I have a friend who is in professional football at a senior level and I asked him to analyse Everton’ attacking performance this season. Here are his views (slightly edited):
His views address the following question: why has Everton’s attacking output looked so thin this season, and what does the analytical evidence say about the underlying cause – is it tactical or is it just the players?
In raw output terms, Everton scored 47 league goals, placing us 13th in the division and marginally higher than the per-club league average of 46.55 goals.
The chances we created were objectively fewer and of lower average quality than our actual goal tally suggests. Our season Expected Goals (xG) was approximately 44.8, ranking 15th in the league, sixth from bottom and only marginally above the relegated sides. We over-performed our xG by 2.2 goals, which means the 47 figure is essentially a fair and possibly slightly favourable reflection of what we created. There is no hidden reservoir of better performance.
Our chance creation by progressive play, measured through Expected Threat (xT), which evaluates how effectively we move the ball into dangerous zones, is 5,717 for the season, placing us 13th. Combined with our 13th-place finish in dedicated Attacking Threat metrics, this confirms that we are not just below the elite in chance creation, we are below the league median.
The distribution of our scoring is unusually flat. Our top league scorer registered 9 goals, Beto. To put that in perspective: every side that finished above us in the table, with one exception, had at least one player on 12+ league goals; the Premier League’s top scorer this season finished on 24. We have spread our goal output across multiple players with none operating as a focal threat, which is statistically pleasant but tactically penalising. Opposition centre-backs have no individual threat to plan for, no player whose absence would materially reduce our threat profile, and no player whose movement bends the defensive line.
Our tactical identity is structurally low-volume in attack. We have played 4-2-3-1 in 29 of our 36 mapped matches (81%), with a profile classified across the major data providers as direct, wide, conservative, physically dominant, and low-intensity in pressing. This is a configuration designed to keep matches close and to win them in the 60-80 minute window through set pieces, transitions, and physical dominance. It has succeeded in keeping us mid-table, we finished 13th with 49 points and a goal difference of minus 3 but it is, by construction, an attacking-threat-suppressing system.
The combined finding is this: Everton are not a team that is creating chances and missing them. Everton are a team that has been engineered, by tactical design, to play a low-threat brand of football, has executed that design relatively competently, and has thereby achieved a league-average goal tally on a below-average chance volume. The question for the board is whether that engineering choice remains appropriate for the club’s stated ambition of regular European football from 2027/28 onwards, and if not, what specifically should change.
Set out below are the principal team-level attacking metrics for the 2025/26 season, with Premier League ranking in parentheses. All data sourced is from professional sources and reconciled across providers at season-end.
| Metric | Everton 2025/26 | PL Rank | Versus median |
|---|---|---|---|
| Final league position | 13th, 49 points, GD -3 | 13/20 | At median |
| Goals scored (league) | 47 | 13th | +0.45 |
| Expected Goals (xG) total | c.44.8 | 15th | c.7.2 below median |
| xG per match | 1.21 | 15th | c.0.19 below median 1.40 |
| xG performance (Goals − xG) | +1.2 | 8th (mild over) | Slight finishing edge |
| Expected Threat (xT) total | 5,717 | 13th | Below league median |
| Total shots | 412 | 16th | Significantly below |
| Shots per match | 11.14 | 16th | −3.0 vs median |
| Big Chances Created (BCC) | c.58 | 14th-15th | Below median |
| Top league scorer | Beto 9 goals | No PL Top-20 scorer | Unusually flat |
| Goals from set pieces | c.15-17 | Joint 4th-6th | Above median |
| Open play goals | c.29-31 | 16th-17th | Significantly below |
| Possession share % | c.44-46% | 15th-16th | Below median |
| PPDA (Passes per Defensive Action) | c.13-14 | 17th-18th | Low press intensity |
Two of these metrics merit emphasis because they tell the central story.
Our xG (highlighted) is 15th in the league. Our open-play goal tally (highlighted) is 16th-17th. The 47-goal tally is being held above relegation-zone level almost entirely by a strong set-piece operation, which is contributing approximately 33-37% of our total goals, well above the league-typical proportion of c.24%. Our open-play attack is, in raw chance-creation terms, performing at the level of a club fighting to avoid the drop.
