Premier League Analyse

Manchester United’s five against Ipswich came with a genuine surge in chance creation: 22 second-half shots, 15 inside the box and 3.82 xG. Even that score included a Jacob Greaves own goal and a Bruno Fernandes penalty, but United also created high-value chances from open play, including Fernandes’s 0.63-xG goal and Mbeumo’s 0.45-xG fast-break finish. Everton’s 2–2 draw was much less convincing support for another United brace. United produced only 0.69 xG all game; Mbeumo scored a chance worth roughly 0.02 xG, and Šeško’s second arrived in the 88th minute from a 0.25-xG opportunity.

Hull and Fulham explain why high attempt totals are insufficient. United took 21 shots without scoring at Hull, then 29 while scoring once at Fulham; eight and 13 attempts respectively were blocked. At Fulham, their only goal was Cunha’s late 0.02-xG finish, rather than the reward for a sequence of clear chances. Last season’s away Everton game produced more United xG, 1.27, but only one goal. The useful target is an opponent allowing repeated valuable chances after United progress into the box, as Ipswich did. Everton-style finishing can deliver two, but it leaves the outcome dependent on converting opportunities that usually do not score.

Manchester City’s 8–3 corner win over Bournemouth and 6–2 win at Crystal Palace came with sustained territorial advantages: City made 63 and 73 final-third entries while their opponents managed 41 and 36. Those were the clear-margin successes. The 7–6 derby win and 6–5 Sunderland win were different: United and Sunderland both entered the final third more often than City. City still collected corners, but the opposition kept enough attacks alive to make the margin fragile. The derby needs an additional qualification: Foden was sent off in the 23rd minute, so United’s attacking response came largely against ten men. Winning the match did not mean shutting down the other side’s corner supply.

Coventry supplies the strongest warning against using possession alone. City had 78% of the ball but lost corners 4–5. After halftime, Coventry increased its shots from four to eight and entries from 18 to 24; City made 23 clearances and collected no further corners. Last season’s visit to United also separates the two questions: City won corners 6–1 despite losing the game 2–0, whereas this season’s winning scoreline accompanied only a one-corner edge. A large City corner margin needs evidence that the opponent’s exits are being cut off, rather than merely that City will have more possession or the better scoring chances.

Liverpool’s 2–2 at Newcastle contained substantial pressure—27 attempts, 16 blocked shots and 2.98 xG—but their second goal still required a penalty at 90+9. Two valuable Alexander Isak attempts in the 47th minute, worth 0.35 and 0.48 xG, were blocked and saved. Cody Gakpo’s equaliser came from a much smaller 0.036-xG chance before Newcastle went back in front. The large chance total described real opportunities, yet it did not translate into a comfortable two-goal return from open play.

Forest offered a different route to the same Liverpool score. Isak converted a 0.84-xG assisted chance in the 60th minute, then Víctor Muñoz scored from a 0.065-xG attempt in the 82nd. One clear chance and one efficient finish produced two goals from 1.75 xG. At Bournemouth, Liverpool created 1.65 xG and Isak scored from a 0.56-xG chance, but the return stopped at one. The useful distinction is therefore repeated access to decisive finishing opportunities, not simply a rising shot count or a total near two xG. A defence blocking many attempts can keep the second goal dependent on a late exceptional event; a match that repeatedly releases Liverpool’s finishers into clear chances provides a stronger route. Newcastle’s stoppage-time penalty should not be treated as part of the normal attacking supply.

Chelsea won the corner count 10–4 against Hull and 12–4 at Brentford, but the two advantages reveal why corners cannot stand in for scoring quality. Hull faced 72 Chelsea final-third entries and made 32 clearances. Chelsea won six corners before halftime despite taking only three shots, and their full-match chance value stayed at 0.89 xG. At Brentford the second-half pattern was more explicit: Chelsea had 66% possession, nine corners and eight shots, but six attempts were blocked and none reached the target. Brentford made 24 clearances in that half. Chelsea kept returning attacks to the defensive line without turning that pressure into an effective finish.

The opposing contribution stayed at four corners in each match, allowing Chelsea’s accumulation to become margins of six and eight. That is a different environment from Fulham and Brighton, where Chelsea supplied four corners while the opponents added six and seven. Last season’s visit to Brentford had also brought seven Chelsea blocked attempts, but only a 6–5 corner win; the same blocked-shot count did not produce the same margin. The useful future condition is sustained Chelsea pressure combined with limited opposition corner production. A low Chelsea goal return does not rule that out, but neither a high possession share nor a count of blocked attempts is enough on its own to justify a large corner advantage.

