution and efficient finishing central to the five-goal total. Across the five recorded league games, only Brentford and Villa matches reached three goals. A defence that repeatedly blocks Spurs before clean contact, as Everton did, supports a lower Tottenham scoring contribution. A match producing valuable unblocked chances at both ends has a different ceiling; neither the Forest blank nor Villa’s late goals should be projected from the score alone.

Villa’s 4–0 defeat showed Brighton carrying the total through clear first-half chances, including three scoring attempts worth 0.96, 0.87 and 0.44 xG, plus an own goal. At Coventry, the 5–0 win instead accelerated after halftime: a Yalcouyé chance worth 0.40 xG and a Groß penalty were followed by two goals from attempts worth only about 0.03 and 0.10. Coventry were down to ten men from the 53rd minute, when already trailing 2–0, so the late expansion also had a numerical-advantage condition. Chelsea’s 4–3 win involved dangerous chances on both sides, but Brighton’s three also included a João Pedro own goal. These were not three identical demonstrations of a seven- or five-goal attacking baseline.

Leeds provides the strongest counterexample to reading the final score as chance quality. That 1–1 draw had 4.67 combined xG, more than the Chelsea match’s 4.34. Leeds created large early opportunities; Brighton later had a 0.35-xG effort saved and missed chances worth 0.42 and 0.55. Last season’s home Villa fixture also reached a high total, but Villa contributed 2.41 xG then compared with 0.29 this time. Another four-goal game needs either Brighton to carry a substantial supply of valuable chances, as against Villa, or meaningful danger at both ends. Penalties, own goals and difficult finishes can lift the score, but should not be assumed to recur.

Manchester United’s 5–2 win over Ipswich contained enough chances to support a high total: 6.57 combined xG, with United creating heavily after halftime and Ipswich still reaching a late 0.97-xG chance. The four goals at Everton had a different foundation. Combined xG was just 1.70, and three of the four scoring shots were each worth less than 0.04 xG. That total depended on exceptional conversion, including two goals after the 87th minute. Treating both games as the same high-scoring trend would hide the difference between repeatable chance supply and a finishing spike.

The low totals need the same scrutiny. Fulham’s 1–1 draw contained 3.22 combined xG, yet required an own goal and an 89th-minute Cunha finish to produce its two goals. United’s 0–2 loss at Hull also contained 2.84 combined xG and several missed United opportunities. The derby was more restrictive, at 2.10 combined xG, despite United playing against ten men from the 23rd minute, with the teams managing only eight shots inside the box between them. Last season’s Fulham visit finished 1–1 from 3.39 combined xG, another reminder that a familiar scoreline need not describe a low-chance matchup. A future opponent permitting sustained good chances supports a higher total; Everton’s difficult finishes are a much weaker reason to carry four goals forward.

Tottenham’s latest visit to Nottingham Forest produced 12 match corners, twice the six in the same away fixture last season, and the change came overwhelmingly from Spurs: their contribution rose from three to ten while Forest’s fell from three to two. Tottenham’s final-third entries moved only from 48 to 52, but their penalty-area touches increased from 13 to 30 and shots from six to 12. The difference was what Spurs achieved after entering the attacking third, rather than a dramatic increase in how often they crossed into it.

That deeper access did not become a goal. Four attempts were blocked and eight missed the target; the best opening was van de Ven’s 0.82-xG chance following a corner. Spurs still won five corners in each half, including a 5–0 second-half count, so the pressure continued while finishing failed. Forest’s own box touches fell from 24 in the prior meeting to ten, reducing the opposing contribution. This matched-venue comparison explains how Tottenham can carry a double-figure total even against a quiet opponent, but it requires attacks to keep reaching the box. It is a different condition from merely holding possession outside the defensive line, as Spurs did during their cornerless first half at Brentford.

With City down to ten men after Foden’s 23rd-minute dismissal, the 13-corner Manchester derby developed through United’s second-half response: United took five corners after the interval while City added three. United also moved from five shots before halftime to 11 afterwards, with 63% second-half possession. Bournemouth and Sunderland each supplied enough alongside City to produce 11 corners. Palace did not: City took six, but Palace’s two first-half corners were their entire contribution. Palace managed only two second-half shots as City held 77% of the ball, leaving the match at eight corners despite a 4–1 scoreline.