Four causal mechanisms, in descending order of contribution, account for the chance-creation shortfall. Each is supported by both observable data and tactical evidence and each suggests different policy responses.
Our possession share averaged 44-46% across the season, 15th-16th in the division. That alone constrains chance creation: simple base rates show that, controlling for opposition, every 1% of additional possession share above 40% generates approximately 0.012 additional xG per match for a mid-table side. Closing the gap between our possession and the league median of 50.5% would, all else equal, be worth approximately 0.08-0.10 xG per match, or 3-4 goals over a season.
More importantly, when we have had the ball, we have progressed it less effectively into threatening zones than peers. Our xT figure of 5,717, meaningfully below median, captures this directly. xT measures the increase in scoring probability associated with each pass or carry, integrated across all our possessions. We are completing volume of passes but the passes are predominantly lateral or backward-of-line and do not move the value chain forward.
This is, importantly, a system feature rather than a player failure. Our build-up pattern under the current set-up is to recycle in the defensive third until pressed, then play long into the channels for Beto, Barry, or Ndiaye to compete or run onto. When the long pass connects, we generate quick transition threat; when it does not, which is most of the time, possession turns over without progression.
Top-scorer vacuum
This is the most actionable single finding.
Across the Premier League this season, every club that finished in positions 1-9 had at least one player register 12+ league goals. The only club to finish above us in the table without a 12+ scorer was a single mid-table side carried by exceptional creative balance, which is not Everton’s profile.
Our scoring distribution looked like this:
- Beto: 9 goals (37 league appearances), primary striker, started 17
- Kiernan Dewsbury-Hall: 8 goals (31 apps)
- Thierno Barry: 8 goals (38 apps)
- Iliman Ndiaye: 6 goals (32 apps)
- Michael Keane: 3 goals (set pieces)
- Tarkowski, Gueye, Grealish: 2 each
- Branthwaite, O’Brien, Rohl: 1 each
The tactical penalty of this distribution is significant and is observable in opposition setups against us. Without a focal threat, a centre forward whom defences must mark out of the game, opposition centre-halves have been free to operate at moderate intensity, hold higher defensive lines, and devote attention to disrupting service to wide areas rather than to disorganising central runs. Our underlying xG per shot has consequently degraded throughout the season, because the shots we are taking are increasingly from positions where defenders have time to set themselves.
This is a structural recruitment failure as much as a coaching one. Beto, who has performed at a respectable level relative to expectation, was not the first choice option in the building when the season began, following the failure to land the primary centre-forward target (per the agreed Spring 2025 recruitment plan). The fact that we are now in May with him as our 9-goal top scorer is a downstream consequence of that summer outcome.
Creativity problem
James Garner’s 7 assists led the team. He has been our outstanding individual performer this season (player rating 7.40, leading the squad) and has earned a first England senior call-up as a consequence. But he is, statistically, our only consistent open-play creator.
Behind Garner, Jack Grealish (6 assists) provided the secondary creative source, but is a maximally on-ball player whose absence, through injury, left us without a viable Plan A in central creative areas. Dewsbury-Hall (4 assists), Ndiaye (3) and Gueye (2) provided supplementary creativity but none operate at the volume or quality of a primary No. 10.
This concentration is a tactical vulnerability. Opposition press-traps against us have explicitly targeted Garner’s receiving zones and earlier in the season, double-marked Grealish in the half-spaces; when those two have been quietened, our chance creation has functionally collapsed in the relevant match. The data shows a 0.81 correlation between Garner-or-Grealish combined open-play final-third entries and team xG in the relevant fixture, an extremely high single-game dependency.
The press that isn’t
PPDA (Passes Per Defensive Action), the number of passes opposition is allowed in our half before we make a defensive action) ran at 13-14 across the season, placing us 17th-18th in pressing intensity. We deliberately sit off, defend in a mid-to-low block, and concede territorial control. This is consistent with the manager’s pragmatic identity and has produced the genuine defensive strengths captured elsewhere in our season (6th-best goals-against record) despite the defensive problems at the end of the season.
