Are we still playing like Union?

Union’s six stated principles, measured on event and tracking data for every league match under Alexander Blessin, Sébastien Pocognoli and David Hubert, and benchmarked against every other club in the league in the same season. The defensive half of the identity is intact and still among the best in Belgium. The attacking half has drifted: verticality has fallen from third in the league to mid-table, and wing play has been below league average for output since 2024/25 even though the wing-backs still touch the ball more than at any other club.

A back three in all 126 matches across the three coaches. The structure has never been the question.

The verdict: six principles, three coaches

Each principle is turned into four to six measurable things, each one compared with the spread across Jupiler Pro League clubs in the same season. 50 is the average club; 60 is a standard deviation better. Spells shorter than fifteen matches are drawn faded.

Elite for four of the five spells. The one dip is Hubert’s first full season.

The most reliably executed principle after duels. Opponents get almost nothing.

Never below second in the league. The principle that has never wavered.

The clear erosion. From third in the league under Blessin to mid-table since.

Structurally intact, but the output has been below league average since 2024/25.

The most volatile. First under Blessin, second under Hubert, thin so far this season.

WHAT THE TRACES SAY

Three principles are alive and well. Duels have never dropped below second in the league in any spell, the single most consistently executed principle at the club. The compact block sits top four in four of five spells: opponents take 2.2 shots per 100 passes against Union, against a league average of 3.3. Pressing has been first, second or third in four of five spells.

Two have drifted. Verticality ran 3rd in the league under Blessin and has been between 6th and 11th ever since. Forward passing has gone from 66.8% (2nd) to 63.1% (league average), and the number of touches needed to reach the final third from 5.6, the second most direct team in Belgium, to 6.8, which is ordinary.

And one was never really there. Wing play has ranked 10th, 13th, 12th, 13th and 16th of the league. Union cross less than almost every club in Belgium and turn crosses into box entries less often than anyone. Section 02 shows why that is not the contradiction it looks.

Principle by principle

The club’s own wording for each principle, then every metric behind it: Union’s figure for that spell, the average Jupiler Pro League club in the same season, and where Union ranked.

HIGH INTENSITY AND PRESSING, HOGE INTENSITEIT EN PRESSING

“Druk zetten op de opbouw van de tegenstander om fouten af te dwingen.”

how much = whether Union does the thing; how well = whether it works. Rank is among Jupiler Pro League clubs in the same season, 1 = most in line with the principle. Union’s figure is the average over that spell; the league figure is the average club that season. The principle index puts 50 at the average club and every 10 points at one standard deviation.

THE WING-BACK PARADOX, AND WHERE THE SET-PIECE EDGE WENT

Wing play. Union’s wide players take 19.8% of all touches under Blessin and 16.8% under Hubert, against a league average that fell from 13.3% to 9.8%. On that measure Union is first or near-first in the league every single season: the wing-backs genuinely are the structure. But crosses per 90 rank 16th of 16 in 2024/25, and only 10.5% of box entries come from a cross against a league average of 16.0%. The ball reaches the flanks more than anywhere else in Belgium and converts into the box less. That is the gap between the principle as written (“om de zestien te bereiken”) and the principle as played.

Set pieces. The 2021 reputation is real but it belongs to specific spells. Blessin’s side generated the most set-piece shots in the league (2.1 per 90) but converted corners into shots only 11th best. Hubert’s 2025/26 is the opposite and better: 24.3% of corners produce a shot, first in the league and well clear of the 14.8% average, with headed shots at 22.2% of all attempts. Volume under Blessin, quality under Hubert.

Pressing. The only real defensive wobble is Hubert’s first full season: PPDA 10.6 against Pocognoli’s 8.7, and losses won back inside five seconds down to 32.7% from 40.1%. The six matches of 2026/27 show it coming back: 28.5 regains in the final third per 90, the highest of any spell.

The block, measured on tracking

Event data cannot honestly measure “compact blok”. Tracking can. This is where Union’s ten outfield players actually stand when the opponent has the ball.

SHAPE OF THE OUT-OF-POSSESSION BLOCK, RANK AMONG JUPILER PRO LEAGUE CLUBS

Tracking frames at one per five seconds in which the opponent has the ball, Union's ten outfield players. Back line = mean height of the four deepest; block depth = back line to the two highest. Raw metres are not comparable between seasons (the league average block width itself moves 42.8m to 35.8m between 2024/25 and 2025/26, a processing change), so only the standing within each season is shown. 2023/24 had tracking for six matches only and is left out.

A HIGH LINE, A WIDE BLOCK

Under Hubert in 2025/26 Union held the highest back line in the league out of possession (37.2m from their own goal, 1st of 16) and spent less time pinned into their own third than any other club (44.1%, 1st). That is the opposite of a team that has gone passive, and it sits oddly beside the softer pressing numbers: they hold a high line but engage later. A high line with a mid-block press is a coherent choice, but it is not the same principle as “druk zetten op de opbouw”.

The block is also wider than almost every side in the league (15th of 16). That is the price of a back five: the wing-backs hold the width. It is a description of the system, not a fault, and the box-protection numbers in section 02 show the trade is working.

Do these principles actually win points?

A fair question before asking any coach to execute a principle harder: in this league, do clubs that score well on it actually finish higher?

CORRELATION BETWEEN A CLUB'S SCORE ON EACH PRINCIPLE AND ITS POINTS PER GAME

Every Jupiler Pro League club-season in the four seasons (66 in total) scored on Union's principles, against its points per game. Correlation, not cause.

READ THE TWO NEGATIVE NUMBERS CAREFULLY

Verticality correlates -0.87 with how much of the ball a team has. It is very nearly a measure of not having possession, so its negative link with points is mostly the possession-and-points relationship in disguise, not evidence that being direct loses games.