Coventry demonstrates the opposite failure mode. Their five corners were enough to contribute to a high total, but City stopped at four, collecting none after halftime despite 71% possession. Only one of City’s six second-half shots was blocked; four were off target. Coventry’s increasing attacks also forced City into 23 clearances. The combined total stalled at nine because City’s own corner production disappeared. Last season’s away derby finished with seven corners, including just one for United, compared with six for United this time. For City matches, neither possession nor goals identifies the full-time corner environment: the key is whether both sides can keep producing corners after the interval, or whether one side can compensate when the other drops out.

Sunderland won the first-half corner count 4–0 at Ipswich, yet the total stayed below five. They also led Fulham 3–1, another four-corner half. Their own early advantage was therefore insufficient when the other side contributed little. Brentford supplied three alongside Sunderland’s three, and City supplied two alongside Sunderland’s three; those were the halves that reached six and five. The difference was the combined supply, not simply whether Sunderland were on top in corners.

The attacking conditions were not identical. Sunderland made 39 first-half final-third entries at Brentford and 30 at City, but only 11 against Arsenal, where the half finished 2–1 in corners. Ipswich still allowed Sunderland eight early shots and four corners without producing a corner themselves. Last season’s away City fixture had the same two City corners but only one for Sunderland, leaving the half at three. A future five-corner half is better supported when Sunderland retain an active attacking contribution and the opponent adds to it. A Sunderland corner lead alone can coexist with a low combined count.

Leeds supplied four of six first-half corners at Brighton and four of five against Palace. The Palace half involved nine Leeds shots, five blocked, and a 38–28 advantage in final-third entries. Against Newcastle, Leeds supplied three of five while forcing 25 clearances and allowing only one Newcastle attempt. These successes had a meaningful Leeds contribution; the total did not require the opponent to carry the whole count.

Brentford was the reverse. Leeds took no first-half corners while defending 11 attempts and making 27 clearances; Brentford’s four still left the half short of five. Forest shows that Leeds activity alone is not enough either: all four Leeds first-half attempts were blocked, but they won only one corner and the half finished at two. Last season’s Brighton visit had three Leeds corners and one for Brighton, compared with four and two this time. Small changes in both contributions moved the total across the threshold. For another five-corner opening, Leeds need attacks that actually generate corners, plus enough opposing supply; blocked shots and defensive work are supporting context rather than substitutes for that contribution.

Ipswich’s 1–1 halftime score against Sunderland combined Emersonn finishing a 0.77-xG fast break with Angulo scoring a much harder direct free kick. At United, Davis scored and Fernandes replied on a break; Ipswich also forced saves from two opportunities above 0.30 xG before halftime. Those games contained genuine early routes for Ipswich themselves. Palace’s three-goal half added another condition: Ipswich’s second arrived after Disasi’s 34th-minute dismissal, although Emersonn had equalised before it.

Liverpool carried the early scoring alone, with Isak scoring in minutes six and nine from chances worth about 0.08 and 0.14 xG. Liverpool’s whole first half produced only 0.43 xG, so the two-goal result depended on efficient finishing rather than repeated clear chances. Everton was the opposite exception: 2.16 combined first-half xG produced only one goal, with Johnson and Davis missing valuable opportunities. Another two-goal half is better supported by transition access and genuine chances on both sides than by assuming Liverpool’s early accuracy will recur. Ipswich can participate in that scoring, but the successful results do not all describe the same attacking balance.

At Fulham, Palace’s 2–1 halftime deficit came with Fulham generating 1.72 xG and Palace only 0.28. King scored from a fast break and Palacios from a substantial 0.33-xG chance, while Palace’s reply was Mitchell finishing a much harder break worth about 0.06. Against Ipswich, Palace again trailed 2–1: Khalaili converted a roughly 0.04-xG chance, whereas Ipswich’s goals came from opportunities worth 0.36 and 0.49. Disasi’s 34th-minute dismissal preceded the second Ipswich goal, adding a condition that should not be treated as normal eleven-a-side play.

Everton provides the clearest failed high-total opportunity. Palace created 1.57 first-half xG, including three significant chances for Mateta and Nketiah, but did not score; Everton’s difficult finish was the half’s only goal. City also generated 1.73 xG before halftime yet led by just one, while Leeds–Palace had only 0.72 combined xG and stayed goalless. A future two-goal half is more credible when the opponent can repeatedly reach good positions and Palace retain a scoring route of their own. The Fulham and Ipswich scores also required Palace to convert difficult chances, so they should not be read as automatic early shootouts.