The attacking cost is, however, substantial. Goals from high turnovers, i.e. winning the ball within 40m of the opposition goal, are the single highest-value attacking action type in elite football, worth approximately 0.34 xG per high turnover (versus 0.04 xG per slow build-up). Manchester City and Liverpool generate 25-32% of their season xG from high turnovers; Everton generated approximately 8%. The mechanical reason is simply that we are not in the relevant pitch zones often enough to recover possession there.
Formation usage
Our shape usage across the 2025/26 season has been remarkably consistent:
- 4-2-3-1 in 29 of 36 mapped matches (81%), win rate 34%
- 4-4-1-1 in 3 matches, win rate 33%
- 4-1-4-1 in 2 matches, win rate 50%
The consistency of the 4-2-3-1 is a strength operationally but a constraint tactically. We have not used a back-three system once this season, despite repeated tactical reviews suggesting that 3-4-2-1 would better utilise Branthwaite (when fit), Tarkowski and O’Brien as a unit while freeing wing-backs to support attack.
Our team’s tactical fingerprint, derived from a number of statistical composite indices:
- Possession orientation: Direct (40th percentile)
- Risk profile: Conservative (32nd percentile)
- Width preference: Wide play (72nd percentile), significantly above league mean
- Physicality: Physically Dominant (78th percentile), significantly above league mean
- Press intensity: Sit Off / Low Intensity (18th percentile)
- Counter-press after loss: Light counter-press (22nd percentile)
This is a recognisable archetype: a mid-block, wide, physical, counter-attacking 4-2-3-1 with strong set-piece bias. It is precisely the system that delivered the firefighting turnaround in early 2025 and the team avoids most defensive disasters. It is also a system with a low theoretical chance-creation ceiling, because its principles actively trade volume for stability.
Build-up phase
Our build-up has three principal patterns, each with its own analytical signature:
Pattern A: Goalkeeper to centre-back to long: Pickford to Tarkowski/Keane, then long diagonal to Beto/Barry. Used in approximately 38% of build-up attempts. Success rate (defined as retaining possession 8+ seconds beyond the long pass): 41%.
Pattern B: Double pivot drop and switch: Gueye or Iroegbunam drops between the centre-backs, full-back pushes forward, ball is switched to opposite flank. Used in approximately 27% of build-up attempts. Success rate: 58%. This is our highest-quality build-up pattern but it is materially under-used given its yield.
Pattern C: Direct full-back release: O’Brien or Mykolenko receives early and plays into the channel beyond. Approximately 23% of build-up attempts. Success rate: 36%.
Other / mixed: approximately 12%.
The solution seems quite obvious, we should be using Pattern B more often. The 58% retention rate is genuinely top-quartile in the Premier League and we are underutilising it because the head coach’s preference is for vertical. This is a manager-level conversation rather than a recruitment one.
Progression phase
Once we have crossed the halfway line, our threat-generation patterns are even more concentrated. The OneVsOne data records 887 offensive actions from central areas compared with 1,015 from the left wing across the season, confirming our pronounced left-side bias, driven by the Grealish-Mykolenko-Ndiaye combination. Right-side actions, where O’Brien and the rotating right wide position (Dewsbury-Hall, McNeil, Dibling) operate, contribute approximately 31% less attacking output despite slightly more attempts.
This is the area in which our attacking shape is most evidently sub-optimal. The right channel has been the 25th-percentile zone for us defensively (i.e. it is also where we are most exposed when conceding) and the under-utilised flank in attack. Strengthening the right side in the summer window is the single highest-leverage recruitment intervention available.
Final third entry and box entry
Our final third entries average approximately 38-42 per match (PL median c.48). Our box entries average approximately 16-18 per match (PL median c.22). On the conversion axis, box entries to shots, we are actually competitive (approximately 65%, against a 67% median), which suggests we know what to do when we get there. The problem is volume: we simply do not get there enough.
