The final 20 minutes of a football match produce more goals per minute than any other period, for reasons that follow patterns. Physical: by 70 minutes players have covered eight to nine kilometres, pressing intensity fades, and the tight marking from minute 20 has loosened. Tactical: trailing teams commit players forward while fresh substitutes run at tired defenders. Psychological: desperation increases risk-taking on one side while the leading team drops deep and invites pressure.
The market knows goals become more likely late, but it prices elapsed time heavily and can underweight the dynamics of the final period. Odds on goals markets drift steadily as the clock runs without a goal. That drift sometimes misses the difference between a game winding down and a game where the conditions for a late goal are stacking up: sustained pressure, accumulating chances, attacking substitutions, a one-goal margin forcing the issue. The late-goals analysis looks for those conditions, and for the time remaining alongside them.
The 70-minute threshold is deliberate. Earlier, too much can still change: substitutions are still coming and the tactical picture is unsettled. Later than about 85 minutes, the window is too short to be useful, multi-goal markets stop being plausible, and even single-goal positions are compressed. Between roughly 70 and 85 minutes you get a clear tactical picture, measurable fatigue, and enough runway for a goal to arrive.
The discipline cuts both ways. When the conditions are not present, the right output is no recommendation. Past a certain point the opportunity has gone, even if a goal eventually arrives.