Arsenal’s clean sheets against Coventry, Villa and Sunderland do not all describe the same defensive strength. Coventry reached just three penalty-area touches and one shot inside the box, while Villa produced 0.33 xG and no shot on target. Those were opponents struggling to turn possession into worthwhile finishes. Sunderland’s blank was different: their attacking output rose from 0.13 xG before halftime to 1.51 after it. Enzo Le Fée’s penalty was saved in the 56th minute, and Nordi Mukiele’s 0.34-xG corner chance was saved in the 84th. Arsenal prevented the goals, but the chances had already breached the first defensive barrier.

Only the Chelsea meeting had both teams score across the five recorded league games. That low frequency still needs context at both ends. Chelsea scored from a Rogers set-piece attempt worth only 0.017 xG despite creating just 0.39 overall; Brighton scored three, but Arsenal failed to contribute. Arsenal had 36 box touches at Brighton yet only two shots on target, with five attempts blocked and four off target. For another matchup, Coventry- or Villa-like restriction supports an opponent blank more strongly than Sunderland-like dependence on saving a penalty and a dangerous header. But a credible opposing scoring route alone is insufficient for both teams to score: Arsenal must also turn their own box access into clean, accurate attempts.

Cody Gakpo had 58 touches at both Newcastle and Ipswich, playing 90 minutes in each, but attempted 28 passes at Newcastle and 38 at Ipswich. At Newcastle he took six shots and scored; at Ipswich he took one shot. The higher passing total also included ten own-half passes rather than three, adding an extra layer of involvement away from the final attacking action. The same touch count can therefore produce a different passing workload depending on what Gakpo does with those possessions.

Forest and Fulham need a minutes adjustment: Gakpo’s 23 passes came in 71 minutes against Forest, while nine came in a 30-minute appearance against Fulham. Bournemouth was a full-match counterexample to assuming more time fixes a low count: he attempted only 26, with 20 beyond halfway. For another higher total, the Ipswich pattern offers the useful condition—advanced involvement maintained while deeper exchanges add to it. A game using Gakpo primarily to receive and shoot can be attractive for his attacking output without delivering the same number of passes.

Dominik Szoboszlai’s 80 passes against Forest combined 30 in his own half with 50 in the opposition half. At Ipswich, his deeper contribution was identical at 30, but advanced passing fell to 31 and his total to 61. Both appearances lasted 90 minutes. The extra Forest workload came from repeated involvement beyond halfway, not extra time or simply receiving the ball deeper.

Fulham brought 51 passes despite four chances created, twice his Forest chance-creation count. Creative actions can therefore coexist with fewer repeated exchanges. At Bournemouth, both parts of his workload fell—13 own-half and 25 opposition-half passes—leaving 38 in 81 minutes. The nine fewer minutes are relevant but do not by themselves explain the drop from the Forest level; his touches fell from 117 to 54. For another high Szoboszlai count, Liverpool need both substantial circulation and repeated use of him across the buildup and attacking phases. A few incisive final passes, or possession concentrated away from him, need not produce the same accumulation.

Bryan Mbeumo took five shots and scored against Ipswich while attempting only 25 passes. At Hull he attempted 44 passes without scoring, and at Fulham 38, again without a goal. All three appearances lasted 90 minutes. The higher passing totals reflected more repeated touches—64 at Hull and 61 at Fulham, compared with 43 against Ipswich—rather than extra playing time. His presence at the end of an attack is therefore a different role from being repeatedly used to connect it.

His deeper passing barely moved: seven own-half passes at Hull, Everton and Fulham. The variation was beyond halfway, where his totals were 37, 16 and 31. United’s larger overall circulation at Hull and Fulham created room for that involvement, but he still needed a share of it; 60% possession against Ipswich did not lift him beyond 25 passes. Last season’s away Fulham match brought just 16 passes in 90 minutes, compared with 38 this time. A higher Mbeumo count is more credible when United keep involving him in the buildup to repeated attacks, rather than reaching him mainly for a shot or final action. A promising goalscoring matchup can leave his passing modest.

