Premiership · Scotland · Regular Season - 1

Dundee Utd–Rangers

1–1

HT 1–0

Kick-off: Friday, 31 July 2026 at 19:00
Status: Match Finished
Venue: Tannadice Park, Dundee

What the model said, and what happened

Ended level — we gave the draw 23%double chance: correct

This verdict goes into the public record unchanged; we do not filter out the ones we get wrong. The model's track record

Goals and red cards

38'Dundee UtdL. Rose (assist: J. Randall)
56'RangersThelo Aasgaard

Match statistics

Match statistics
Dundee Utd		Rangers
11	Shots	21
3	Shots on target	7
10	Shots inside the box	16
6	Goalkeeper saves	2
5	Corners	10
1	Offsides	2
45%	Possession	55%
299	Passes	362
210	Accurate passes

What we expected, and what happened

Expected goals before kick-off: Dundee Utd 1.21 · Rangers 2.03. On the pitch: 1-1.
Shots Dundee Utd 11 · Rangers 21; on target 3 and 7.
Rangers kept the ball more (55%).

Referee

Nick Walsh, Scotland→

4.34

yellows per match

0.16

reds per match

matches of data

Model prediction

Why this prediction?

The model puts the most likely outcome at 56.4%: “Rangers to win”. Second comes “a draw” at 22.9%.
Behind that distribution are the expected goals: 1.21 for Dundee Utd, 2.03 for Rangers.
The highest double chance value is 79.2% (X2 — draw or Rangers win).
These values were computed from 867 past matches.

This is a prediction, not a guarantee; a probability is not a certainty.

These sentences are built by a fixed template; every number in them comes straight from the model's computation, none invented while the text was written.

Dundee Utd to win20.8%

Draw22.9%

Rangers to win56.4%

When the model has said 55–60% in the past, 1,158 of 1,982 predictions were correct (58.4%). Accuracy record →

16	Fouls	14

1	Yellow cards	3

0	Red cards	0

First half

Actual: 1–0

Dundee Utd lead at half-time19.3%

Level at half-time39.4%

Rangers lead at half-time41.3%

First-half double chance

1X — Dundee Utd ahead or level58.7%

12 — not level at half-time60.6%

X2 — level or Rangers ahead80.7%

These values are not a fraction of the full-match prediction; they come from a separate model trained on the first-half scores of 867 matches. Its edge over the base rate is markedly smaller than the full-match model's — the measured figure is on the accuracy record. Read them more cautiously. First-half goal markets are not published because they failed to beat the base rate.

Expected fouls and corners

Fouls

Dundee Utd: at least 12Actual: 16
Rangers: at least 9Actual: 14

Corners

Dundee Utd: at least 3Actual: 5
Rangers: at least 5Actual: 10

These figures are set to hold in at least 60% of matches; in the rest the count comes in below them. No upper bound is stated — we make no "at most" claim. Yellow cards were measured and are not published: there our model did not beat the historical average.

What these probabilities rest on

Expected goals comes out of a team's attack weighed against its opponent's defence. A strong attack is damped when it meets a strong defence — which is why comparing the ratings on their own is misleading.

How many goals Dundee Utd score1.21

Attack: Dundee Utd score close to the league average.

Opponent's defence: Rangers concede 18% fewer than the league average, which pulls the expectation down.

Result: These two effects give an expectation of 1.21 goals (including home advantage).

How many goals Rangers score2.03

Attack: Rangers score above the league average (46% more).

Opponent's defence: Dundee Utd concede 20% more than the league average, which pushes the expectation up.

Result: A strong attack meets a weak defence. The effects compound and the expectation rises to 2.03 goals.

Most likely scorelines

1–110.8%1–210.8%0–28.1%

A single scoreline is always less likely than an outcome: “home win” is the sum of 1-0, 2-0, 2-1 and dozens more. So the most likely single score can be a draw while the most likely outcome is a home win; the two do not contradict each other.

Week 1 of the season. This prediction was computed from previous seasons. Summer transfers, budget changes and squad rebuilds aren't in the ratings yet — the model can only see that a team has been rebuilt once new results come in.

We measured this and we're not hiding it: in the opening week of past seasons the model came in 3.3 points below the accuracy it claimed (n=220) — for the rest of the season it beats its own claim. The tendency points this way, though it is not proof on its own: in the opening week it slightly over-calls home wins (said 42.7%, got 39.5%) and slightly under-calls draws (said 26.1%, got 28.2%).

This sample is small and not statistically established, which is why we do not correct the model for it. The correction we tried (pulling probabilities toward the league average) didn't win on holdout, so it wasn't added. Read opening-week predictions with a little more caution.