Match analysis
Manchester United 1–1 Tottenham Hotspur: what changed at minute 73?
Manchester United's win probability moved from 24.8% to 87.6%. Compare the match sequence with 0.77–1.57 xG.
By xG Lab ·
The recorded evidence
On 10 October 2026, Manchester United and Tottenham Hotspur finished 1–1. The verified goal sequence was: Manchester United at minute 73 (1–0); Tottenham Hotspur at minute 90 (1–1).
At kick-off the model assigned Manchester United 54.8%, the draw 24.7% and Tottenham Hotspur 20.6%. At minute 72, the home win probability was 24.8%; at minute 73 it was 87.6%. This was the largest consecutive home probability change before the settled endpoint, at 62.8 percentage points. Values are independently rounded. At the later checkpoint the draw was 11.2% and the away win probability was 1.1%.
Manchester United recorded 0.77 xG and Tottenham Hotspur recorded 1.57. Total xG sums the scoring probabilities of chances; it does not establish average shot quality or a deserved result. The replay model uses pre-match strengths, score, remaining time and recorded red cards, rather than updating from live xG. The final checkpoint is settled by the actual result.
Reading the minute 73 movement
The question in the title has a narrow answer in the recorded data. The goal timeline lists Manchester United at minute 73 (1–0); Tottenham Hotspur at minute 90 (1–1). The largest checkpoint change, 62.8 percentage points, sits at minute 73, and that is the same recorded minute as the Manchester United goal. That is an observation about matching minutes. This analysis does not add any cause beyond the score change.
The sequence also explains why the earlier checkpoint looks unusual. At minute 72, the model had Manchester United at 24.8%, the draw at 62.4% and Tottenham Hotspur at 12.8%. After the recorded score change to Manchester United ahead, the checkpoint at minute 73 showed 87.6% for Manchester United, 11.2% for the draw and 1.1% for Tottenham Hotspur. The later Tottenham Hotspur goal at minute 90 leads to the settled endpoint: 0.0%, 100.0% and 0.0%. That endpoint reflects the result of 1–1, not a pre-whistle estimate.
What xG can and cannot add
The xG totals, 0.77 for Manchester United and 1.57 for Tottenham Hotspur, are a separate lens. They sum the scoring probabilities of recorded chances. They do not show average shot quality, and they cannot undo the actual goals. They also did not feed the probability curve, which uses pre-match strengths, score, remaining time and recorded red cards. A possible interpretation is that the result and the xG totals tell different stories. The recorded data cannot say whether that reflects finishing, chance quality, game state or something else.
The table shows selected checkpoints, not the full series. The scan covered every sampled checkpoint, but the intermediate values between those rows are not displayed here. Checkpoint times are recorded minutes, so no claim about seconds or duration is possible.
For a next check, open the full match replay and compare the xG timeline with the minute 73 and minute 90 score changes. Then read the model methodology to confirm which inputs move the curve. Look at when each side's chances occurred relative to the Manchester United goal. Treat any pattern you find as a question for further study, not as proof of what should have happened.
The data behind this analysis
| Minute | Manchester United win | Draw | Tottenham Hotspur win |
|---|---|---|---|
| 0′ | 54.8% | 24.7% | 20.6% |
| 72′ | 24.8% | 62.4% | 12.8% |
| 73′ | 87.6% | 11.2% | 1.1% |
| 90′ | 0.0% | 100.0% | 0.0% |
Method, sources and limitations
This article uses a fixed snapshot captured on 2026-10-19. The latest data used is dated 2026-10-10. The snapshot stays fixed as the underlying tools update.
xG values use xG Lab’s licensed TheStatsAPI feed; other providers may produce different values. Probabilities come from xG Lab’s model using pre-match strengths, score, remaining time and recorded red cards. The model does not update strengths from live xG. The replay follows recorded minute checkpoints, including available stoppage time, with a settled final endpoint. Figures are rounded.
Short samples are descriptive and are not adjusted for opposition strength. A goals-minus-xG gap does not by itself isolate finishing skill or forecast future results. Pre-match probabilities are uncertain snapshots and may change with later data.
Prepared with AI assistance. Numeric statements are checked against the snapshot above; interpretations should be read with these limitations.