Wing play is different. Its correlation with possession is only -0.15, so that excuse does not apply. In this league, the clubs that lean hardest on the flanks do tend to finish lower, which makes it worth asking whether the principle is written the way the club actually wins.

Union is not a possession team in any of these seasons (49.7%, 48.2%, 54.8%, 51.9%, 50.5% of open-play passes), so its drift away from verticality is a real change of method, not a by-product of taking the ball off opponents.

WHAT TO DO WITH THIS

The three principles Union executes best, pressing, the compact block, duels, are also the three that correlate positively with points in this league (0.44, 0.50, 0.21), and set pieces with them (0.36). The club is strongest exactly where strength pays.

The honest conclusion on the attacking half is not “the coaches failed”. It is that verticality and flank play, as the club has written them, are the principles of a challenger, and Union have spent three years as the team everyone else sets up against. What has not happened is anyone rewriting them. If the club still wants them, the wing-play gap, most touches in the league on the flanks, near-fewest box entries from crosses, is the single most addressable thing in this report.

The most, and least, Union performances

Every league match scored against Union’s own norm rather than the league’s, so this answers “how Union was this performance” rather than “how good was it”.

EVERY LEAGUE MATCH, MOST UNION TO LEAST

All 126 league matches, most Union on the left. Green = closer to how Union usually plays, red = furthest from it. 50 is Union’s own average performance, not the league’s. A white outline marks a win. Hover for the match.

THE TEN MOST UNION PERFORMANCES

THE TEN LEAST UNION PERFORMANCES

PLAYING LIKE UNION DOES NOT DECIDE THE RESULT

The most Union performance of the three years is La Louvière at home, 8 Feb 26 (2-1, index 70.3), and the least is Standard Liège at home, 9 Mar 25, which Union won 3-0.

That is the pattern, not an exception. The average index is 50.5 in wins, 48.9 in draws and 49.8 in defeats: effectively no difference. Several of the least-Union matches are comfortable home wins, because a side that is 3-0 up against a passive opponent stops pressing, stops countering and stops crossing. The index measures how strongly the identity was expressed, and a team only gets to express a counter-pressing, vertical identity when the opponent gives it something to press and counter against.

The exception worth reading is the bottom of the list: the 0-5 at Club Brugge in May 2026 is the one match that is both far from the identity and a heavy defeat, with pressing -2.1 and the block -1.6 standard deviations below Union’s own norm.

The most, and least, Union players

Which players express these principles most, measured against every player in the league in the same position and season, not against their own team-mates, which would only say who is most like Union.

CENTRE-BACKS, 13 PLAYER-SEASONS OF 600 MINUTES OR MORE

Each player is measured against every Jupiler Pro League player in the same position group and season, on the principles his role is accountable for (winning duels, high intensity and pressing, set pieces, verticality). 50 is the average player in that group. A low score is not a verdict on quality, it says the player expresses this particular set of principles less than his peers, which for a creator is often the point.

THE IDENTITY IS CARRIED FROM THE BACK, AND THINS OUT GOING FORWARD

Ross Sykes is the most Union player of the three years, and the centre-backs own the top of the list, Sykes appears in the top four in 3 separate seasons, on set-piece threat and duels. The group averages fall in a straight line from back to front: centre-backs 55.8, wing-backs 54.8, midfielders 52.6, forwards 50.8.

The wing-backs are not the problem. All 11 wing-back seasons score above the league average for their position, from 52.6 to 58.5. The ball reaches them and they do their job. The gap in section 02, most flank touches in Belgium, near-fewest box entries from crosses, is not created on the flank.

It is created in the box. Every one of the four lowest scores in the squad is a forward, and they share a profile: Ait El Hadj contests 0.4 aerial duels per 90 (1.6 standard deviations below a league forward) and took no headed and no set-piece shots at all in 2025/26, while entering the box more than almost any forward in the league (3.3 above average). Fuseini and Zeneli are the same shape. Set against them, Gustaf Nilsson in 2023/24 contested 6.3 aerial duels per 90 and Mateo Biondic in 2025/26 took 1.3 headed shots per 90, 2.2 above the league forward.

The forward line has moved from aerial target men to carriers and creators. That is a coherent way to play, and it is also the mechanism behind three of this report’s findings at once: fewer box entries from crosses, a smaller set-piece share of shots, and a wing-play principle that reads as unfulfilled. This is a recruitment question, not a coaching one, and it is the one decision in this report that sits squarely with the sporting director.

How this was measured

METHOD

Every Jupiler Pro League match of the four seasons was processed, 66 club-seasons in total, so Union is always compared with the league of the same season. That matters: parts of the event feed changed density between seasons, and a league-wide shift cancels out inside a within-season comparison while a raw season-on-season number would not.

Only measures that are stable across all four seasons are used: coordinates, pass end points and distances, pass direction, completion, set-piece type, event type and position. Deliberately excluded are expected goals, phase-of-play tags and off-ball run tags, all of which only exist from 2025/26 and would have invented a trend.

Coaching spells follow the club record: Blessin to June 2024, Pocognoli from July 2024 to 11 October 2025, Hubert from 13 October 2025. Champions League matches are excluded so every comparison stays inside one league.

WHAT THIS CANNOT TELL YOU

It cannot separate the coach from the squad. A spell’s score reflects the players available as much as the instructions given.

Two spells are small. Pocognoli’s ten matches in 2025/26 and Hubert’s six in 2026/27 are indicative only; single-figure samples move a long way on one match.

It measures expression, not merit. A low index on a match or a player means distance from this particular set of principles, nothing more.

Equal weight is an assumption. All six principles count the same because the club lists them as six. Weighting them differently would change the overall index.