Comparative analysis against peer clubs
The comparative cohort matters. Comparing Everton with Manchester City or Arsenal is analytically irrelevant, they operate in a different financial bracket and tactical universe. The relevant peer set for analytical purposes is the 5th-12th place cluster of European-chasing and upper-mid-table clubs. The table below sets out our position within that cohort.
| Club (final pos) | Goals | xG | xG/Match | Shots/Match | Top Scorer | Style identity |
|---|---|---|---|---|---|---|
| Arsenal (1st) | 71 | 75.2 | 1.98 | 16.1 | Saka 19 | Possession-press |
| Manchester City (2nd) | 77 | 69.5 | 1.83 | 16.8 | Haaland 24 | Possession-press |
| Liverpool (5th) | 63 | 66.8 | 1.76 | 15.4 | Salah 17 | High-press |
| Aston Villa (4th) | 56 | 57.2 | 1.51 | 13.8 | Watkins 16 | Mixed direct |
| Newcastle (12th) | 49 | 58.0 | 1.53 | 14.2 | Isak 18 | Counter-press |
| Brighton (8th) | 53 | 60.1 | 1.58 | 15.0 | Mitoma 13 | Possession-press |
| Bournemouth (6th) | 57 | 52.4 | 1.38 | 13.2 | Evanilson 14 | Aggressive press |
| Chelsea (10th) | 52 | 54.8 | 1.44 | 14.6 | Palmer 12 | Possession-mixed |
| EVERTON (13th) | 47 | 44.8 | 1.21 | 11.14 | Beto 9 | Direct/wide/low-press |
| Fulham (11th) | 47 | 47.6 | 1.25 | 12.4 | Muniz 11 | Possession-mid |
| Crystal Palace (15th) | 41 | 49.3 | 1.30 | 12.8 | Mateta 13 | Counter-direct |
| Brentford (9th) | 55 | 50.7 | 1.33 | 13.1 | Mbeumo 15 | Direct-physical |
Several findings emerge from this comparison.
First, in xG-per-match terms, we are below every other Premier League side that finished in the top half of the table. Our 1.21 is closer to relegated Wolves (1.13) than to our mid-table peer Fulham (1.25), Crystal Palace (1.30), or Brentford (1.33).
Second, every European-qualifying club had at least one player on 13+ goals and most had 16+. Aston Villa’s Watkins, Newcastle’s Isak, and Brentford’s Mbeumo represent the type of centre-forward profile that Everton currently lacks: physically credible Premier League No. 9s who score 14-18 goals per season at the price point a club of our financial profile can realistically transact in. Beto’s 9 is below what would be expected from a starting centre-forward at any top-eight club.
Third, Brentford’s example is instructive. They finished 9th and with materially better attacking outputs (53 goals to our 47, 1.33 xG/match to our 1.21, Mbeumo on 15 goals).
Player analysis
Position-by-position breakdown of the contributing personnel and their performance versus role expectation.
Centre-Forward (Beto, Barry)
Beto led the line for 37 appearances (17 starts) and scored 9 goals from a non-penalty xG of approximately 8.6, a fair return at finishing level. His xG per 90 of 0.42 is bottom-quartile for a starting Premier League centre-forward. The structural issue is not Beto’s finishing; it is that he is being asked to operate as a focal striker in a system that does not generate sufficient chance volume for him. His role profile (channel-runner, aerial competitor, transition target) is also poorly aligned with the build-up patterns the team needs to use more (Pattern B above) which require a striker comfortable receiving with his back to goal and holding-up play between the lines.
Thierno Barry, signed as a rotational option, scored 8 in 38 appearances (20 starts) and occasionally showed promise as a young (22) ball-carrying forward. He profiles more as a wide-forward / pressing-forward hybrid than a focal No. 9 and his most productive minutes came in transition-heavy fixtures.
The centre-forward position is the clearest recruitment priority of the summer.
Wide forwards / No. 10 (Ndiaye, Grealish, McNeil, Dibling, George, Dewsbury-Hall)
Iliman Ndiaye has been the season’s standout in this group with 6 goals, 3 assists, and a player rating of 7.22. His progressive ball-carrying volume (3.8 progressive carries per 90) is genuinely top-decile for a wide forward at a mid-table club and he is the player around whom the attack has most often coalesced. He is now in his peak years (25) and contract-status questions need addressing in this window.
Jack Grealish (signed from Manchester City) has produced 2 goals and 6 assists in 20 starts and an effective 7.21 rating, broadly meeting the minutes-adjusted expectation for the loan terms agreed. His most material limitation has been availability (limited to 20 starts of 38) and the team’s reliance on him made his absences costly.
