Glasgow offers Celtic a chance to break an uncomfortable pattern
Celtic welcome Beşiktaş on Thursday, 5 November 2026, for a Europa League fixture scheduled at 20:00 UTC. Three games into the recorded league phase, the hosts have no points. Beşiktaş have seven, following two wins and a draw. For Celtic, this is no longer simply a question of building momentum. Their European results need to change.
The challenge is unusual because Celtic's domestic performances have not been uniformly poor. They have recently scored seven against Hearts and created enough chances to win at Falkirk, yet their European attack has struggled to impose itself. Beşiktaş, meanwhile, have scored freely in the competition, including three away at Hoffenheim.
Celtic's European defeats reveal more than a scoring problem
Celtic opened the league phase with a 3-1 home defeat to Ferencváros. The underlying numbers were closer than the score: Celtic created **1.30 xG** and allowed **1.15 xGA**. Conceding three from that chance profile made the result particularly frustrating.
The second match was a 4-1 defeat away at Benfica, where the chance balance was much less favourable. Benfica generated **3.18 xG**, while Celtic managed **1.36**. Against Celta Vigo on 22 October, Celtic lost 2-1 at home and produced only **0.51 xG** to the visitors' **1.21**. Their most recent European display was therefore also their least productive in attacking chance quality.
Across those three fixtures, Celtic have scored **three goals from 3.17 xG** and conceded **nine from 5.54 xGA**. Their goals scored are broadly consistent with the chances created; the larger discrepancy is defensive. Nine conceded from 5.54 xGA points to an especially costly sequence of opposition finishing and match events, though the Benfica and Celta games also show genuine weaknesses in chance prevention.
Domestic results complicate the picture. Celtic beat Hearts 7-1 on 18 October with **4.16 xG**. Yet they drew 2-2 at home to Dundee United on 28 October despite a **1.55 to 0.87 xG** advantage, then lost 2-0 at Falkirk on 1 November after creating **2.77 xG** against **0.99 xGA**. That latest defeat suggests a failure to convert chances rather than a lack of attacking opportunities.
The contrast is striking: Celtic have recently generated substantial xG in Scotland, but their Europa League output has declined across three successive games. The Scottish Premiership page provides the wider domestic picture.
Beşiktaş have made chances count in Europe
Beşiktaş began with a 4-1 home victory over Marseille, backed by **2.93 xG** against **1.26 xGA**. They then won 3-2 away at Hoffenheim, creating **3.33 xG** while allowing **1.15**. Those were two genuinely productive attacking performances, not simply scorelines inflated by a handful of unlikely finishes.
The latest game was different. Beşiktaş drew 2-2 at home to Crystal Palace on 22 October despite producing only **0.67 xG** and allowing **0.40 xGA**. Both sides scored twice from relatively low-quality chances. The draw kept Beşiktaş unbeaten but should not be read as another dominant attacking display.
Their three-game league-phase totals are **nine goals from 6.93 xG** and **five conceded from 2.81 xGA**. The attack has generated plenty of opportunities overall, particularly against Marseille and Hoffenheim. The defensive goals-against record looks worse than the underlying chances conceded, which is important when assessing how vulnerable the visitors really are.
Beşiktaş have already shown that their attack can travel. The Hoffenheim win was their clearest demonstration, with more than three expected goals away from home. Celtic will need to avoid the defensive lapses that have turned manageable xGA totals into heavy European scorelines.
Where the match could change
The obvious contrast is in European chance creation. Beşiktaş have generated **6.93 xG** in three matches, more than double Celtic's **3.17**. But the visitors' goals conceded also warrant attention. Five goals allowed from 2.81 xGA suggests their opponents have converted chances efficiently, potentially leaving Celtic an opening if they can reproduce their stronger domestic attacking performances.
For Celtic, the first priority is to create better chances than they did against Celta Vigo. The second is to prevent Beşiktaş from reaching the high-value areas they exploited against Hoffenheim. The hosts cannot assume that a productive domestic xG performance will automatically carry into Europe, particularly against an opponent that has been much more effective in the competition.
Beşiktaş must avoid drawing too much confidence from the 2-2 Crystal Palace game. Their attack produced just 0.67 xG in that fixture, a reminder that their attacking level can fluctuate. If Celtic establish sustained pressure, the visitors may need to defend for longer periods than they did in their two wins.
For a wider look at the tournament's chance-quality trends, see SmartMatchData's Europa League xG analysis and guide to expected goals.
Match outlook
Beşiktaş enter with the stronger European record and more convincing league-phase attacking numbers. Celtic have home advantage and evidence of substantial domestic chance creation, but three straight European defeats make their situation difficult to dismiss as bad luck alone.
**Celtic's opportunity lies in raising their European chance creation while preventing Beşiktaş from turning the match into another open, high-scoring contest.** The visitors have shown they can punish defensive gaps, especially away at Hoffenheim, and will be confident of extending their unbeaten start.
No verified locked SmartMatchData model forecast was established for this fixture. Numerical 1X2, Over 2.5 and BTTS probabilities are not invented.
Data scope
The European comparison covers three completed 2026/27 Europa League league-phase games per team through 22 October. Celtic's domestic context includes Scottish Premiership matches through 1 November. The results and xG figures are from SmartMatchData's recorded match statistics. Unverified lineups, injuries and additional competition statistics are not assumed.