Phil Foden’s eight passes against Manchester United belong to a 23-minute appearance ended by his own red card. His 18 against Coventry came in only 24 minutes. These are not comparable to a normal-length appearance, and combining them with his longer outings would make his passing role look smaller than the evidence supports. The reasons for limited minutes also differ: the United dismissal is a documented interruption, while Coventry is simply a short recorded appearance and should not be treated as evidence of the same problem.

In the longer appearances, Foden attempted 54 passes in 82 minutes against Bournemouth and 48 in 90 against Crystal Palace. City’s team passing total increased from 658 to 788 between those matches, yet Foden’s own count fell. His opposition-half passes dropped from 40 to 35 and his own-half passes from 14 to 13. That is a change in personal involvement rather than a lack of City possession. A substantial appearance gives him access to a much higher total than the eight-pass derby, but City dominating the ball does not guarantee he becomes the route through which it moves. Selection, eligibility and expected playing time need to be established before applying either the low or high examples to another fixture.

Marcus Rashford attempted 24 passes in 45 minutes at Hull, including 20 beyond halfway. Against Ipswich he needed 80 minutes to reach 31, with the same 20 advanced passes and the additional work coming deeper. The Hull appearance therefore contained much more frequent advanced circulation, even though the raw total was lower. Treating those totals without minutes would conceal the stronger rate of involvement.

Fulham brought 35 passes in 77 minutes, including 28 beyond halfway and four chances created. The derby produced only 27 in 83 minutes despite United playing against ten men from the 23rd minute. Rashford had 54 touches there, close to Fulham’s 56, but fewer became passes. Everton’s 26 in 70 minutes adds another shorter appearance rather than a full-match baseline. For a larger count, Rashford needs both enough time and a role involving repeated exchanges, as at Fulham. An extra player, plenty of touches or simply more minutes does not guarantee the ball will keep moving through him in the same way.

Morgan Rogers attempted 42 passes at Brentford, with 36 beyond halfway, despite Chelsea losing 3–0. Against Hull he scored but attempted only 23, including 19 in the opposition half, although Chelsea had more possession and more team passes. Both appearances lasted 90 minutes. The larger Brentford total came from more repeated personal involvement—66 touches rather than 43—not from a better attacking result.

Fulham also brought 23 passes, but with five chances created in 82 minutes, while Brighton brought four chances created from only 16 passes over 90. Rogers can therefore deliver final actions without accumulating a large count. Brighton’s restriction of Chelsea to 26% possession reduced the passing pool; Hull shows that a large pool can still be directed elsewhere. For another high total, the useful condition is Chelsea repeatedly connecting attacks through Rogers beyond halfway. A matchup offering a goal or several chances created may use him more selectively and leave his passing much lower.

Cole Palmer attempted 47 passes at Arsenal with Chelsea holding 45% possession, but only 34 against Hull with 68%, and 30 at Brentford with 60%. All three were 90-minute appearances. The Arsenal workload was concentrated beyond halfway: 38 passes there, compared with 28 against Hull and 23 at Brentford. Chelsea having more of the ball did not automatically direct more of it through Palmer. At Arsenal he took a substantially larger role in circulating it.

The contrast with Fulham shows why touches alone also need interpretation. Palmer had 62 touches in 83 minutes there, more than his 55 at Arsenal, yet attempted only 34 passes while taking five shots and creating three chances. At Arsenal, 47 of his 55 touches were accompanied by pass attempts, with three shots and one chance created. He could be busy as an attacker without repeatedly returning the ball into combinations. Brighton’s restriction of Chelsea to 26% possession then lowered the whole opportunity pool and Palmer finished on 23 passes. For another high count, the important condition is not simply a favourable possession forecast: Palmer needs frequent involvement that ends in another pass rather than a carry, shot or lost attack.

Sandro Tonali played 90 minutes against both Forest and Everton, but his passing total more than doubled from 41 to 88. Tottenham’s possession changed much less, from 59% to 62%. The difference was where and how often Tonali participated: his own-half passes moved only from 22 to 24, while his opposition-half passes climbed from 19 to 64. Spurs’ team passing total also rose from 443 to 643. The Everton performance combined a larger passing pool with much more of the advanced circulation involving Tonali.