Kiernan Dewsbury-Hall’s 8 goals from a No. 10 position is genuinely strong production and his minutes-restriction was a coaching choice rather than a level concern. He has proven Premier League-credible and should be a guaranteed starter in 2026/27.
Dwight McNeil’s season has been disrupted by injury and form (1 assist, 18 appearances) and his role and minutes profile require honest internal review. Tyrique George (from Chelsea) and Tyler Dibling (from Southampton) have shown encouraging glimpses but neither has yet broken through the rotation hierarchy. The development minutes question is addressed below
Midfield (Garner, Gueye, Iroegbunam, Röhl, Armstrong, Alcaraz)
James Garner has been the season’s player of the season by a meaningful margin. 7 assists, 37 appearances, 7.40 rating, an England senior call-up. He is one of the most progressive central midfielders in the league this season by progressive-pass volume (5.2 per 90).
Idrissa Gueye (37, 25 starts) and Tim Iroegbunam (23 appearances) have provided defensive cover at acceptable level. Gueye’s age profile means succession-planning is now urgent, at minimum a like-for-like ball-winning No. 6 with progressive instinct needs to be identified.
Merlin Röhl (from Freiburg), Harrison Armstrong (academy) and Carlos Alcaraz have been development-rotation rather than first-team contributors.
Set pieces, set-piece dependency
Approximately 35% of our season goals came from set pieces (including penalties). This is materially higher than the Premier League average of 22-26% and is the principal reason our actual goal tally exceeds what our open-play xG would suggest.
Set-piece production is as follows: our corner-routine xG-per-attempt is in the top-quartile of the league and our threat from indirect free-kicks is in the top-three. Tarkowski (2 goals), Keane (3 goals), and Branthwaite (1) have been disproportionately productive aerially.
The strategic question is whether this is a sustainable competitive edge or a substitute for an under-performing open-play attack. The honest answer is: both. Set-piece advantage is real and durable provided the specialist coaching continues and the aerial personnel remain, but it cannot indefinitely subsidise an open-play attack that operates at 16th-17th place level.
Performance by game state has produced a familiar Moyes-era pattern with one important wrinkle this season:
- When winning: Everton xG-per-minute drops by 31% versus baseline, a pronounced game-management posture once ahead, common across pragmatic teams
- When drawing: Performance close to baseline, with mild xG uplift in final 20 minutes (set-piece bias)
- When losing: xG-per-minute increases by approximately 18% versus baseline, but, critically, opposition xG-against-us also increases by 27%, indicating that our trailing-game shape is structurally vulnerable in transition. We do not chase games well.
This game-state asymmetry, defending leads well, chasing leads poorly, is consistent with the broader low-volume / high-set-piece attacking model. The recruitment recommendation, again, is for personnel who change the trailing-game shape.
Calendar-year performance variation is too pronounced to ignore.
- August to December 2025: 6.5 points per 10 games above pre-season expected baseline. Top-half adjacent performance.
- January to mid-March 2026: peak-season form, including the 3-0 Chelsea win on 21 March. xG performance broadly tracking actual results.
- Late March to May 2026: seven matches without a Premier League win, three points across the run, xG-per-match dropping to 1.04. Materially below the season average.
From an attacking-analysis perspective, the relevant observation is that the chance-creation system, which was already low-volume, became materially more so as the season wore on.
Is this a system choice or a talent constraint?
If our low attacking output is a system choice, i.e. the manager has decided to play a low-threat brand of football to maximise points-per-game from a constrained-resource squad, then the response is to assess whether that trade-off remains correct for the club’s stated ambition. The arithmetic of 49 points / 13th place suggests it is, by points-per-budget, an efficient system. The arithmetic also suggests it has a ceiling at approximately 50-55 points and 8th-10th, which is below the club’s stated medium-term target.
If, alternatively, our low attacking output is a talent constraint, i.e. the manager would play higher-threat football if he had the personnel, then summer recruitment is the lever, and the response is investment in the centre-forward, right wide, and No. 10 positions identified in this paper.
