Wednesday, 28/10/2026 at 20:45
Full timeCeltic 2:2 Dundee United
Celtic were the model's likeliest winner before kick-off at 64%. It finished 2–2. The biggest difference was Celtic's attack: 1.95 xG against 2.30 projected (−0.35). Dundee United produced 0.78 xG against 1.10 projected (−0.32).
Goals
- 18' Camilo Durán Celtic · 0.28 xG chance 1–0
- 28' Zachary Sapsford Dundee United · 0.04 xG chance 1–1
- 31' Neil Farrugia Dundee United · 0.14 xG chance 1–2
- 65' Liam Scales Celtic · 0.04 xG chance 2–2
The model against the result
- Celtic
- 64%
- Draw
- 22% What happened
- Dundee United
- 14%
Data coverage Both teams have the full 20-match model window available.
Projected vs produced
Celtic produced 0.35 xG less than projected
Projected before kick-off Produced on the night
The shots behind it
- Shots
- 29
- 11
- On target
- 8
- 5
- Inside the box
- 15
- 7
- xG per shot
- 0.07
- 0.07
- Open play xG
- 1.34
- 0.57
- Set-piece xG
- 0.60
- 0.23
- Penalties
- —
- —
Celtic's shots add up to 1.94 xG and Dundee United's to 0.80; the published figures are 1.95 and 0.78. Leaving out penalties and own goals, Dundee United's shots come to 0.80 against a published non-penalty xG of 0.79: more than rounding can explain, so the two figures were not read at the same point of the provider's revisions. The provider computes a match's total xG and its non-penalty xG separately, so the difference between the two is not what a penalty was worth.
Shot map
Both teams shown attacking the same goal for comparison.
All shots
40 shots · 2.74 xG
Hover or tap a shot, or step through them with the arrows.
- Celtic
- Dundee United
- Filled = goal
- Dot area = xG (the smallest chances drawn at a minimum size)
Figures read 7 November, 17:06 · final
Recent form
2.26 xG created 1.31 xGA
- W 3.19 created, 1.95 conceded against St. Mirren
- W 1.80 created, 0.69 conceded against St. Johnstone
- L 0.39 created, 1.75 conceded against Rangers
- W 2.85 created, 1.05 conceded against Motherwell
- W 4.16 created, 1.20 conceded against Heart of Midlothian
- W 1.15 created, 1.19 conceded against Hibernian
1.48 xG created 1.38 xGA
xG created xG conceded (xGA) W / D / L: the match result
Head to head
Pre-match model & market record
9 signals flagged4 won5 lost
- Lost
- Won
- Lost
- Lost
- Lost
- Won
- Won
- Won
- Lost
View pre-match record Hide pre-match record
| Selection | Price | Model-implied edge | Result |
|---|---|---|---|
| Dundee United to win | 10.00 | +41.9% | Lost |
| Draw | 6.25 | +35.1% | Won |
| Celtic at most 1 goal | 4.33 | +33.1% | Lost |
| Celtic not to score | 15.00 | +24.2% | Lost |
| Under 2.5 goals | 3.75 | +18.9% | Lost |
| Dundee United 2+ goals | 3.75 | +14.4% | Won |
| Dundee United to score | 1.53 | +7.2% | Won |
| Both teams to score | 1.62 | +5.4% | Won |
| Under 1.5 goals | 9.00 | +2.2% | Lost |
Pre-match model vs market
| Outcome | Model | Market | Disagreement | Model price |
|---|---|---|---|---|
| Celtic | 64.2% | 75.5% | −11.3 pp | 1.56 |
| Draw | 21.6% | 15.1% | +6.5 pp | 4.63 |
| Dundee United | 14.2% | 9.4% | +4.8 pp | 7.04 |
Other markets
| Outcome | Model | Market | Disagreement | Model price |
|---|---|---|---|---|
| Celtic to score | 91.7% | 93.6% | −1.9 pp | 1.09 |
| Celtic not to score | 8.3% | 6.4% | +1.9 pp | 12.08 |
| Dundee United to score | 69.9% | 60.8% | +9.1 pp | 1.43 |
| Dundee United not to score | 30.1% | 39.2% | −9.1 pp | 3.32 |
| Both teams to score | 65.3% | 57.7% | +7.6 pp | 1.53 |
| Both teams not to score | 34.7% | 42.3% | −7.6 pp | 2.88 |
| Celtic 2+ goals | 69.3% | 78.3% | −9.0 pp | 1.44 |
| Celtic at most 1 goal | 30.7% | 21.7% | +9.0 pp | 3.26 |
| Dundee United 2+ goals | 30.5% | 25.0% | +5.5 pp | 3.28 |
| Dundee United at most 1 goal | 69.5% | 75.0% | −5.5 pp | 1.44 |
| Over 1.5 goals | 88.6% | 89.4% | −0.7 pp | 1.13 |
| Under 1.5 goals | 11.4% | 10.6% | +0.7 pp | 8.81 |
| Over 2.5 goals | 68.3% | 74.5% | −6.2 pp | 1.46 |
| Under 2.5 goals | 31.7% | 25.5% | +6.2 pp | 3.15 |
Prices and probabilities are the ones stored before kick-off; nothing here is recalculated. The model-implied edge is the model's probability times the price, minus one: what a stake of one would return on average if the model is right, not observed return, so a flagged outcome that lost is the ordinary case rather than a contradiction.
Model details
Show Hide
- Fit window
- the last 20 matches, recency weighted (a match 107 days old counts half)
- Celtic sample
- 20 matches
- Dundee United sample
- 20 matches
- Computed
- 2 hours before kick-off
- Prices taken
- 2 hours before kick-off
Limitations
The model reads expected goals and nothing else. It does not know the lineups, who is injured or suspended, what either side has to play for, the weather, or how congested the schedule around this match is.