Celtic vs Rangers: Old Firm Preview by the Numbers
Celtic vs Rangers returns at Celtic Park. We examine first goals, set pieces, central chance quality, discipline and schedule fatigue.
Celtic host Rangers at Celtic Park on September 20 in a rematch shaped by two very different signals: Celtic are first with 18 points from six Scottish Premiership matches, while Rangers sit second on 13. Yet Rangers won the clubs’ most recent meeting 3-0 in the League Cup quarter-final, and Celtic arrive after a 3-1 Europa League defeat by Ferencvaros. This Old Firm preview therefore needs more than the league table. Rivalry volatility can make raw form a poor guide unless the model separates repeatable performance from derby-specific events.
Celtic vs Rangers Form Split
Celtic’s league numbers remain the strongest in the match-up. Martin O’Neill’s side have won all six Premiership fixtures, scoring 14 goals and conceding three, with an average of 2.33 goals scored and 0.50 conceded per game. They also lead the available efficiency indicators: 61.6% possession, 65.2% field tilt and 9.5 shots on target per match.
Rangers’ league record is 4-1-1, with 13 goals scored and four conceded. Their averages are lower in attack—1.17 goals and 6.5 shots on target per game—but their defensive profile is close to Celtic’s, with 0.67 goals conceded per match and three clean sheets. Derek McInnes’s side have also won their last five matches across competitions, including the 3-0 derby victory.
The key modeling point is sample conflict. Celtic’s six-match league sequence says control and chance volume; the latest head-to-head says Rangers found a way to disrupt that control. The 3-0 result should not erase Celtic’s season-long superiority, but it does increase the uncertainty around any simple home-win projection.
Why the First Goal Matters
Public pre-match data does not provide each team’s first-goal record or score-state expected goals, so a precise first-goal multiplier would be invented rather than measured. The available evidence still points to a major state change. Celtic’s possession and field-tilt advantages suggest they are more likely to establish territory early, while Rangers’ recent 3-0 win demonstrates the value of converting disruption before Celtic settle.
For a ScorePoint AI-style model, the useful question is not simply who scores first, but how each team performs after the state changes. Celtic’s 14 league goals indicate attacking depth, while Rangers’ four conceded in six matches suggest they can protect a lead. A Rangers first goal would likely make Celtic’s midweek fatigue more important; a Celtic opener would test whether Rangers can create enough volume from a deeper or more cautious game plan.
Central Chances and Set Pieces
The public figures strongly favor Celtic in overall attacking territory, but they do not include expected goals split by central versus wide locations. That limitation matters. Celtic’s 9.5 shots on target per game and 61.6% possession show pressure, not necessarily the quality of chances against Rangers’ compact defensive structure. The model should therefore avoid treating possession as a direct substitute for central-area xG.
Set pieces offer a measurable secondary route. Celtic have won 58 corners in the league data supplied, compared with Rangers’ 48. No source figure identifies goals from corners or free kicks, so corner totals should be treated as opportunity volume rather than proof of set-piece superiority. Cameron Carter-Vickers and Liam Scales give Celtic established defensive targets, while Rangers can draw on Olwethu Makhanya’s aerial presence after he impressed in the previous derby. The key distinction is delivery quality and second-ball recovery, not the corner count alone.
Discipline and Schedule Load
Rangers have committed 78 fouls to Celtic’s 60, a difference of 18 across six league matches. That is a meaningful disciplinary-risk signal in a derby where repeated midfield challenges can affect pressing, set-piece concessions and the availability of aggressive defenders. It is not a prediction of cards—the supplied data contains no card totals—but it does justify a higher volatility adjustment for Rangers’ defensive actions.
Schedule fatigue is the clearest contextual edge for Rangers. Celtic played their League Cup quarter-final against Rangers on September 13 and then lost 3-1 to Ferencvaros on September 17, leaving only three days before this league meeting. Rangers have also played since the cup derby, but their last listed match before the weekend was a 1-0 league win over St Mirren on September 9, giving McInnes’s team a longer recovery window. That gap should be reflected in projected pressing intensity, especially for Celtic after the hour mark.
Celtic are also without Alistair Johnston because of a leg injury, while Callum Osmand, Landon Emenalo and Yang Hyun-jun are unavailable for the reasons reported before the match. Rangers remain without long-term absentees Lawrence Shankland, Youssef Chermiti and Daisuke Yokota; Ryan Naderi is a doubt. Celtic’s Camilo Duran leads their league scoring and assisting figures with three goals and three assists, while Shankland had three goals in five league appearances before his injury.
Old Firm Model Outlook
Celtic’s home advantage, unbeaten league record and superior attacking indicators support them in this analysis. Rangers’ recent 3-0 win, lower foul discipline, fresher schedule and stronger immediate momentum prevent a routine projection. The practical outlook is a match in which the first goal has unusually high leverage: Celtic need their possession to become central chances, while Rangers need their pressing and transition moments to offset Celtic’s territorial control. Any prediction should widen its uncertainty band rather than rely on the six-match table alone.
Research references
These sources were checked while preparing this ScorePoint AI analysis.
