The Finishing Gap Behind Celtic’s Tannadice Win

Celtic left Tannadice with four goals; Dundee United left with five reminders that a football can travel inches in the wrong direction. That is the blunt tension inside a 4-0 League Cup defeat. United had 15 shots to Celtic’s 11, won seven corners to three and struck the frame of the goal five times. They also registered zero shots on target. Celtic had eight, and scored four.

The quarter-final place therefore belongs to the side that turned its useful moments into goals, rather than the one that made more noise around the penalty area. Celtic’s clinical edge was not a minor detail added to an uneven afternoon. It was the match. Auston Trusty, Camilo Durán, Sebastian Tounekti and Benjamin Nygren supplied the difference, while Dundee United’s high press and repeated near-misses supplied the uncomfortable context behind a scoreline that looks straightforward.

Celtic made precision look like control

Martin O’Neill’s side had 58% possession, yet this was not a match in which possession alone explains the result. Dundee United began with enough purpose to put Celtic under strain. Julius Eskesen and Emmanuel Agyei each hit the same post before the visitors had scored. The home side then started the second half strongly and continued to find routes into dangerous territory.

Celtic nevertheless established their lead in a brief spell. Trusty met Nygren’s 22nd-minute corner with a glancing header. Soon afterwards, Alex Oxlade-Chamberlain broke into the box and squared for Kasper Høgh. Høgh did not finish, but Durán prodded home from close range.

Those goals altered the terms of every later United attack. A high press can produce territory and loose touches; it cannot undo an early margin unless the final action carries the same accuracy. Celtic had four fewer shots, yet eight more shots on target. The gap between United’s zero and Celtic’s eight is a harsher explanation than the possession split.

Jim Goodwin argued that the scoreline was not a fair reflection of his side’s performance, and the figures give that view some support. His team created 15 attempts, compared with 11 for Celtic, and had four more corners. The result is still not an argument against Celtic’s superiority where it mattered most. It is an argument about the particular form that superiority took: not monopoly of the game, but certainty in both boxes.

The disputed second goal changed an already difficult task

Goodwin’s strongest objection concerned Celtic’s second goal. Dundee United wanted it ruled out for a handball by Høgh before Durán scored. A VAR check found that the handball was not deliberate and the goal stood. Goodwin said the ball striking the Celtic attacker’s hand “directly impacts on the goal”.

The manager also acknowledged that two Celtic players ran into the box unopposed. That is why the incident matters beyond the decision itself. United saw a potential route back closed, then had to confront a second defensive problem in the same phase. With that early margin established, Celtic had converted their threat into an advantage that made United’s misses carry still greater weight.

There is no need to pretend the complaint settles the match. United still had later opportunities, including Zac Sapsford’s effort off the crossbar early in the second half and Bert Esselink’s powerful header that also hit the frame. The decision instead sits inside the wider frustration: United created openings, but the afternoon repeatedly refused to give them a finish or a favourable break.

Five hits of the woodwork, no save required

The five strikes of the woodwork are unusual enough to demand more than a shrug. Eskesen and Agyei struck the same post before Trusty’s opener. Sapsford hit the crossbar after the interval. Esselink’s header followed, and Mehdi Merghem’s long-range effort kissed the junction of crossbar and post late on.

Yet the goalkeeper-saves column is the awkward companion to that sequence: Celtic goalkeeper Viljami Sinisalo made none. Former Dundee United defender Mark Reynolds made the point on BBC Radio Scotland, saying United had not worked the goalkeeper despite their contact with the bar and post.

That does not make every United attack poor. It does explain why the woodwork count should not be treated as a substitute for shots on target. The frame measures proximity. A save, or a goal, forces the contest into a different category. United’s 12 off-target attempts and 11 attempts from outside the box suggest a side capable of reaching shooting positions, but unable to make Celtic’s goalkeeper intervene.

Celtic’s numbers point the other way. Of 11 attempts, eight were on target. Jack Walton made four saves, which means the visitors gave themselves repeated chances to make their edge visible on the scoreboard. In a cup tie, that economy is a serious weapon. Celtic did not require a flood of opportunities because they gave United so few comfortable exits once chances arrived.

Nygren gave Celtic the extra answer

Nygren’s contribution explains why Celtic’s attack felt deeper than the final score’s sequence. He assisted Trusty’s first-half header, helped create Tounekti’s 54th-minute goal and then scored the fourth himself in the 63rd minute with a strike into the top corner. One goal and two assists is direct involvement in three of Celtic’s four goals.

Tounekti’s finish arrived after Celtic broke quickly, cut inside Joshua Rawlins and beat Walton. That third goal effectively ended the tie as a contest, because it gave United a three-goal deficit after another spell in which they had pressed and threatened without scoring. Nygren’s fourth then placed a definitive gloss on a game that had offered United much more territorial encouragement than the result records.

Durán and Høgh also gave Celtic a different kind of problem to defend. Høgh’s physicality, movement and work-rate, together with Durán’s pressing and pace, caused difficulties for the home defence. Durán’s recovery run after losing the ball in United’s penalty area, then winning it back near halfway, drew a large cheer from the travelling support. The attacking work did not stop once Celtic had their lead.

O’Neill was pleased with much of Celtic’s play, though he said there were moments when his side were “really casual”. That qualification matters. A 4-0 win should not erase the early double scare or the further woodwork strikes. Celtic’s advantage was clear in finishing, rather than proof that Dundee United never troubled them.

A cup win, then a sharper examination

Celtic now move into the last eight of the Premier Sports Cup, joining Aberdeen, Ross County and Kilmarnock at that point in the draw. The immediate demand is different: Austrian champions LASK visit in the first leg of Celtic’s Champions League play-off on Wednesday at 20:00 BST.

O’Neill called the coming matches big games because of the attempt to reach the Champions League, its prestige and its finance. LASK arrived after winning their first three Austrian Bundesliga matches and beating SV Ried 4-1 on Friday. Celtic’s finishing at Tannadice offers encouragement; the chances they allowed offer the warning. Against Dundee United, five shots reached woodwork rather than net. Better opposition may ask for a less forgiving version of the same performance.

For Dundee United, the next fixture is the Premiership derby against Dundee next Saturday at 17:45. They carry two successive 4-0 defeats, following the loss to Hearts at Tynecastle. Goodwin will reasonably see the two results as insufficient descriptions of his team’s performance, yet scorelines become patterns if the defence continues to concede decisive chances and the attack cannot turn pressure into shots on target.

Tannadice did not witness a simple dismantling. It witnessed a harsher football calculation. Celtic had eight shots on target and four goals. Dundee United had 15 attempts, five meetings with the frame and no goal. The arithmetic is not glamorous, though it is usually what cup ties remember.

Research: This article uses the archived evidence supplied for this story. Story inspiration: BBC Sport's original report.