Villa gives a useful intermediate example. Tottenham held 64% possession, slightly more than against Everton, but Tonali attempted 65 passes in 84 minutes, including 49 in the opposition half. That is still substantial advanced involvement, without reproducing the Everton ceiling. His 63 passes against Brentford included 43 in the opposition half, whereas the Newcastle appearance was shorter at 75 minutes and produced 36 overall. These distinctions matter because a possession percentage alone cannot tell whether Spurs are exchanging the ball through Tonali or elsewhere. The stronger case for another high total is a match that lets him remain involved in repeated advanced phases; possession that leaves him deeper can produce a much smaller count even across a full game.

Omar Marmoush’s passing totals depend on his own involvement in Tottenham’s advanced exchanges, rather than how much of the ball Spurs hold. He attempted only 13 passes in 77 minutes against Everton despite Tottenham making 643 passes with 62% possession. Against Newcastle, a smaller team total of 498 still produced 19 Marmoush passes and three key passes in 90 minutes. Spurs having more possession therefore did not automatically put more of their circulation through him.

The Villa match shows what a stronger personal passing performance actually required. Marmoush reached 23 passes, with 18 in the opposition half, compared with 12 passes and eight in the opposition half against Forest. His own-half contribution barely changed, from four to five. The increase came almost entirely from more advanced involvement, with six additional minutes also available against Villa. Even then, 23 passes was a modest share of Tottenham’s 581. This is a player whose attacking contributions can arrive without a large passing workload: his three shots against Forest matched his Newcastle total despite seven fewer passes. A possession-heavy Spurs matchup alone is consequently a weak reason to expect a substantial rise in his passing.

Morgan Gibbs-White attempted 21 passes at Liverpool and 56 against Coventry, playing 90 minutes in each. His own-half passing barely changed, 12 versus 11. The entire swing came in the opposition half, from nine passes to 45. Forest’s ability to keep him involved in advanced possession was therefore the decisive difference; deeper exchanges did not compensate when Liverpool limited that access.

Villa brought 37 passes with seven chances created, more creative output than his six against Coventry despite far fewer passes. A productive final ball can emerge from fewer possessions, so chance creation is not a direct proxy for passing volume. Tottenham also held him to 22 passes over 90 minutes, with just 14 beyond halfway. Coventry’s 62% Forest possession and 508 team passes created a larger pool, but Gibbs-White still needed to receive a substantial share of it. For another high count, repeated access in the opposition half is the condition to seek. An isolated goal, penalty or clever final pass does not replace that recurring involvement.

Martin Ødegaard created five chances at Sunderland but attempted only 42 passes in 85 minutes. Against Chelsea he attempted 55 in 77 minutes and created four. The difference was not lack of creativity at Sunderland; it was less repeated circulation. He made 29 passes beyond halfway there, compared with 41 against Chelsea, while his own-half work barely changed. Direct involvement in creating a shot can be valuable without producing the repeated exchanges that accumulate a large passing count.

Brighton offered a second route to a lower total. Ødegaard’s 44 passes included 23 in his own half and only 21 in the opposition half; he created no chances. Against Villa, the same 44-pass total came from just nine own-half passes and 35 beyond halfway. Identical totals therefore concealed very different attacking access. His recorded appearances also ranged from 71 to 85 minutes, so a full 90 should not be assumed. For a higher count, Chelsea- or Coventry-style repeated advanced involvement is stronger support than creativity alone. A matchup that forces him deeper may preserve some passing, but it does not necessarily replace the combinations lost nearer the opposition goal.

Kai Havertz registered a shot on target in each of Arsenal’s five recorded league matches, but that consistency hides a wide range of chance quality and timing. Against Villa, his only attempt was saved after 74 minutes and carried just 0.025 xG. Against Brighton, his saved attempt after 78 minutes was worth approximately 0.59 xG, following a missed corner chance two minutes earlier. Both games produced one shot on target and no goal, yet Brighton offered a far stronger scoring opportunity. A target alone therefore does not tell whether Havertz is receiving the sort of chance that supports an expectation of a goal.

The early opportunities came through different patterns. He scored from an assisted chance after 15 minutes against Coventry, while all three Chelsea attempts arrived between minutes 15 and 25: an assisted save, a blocked fast-break attempt and an assisted goal. Sunderland supplied two chances before the break, including a fast-break shot saved after 33 minutes, followed by a missed corner chance after 75. There is no consistent half in which his opportunities must arrive. Instead, the useful distinction is whether Arsenal can supply him in open play or transition, with corners providing an additional route. A quiet first half did not prevent the substantial late Brighton chance, but one early target against Coventry did not develop into repeated shooting.