Yéremy Pino’s 38 passes against Fulham came in 89 minutes with Crystal Palace holding only 35% possession. Against Leeds he attempted 35 in 82 minutes despite Palace having 49%. His opposition-half passes were similar, at 21 and 23, while own-half passes were higher against Fulham, 17 versus 12. The lower-possession match still supported the larger total because Pino played longer and contributed more to exchanges deeper in Palace’s half.

The short appearances need separate treatment. He attempted 19 passes in 35 minutes against Everton, while his Ipswich outing lasted 37 minutes and produced 15. Palace’s dismissal after 34 minutes against Ipswich changed the team’s match, but most of Pino’s recorded playing time preceded it; the full team’s subsequent numerical disadvantage cannot simply explain his entire count. Against Manchester City he played 68 minutes and attempted 24 passes, with 13 in his own half and 11 in City’s. That balance suggests a different route from the longer Fulham and Leeds appearances, where more advanced exchanges were available. A useful expectation for Pino starts with his role and likely duration, then considers where Palace can involve him. The team’s possession share alone misses both.

Pascal Groß played 90 minutes in each of these Brighton appearances, so the range from 32 passes against Arsenal to 93 against Coventry cannot be explained by extra playing time. Against Arsenal, Brighton made 343 passes with 40% possession and Groß attempted 20 in the opposition half. Coventry allowed a much larger platform: 584 team passes, 69% possession and 61 Groß passes in the opposition half. Coventry also lost a player after 53 minutes, so that 93-pass ceiling includes a substantial spell against ten men.

The comparison with Chelsea shows that team possession still does not settle Groß’s individual total. Brighton made 647 passes against both Chelsea and Villa, yet Groß finished with 63 against Chelsea and 77 against Villa. His opposition-half passes rose from 36 to 53, while his own-half total actually fell from 27 to 24. Villa’s dismissal after 40 minutes is material context for the higher figure. Coventry and Villa therefore provide evidence of a high-volume role when Brighton can sustain advanced exchanges, but neither is a clean eleven-against-eleven benchmark. Leeds offers another route: Groß attempted only 47 passes but took five shots, showing that attacking involvement can end in attempts instead of another pass.

Oscar Bobb’s 24 passes against Crystal Palace came in 79 minutes while Fulham made 672 passes and held 65% possession. Against Manchester United he attempted 25 in 71 minutes, although Fulham’s total fell to 457 and possession to 42%. The bigger team passing pool did not generate a bigger individual total. Bobb remained a relatively small part of the circulation, and his recorded appearances of 63–79 minutes also limit how much time he has to accumulate passes.

Liverpool was the clearer example of his passing supply being restricted. Bobb finished with 16 passes in 63 minutes, including only nine in the opposition half. Against Chelsea and Sunderland he reached 25 in 67 and 68 minutes, with 21 and 19 opposition-half passes respectively. Those similar playing times make the loss of advanced involvement more informative than the Liverpool total alone. Palace also shows why passing and attacking output should be separated: his 24 passes accompanied three shots and four key passes, much more direct attacking activity than the 25-pass United appearance. A more productive final-third game can improve his chances of creating or shooting without producing a higher pass count.

João Pedro’s early scoring against Fulham and Brighton came through the same useful route: a fast break delivering a valuable chance. He scored after one minute at Fulham from an opportunity worth 0.48 xG and in the 32nd minute against Brighton from one worth 0.46 xG. Brighton also allowed a saved corner attempt and a missed 0.40-xG assisted chance before halftime, so his two early targets were supported by several ways of reaching a shooting position. Chelsea had only 39% and 30% first-half possession in those games. The player opportunity came from what Chelsea did when they attacked, rather than from controlling the ball for most of the half.

Arsenal and Hull restricted the volume despite João Pedro again playing 90 minutes. At Arsenal he managed one 0.018-xG attempt and no target. Against Hull his only shot was the assisted chance he scored in the 66th minute, worth 0.29 xG. He had 42 touches against Hull, close to 43 against Brighton, but five attempts became one: general involvement was not translating into the same access to shooting situations. Across these four recorded appearances, the stronger early-shot condition is an opponent allowing Chelsea to release him into a valuable transition chance or find him repeatedly near goal. Hull demonstrates that one well-taken later chance can produce a goal without recreating either the early targets or the larger shot count.