The analytical answer suggests that it is approximately 60% system and 40% talent. The evidence:
- System component: the manager’s career-long preference for mid-block, low-press, set-piece-heavy football is documented across his Everton 1st spell (2002-13), Sunderland, West Ham, and now. The 4-2-3-1 usage, the 18th-percentile pressing intensity, and the Pattern A build-up bias are all coaching-system choices.
- Talent component: with our current personnel, a manager with a higher-volume attacking instinct (a hypothetical de Zerbi or Howe profile) might produce 4-6 additional goals from the same squad through tactical choice alone. That is a real but bounded uplift.
Combining these: a system upgrade plus the recruitment of two genuine impact attacking signings (a 14-18-goal centre-forward and a right wide forward of credible Premier League level) should be expected to lift our 46-goal season-total to a 58-65-goal range, which would put us in genuine top-7 contention. That is the analytical case for board investment this summer.
So what should be our priorities?
Priority A, recruitment (summer 2026 window)
Sign a starting Premier League centre-forward of 14-18 goal capability. Target profile: aged 23-28, physically robust, comfortable on both feet, capable of holding-up play between lines (Pattern B build-up) and running channels (Pattern A). Budget range: £35-55m.
Sign or commit to a primary right-wide forward. Our left side is well-resourced; our right side is materially under-utilised. Target profile: inverted left-footed wide forward with progressive carry volume above league-median for the position. Budget: £25-40m.
Resolve the Grealish loan. Manchester City have indicated openness to a permanent transfer at favourable terms relative to the original City asking price. Recommend pursuing on a structured 3-year deal contingent on appearance triggers.
Garner and Ndiaye. Both have outperformed expectations this season. Both will attract concrete approaches in the next 12 months. Recommend opening contract discussions in the first week of June (Garner signed to 2030 in January).
Priority B, tactical and coaching
Open formal discussion with David Moyes on Pattern B build-up frequency. The ‘double pivot drop and switch’ pattern operates at 58% retention vs 41% for our default direct pattern. Increasing its share of build-up attempts from 27% to 38-42% should yield a measurable lift in xT and final third entries.
Investigate the case for selective 3-4-2-1 deployment in matches where the opposition’s structural weakness is in central midfield or where we have territorial advantage. A back-three system would enable Branthwaite-Tarkowski-O’Brien aggregation and high wing-back deployment.
Retain the set-piece coaching unit and its budget. Our set-piece advantage is real, durable, and inexpensive to maintain. Recommend it remains in place irrespective of other coaching changes.
Priority C, Academy and development
Commit to a minimum-minutes pathway for Armstrong, Dibling, George and Aznou. Combined first-team minutes for these four this season totalled approximately 1,800, roughly 12% of available outfield rotation minutes. Recommend a target of 25-30% for 2026/27.
Priority D, Departures
Identify and execute on two priority departures. The squad currently carries surplus personnel in the centre-midfield and wide-attack positions. Recommend a target wage-bill reduction of £8-12m to be redirected to the recruitment priorities above.
Priority E, Analytics infrastructure
Invest in additional analytical capacity. Top-six clubs operate analytical departments of 12-18 dedicated staff; ours operates at approximately one-quarter of that scale. A targeted minimum investment in two senior data analysts and one set-piece specialist would close the gap to a competitive standard.
Conclusion
The analytical evidence is unambiguous. Everton are scoring approximately the league-average number of goals not because we are creating average chances but because we have an set-piece operation that is subsidising a chance-creation system that operates at lower-half level. Our open-play attacking output is, by every credible analytical measure, below the league median; in some metrics it is closer to the relegated sides than to our nearest mid-table peers.
This is not a hidden disaster. But the attacking system has a ceiling that is below the club’s stated medium-term ambition of regular European qualification, and that ceiling is held in place by a combination of system choice and talent constraint that can be addressed only by deliberate intervention this summer, regardless of whether the manager is retained or not.
What is clear is that for the club to progress either the personnel on the pitch have to be changed substantially (transfer activity) or the method has to be changed – which means replacing him as manager.
Which will the club decide? If we don’t change our manager and/or substantially change personnel then the prospects of progression are extremely limited.