Granit Xhaka attempted 75 passes at Ipswich and 58 against Fulham, playing 90 minutes in both. His opposition-half count was exactly 37 in each; the entire difference came from own-half passes rising from 21 against Fulham to 38 at Ipswich. Sunderland’s 62% possession at Ipswich created additional buildup work rather than a larger advanced passing contribution. That is the specific route to the higher total.

The lower-possession games still gave Xhaka a substantial role. Against Arsenal, Sunderland made only 282 team passes, yet Xhaka supplied 56; against City, the team made 388 but he supplied 49. His share of the available circulation mattered as well as its size. Fulham also brought four chances created, twice the Ipswich figure, despite the lower pass total. For another very high count, ordinary advanced involvement needs to be supplemented by repeated deeper exchanges. A dangerous attacking performance or a bigger team passing number alone does not establish those extra passes through Xhaka.

Lewis Dunk’s 114 passes against Villa and 94 at Coventry both came in matches where Brighton gained an extra player, from the 40th and 53rd minutes respectively. Those are useful examples of a high ceiling, but not clean baselines for eleven against eleven. Chelsea provides the stronger equal-numbers comparison: Dunk attempted 99 passes as Brighton held 74% possession and made 647 team passes. Against Villa, the team total was also 647, but Dunk’s larger personal share and greater opposition-half contribution—63 passes there versus 37 at Chelsea—lifted him further.

Leeds brought only 80 Dunk passes despite 67% possession, because Brighton’s total circulation fell to 492. His own-half work was almost identical to Coventry, 50 passes in each, but the opposition-half component was 30 against Leeds and 44 at Coventry. Last season’s home Leeds fixture yielded 98 passes in 90 minutes, while the away Chelsea game yielded 57; the opponent’s name does not preserve the same possession conditions. Dunk’s high count needs both a large team passing pool and repeated use of him within it. His recorded current-season appearances stop at Coventry, so this analysis does not establish his availability or selection for the next match.

Declan Rice reached 75 passes in only 68 minutes against Coventry and 77 in 85 minutes at Villa, through different distributions. Coventry allowed 50 passes in the opposition half; at Villa that fell to 35, but 42 own-half passes compensated. Rice’s passing therefore has more than one route to a high total: he can participate in sustained advanced possession or remain heavily involved earlier in Arsenal’s buildup. The shortened Coventry appearance also shows why a simple full-match average misses how intense the involvement was.

Brighton disrupted the balance. Rice played 90 minutes and Arsenal had 60% possession, yet he attempted 60 passes, with only 28 beyond halfway. His 32 own-half passes did not fully replace the lost advanced work. At Sunderland he managed 68 in 81 minutes even though Arsenal’s team total was only 388, compared with 511 at Brighton, indicating a larger share of the available circulation. Last season’s away Brighton game also held him to a relatively modest 54 in 90 minutes. For another high Rice count, either advanced access must improve or deeper possession must keep running through him; Arsenal’s overall possession advantage alone does not resolve which.

Bruno Fernandes attempted 88 passes at Hull, with 70 in the opposition half, while Manchester United chased a two-goal halftime deficit. Against Ipswich, his hat-trick accompanied only 63 passes and 52 beyond halfway, despite another full 90 minutes. United created far more scoring chances against Ipswich, but Fernandes was more involved in finishing those attacks: eight shots rather than three. A productive attacking performance therefore need not deliver his largest passing workload. The Hull game offered repeated involvement in possession without the same successful endings.

Everton and Fulham sharpen the distinction. Fernandes’ own-half passes were almost unchanged, 17 and 16, but his opposition-half total rose from 35 to 47, taking his overall count from 52 to 63. That is where the additional workload came from. Last season’s away Fulham game produced 54 passes in the same 90 minutes, so the fixture name alone does not fix the total. Even the derby, played largely against ten men after Foden’s dismissal, yielded only 59. For a higher Fernandes count, the useful condition is United repeatedly using him to sustain attacks beyond halfway, particularly when attacks must be rebuilt. Goals, an extra player or a full match on their own do not guarantee that involvement.