Chelsea and Newcastle both scored against Hull, but Hull’s replies came in different circumstances. Belloumi scored twice before halftime at Chelsea, including a corner finish, and Hull later created better chances for McBurnie and Cho that were saved. At Newcastle, Hull were two goals down after seven minutes and produced 19 second-half attempts, eventually scoring through Cho’s 0.27-xG chance. The latter is evidence of a response while chasing, not proof that Hull will attack at that rate from kickoff.

The clean sheets do not erase opposition threat. United created 1.83 xG without scoring; Coventry had Awoniyi’s 0.72-xG chance saved; Villa missed valuable McGinn and Buendía openings. Hull’s own scoring can be the tighter condition: they generated only 0.58 xG against Villa and needed an 82nd-minute fast break to beat Coventry. Their United goals came from restarts before an almost inactive second-half attack. For another game with goals at both ends, a dangerous opponent alone is insufficient. Hull need a credible set-piece, transition or sustained chasing route of their own, while the clean-sheet history should be assessed against the chances opponents actually received.

Crystal Palace led Ipswich 5–0 on corners at halftime and won the final count 6–3, despite playing with ten men from Disasi’s 34th-minute dismissal. The advantage was largely banked in the opening half; after halftime, Palace added one corner while Ipswich took three. It is evidence that an early lead can survive reduced attacking access, but not that Palace maintained a dominant corner rate throughout the match.

The other games show the difficulty of reproducing a positive margin without that opening contribution. Palace took no corners at Everton despite 58% possession, lost 2–6 against City as their second-half attack faded, and lost 1–7 at Fulham while the hosts made 63 final-third entries to Palace’s 33. Leeds was closer at 3–4, with Palace recovering from a 1–4 halftime deficit after Leeds stopped adding corners. Last season Palace won corners 4–2 at Everton, whereas this season the same venue produced 0–4. The opponent label or possession forecast therefore does not establish the margin. Palace need a reliable source of their own corners, and a large early advantage is vulnerable if the opponent sustains pressure while Palace’s supply dries up.

Newcastle’s fifth-minute opener against Liverpool came from an Elanga fast break. Newcastle finished that half with only 0.36 xG and one shot on target, while Liverpool made more final-third entries and took more shots. Hull was a faster Newcastle start in a different setting: Willock scored in the third minute and Hall added a fast-break goal in the seventh. These successes show that early decisive actions can secure the opening goal before sustained dominance develops.

Tottenham illustrates why scoring first is not the same as scoring early. Newcastle had no first-half shot on target there and did not break through until Elanga’s 62nd-minute goal. Bournemouth and Leeds exposed the opposite risk: Newcastle allowed 13 and 11 first-half attempts while offering only three and one themselves, and conceded first in both. Last season’s home Bournemouth game also saw the visitors lead at halftime after creating substantially better chances. For the next opening goal, Newcastle need either an immediate break into a useful shooting position or enough early access to avoid being pinned back. Their ability to finish first against Spurs should not be converted into an expectation of a quick goal.

Erling Haaland took no first-half shot at Manchester United, where Foden’s 23rd-minute dismissal left City with ten men, then scored from a 0.44-xG chance in the 60th minute and missed a 0.42-xG set-piece opportunity in the 76th. Those were his only two attempts. Sunderland also kept his first-half efforts off target, but an assisted chance worth 0.67 xG brought his goal in the 81st minute. In both matches, the scoring route depended on a valuable opportunity arriving after halftime rather than a steady stream of accurate shots throughout the game. Haaland played 90 minutes in all five recorded league appearances, preserving access to those late opportunities.

Coventry supplied a different timing pattern: all three of Haaland’s shots on target arrived before halftime, including a goal and a saved 0.46-xG chance. Bournemouth allowed five attempts but saved his best opening, worth 0.49 xG in the 75th minute, and he finished without a goal. Volume and timing therefore need a chance-quality check. Last season’s away derby also produced only two attempts, but just 0.13 xG in 80 minutes; the current two-shot return carried 0.87 xG. The same opponent and shot count concealed much better scoring positions this time. A match restricting early service can still offer a late Haaland goal route if City eventually find him in valuable positions, but an early shot or target expectation needs the sustained first-half supply seen against Coventry. His full-match scoring record cannot answer both questions equally.

Virgil van Dijk’s 114 passes against Forest came from more than routine exchanges near his own goal. Liverpool made 631 passes with 69% possession, and Van Dijk attempted 69 in the opposition half alongside 45 in his own. Against Newcastle he played the same 90 minutes but attempted only 58, split 30 in Liverpool’s half and 28 in Newcastle’s. Forest provided both a bigger team passing pool and much greater scope for him to join advanced circulation.

His 90-pass Ipswich match had a different balance: 50 passes in his own half and 40 in the opposition half. Liverpool’s possession was only 55%, below the Newcastle figure of 61%, yet Van Dijk still accumulated far more passes. Deep buildup involvement can therefore support a substantial total even without exceptional territorial dominance. Fulham and Bournemouth show the same distinction within almost identical headline totals. He attempted 78 and 79 passes respectively, but the Bournemouth match shifted his own-half contribution from 34 to 46 while opposition-half passes fell from 44 to 33. For his next matchup, the question is whether Liverpool can keep using him in either phase. A loss of advanced involvement need not collapse his total if more buildup work returns through him.

Arsenal’s opener against Coventry came in the 15th minute through Havertz, followed by a Saka fast-break goal eight minutes later. Coventry created just 0.04 first-half xG while Arsenal produced 1.32, a clear example of early attacking imbalance. City’s opener was less one-sided in its sequence: Coventry had already missed a 0.31-xG Awoniyi chance and seen a 0.29-xG Rudoni attempt saved before Haaland scored in the 26th minute. Coventry had a route to lead, but failed to use it.

Hull was a stronger version of that missed opportunity. Coventry generated 1.09 first-half xG to Hull’s 0.01, including Awoniyi’s saved 0.72-xG effort, then conceded the only goal to a Millar fast break in the 82nd minute. Brighton scored first in a half where each side produced 0.97 xG; Awoniyi’s later red card came at 0–2 and cannot explain the opener. Against Forest, Coventry instead held out through a high volume of largely blocked attempts before DaSilva scored. The risk of conceding first is strongest when Coventry are genuinely denied an attacking route, as at Arsenal. Where they create the better early chances, conversion and transition protection matter more than the recent first-concession count.

Newcastle scored twice at Tottenham from just 0.72 xG, with Elanga finishing a 0.14-xG chance and Wissa scoring from a corner worth 0.22. Against Bournemouth, Barnes’s first was roughly 0.05 xG and Ramsey’s late set-piece equaliser 0.11; Newcastle generated only 0.71 xG overall. Hull’s two-goal Newcastle return also needs qualification: the goals arrived in minutes three and seven from chances worth 0.14 and 0.06, while much of the later xG came from Wissa’s saved penalty. The final totals make the scoring look more routine than the goal routes do.

Liverpool supplied stronger non-penalty opportunities, including Wissa chances worth 0.51 and 0.41 xG, although Newcastle’s actual goals came from lower-value fast breaks. Leeds then limited Newcastle to 0.85 xG and one late consolation. Last season’s Tottenham visit produced two Newcastle goals from 2.48 xG, far more substantial supply than this season’s same scoring return. Another brace is more convincing when Newcastle repeatedly reach valuable opportunities beyond a penalty or a handful of precise finishes. The Liverpool chance supply offers that route; the Tottenham and Bournemouth results require greater caution about repeating the conversion.

Bukayo Saka put three attempts on target against both Chelsea and Sunderland, but the matching totals conceal different scoring opportunities. Chelsea kept all four of his shots below 0.07 xG each: he could test the goalkeeper repeatedly without reaching a high-value finish. Sunderland allowed a 0.51-xG chance in the 13th minute, then a smaller fast-break chance and a stoppage-time penalty. The penalty supplied about 0.79 of his 1.45 xG and his only goal, so neither that goal nor the full xG total describes an open-play breakthrough. At Villa, Saka needed only two attempts to score because the decisive assisted chance was worth 0.57 xG—more than twice his entire Chelsea shooting sample.

Timing also matters for interpreting the volume. Saka had one first-half effort on target in each of the Chelsea and Sunderland games; the additional targets came after halftime, with Sunderland’s third arriving from the penalty at 90+7. He played 89 and 90 minutes respectively. Against Brighton, his only target was a 0.29-xG fast-break chance before halftime, and he left after 81 minutes with two shots overall. The stronger condition for Saka scoring is access to a valuable assisted or transition chance, as at Villa or in his 0.52-xG fast-break goal against Coventry. A Chelsea-like opponent permitting several low-value efforts can support shot and target volume without offering the same goal threat. These five matches do not establish that opponent ranking alone explains the difference.