Note: The data that powers the chart above includes ~4.5k Premier League games since 2000. At each minute, you can see the win/draw/lose probabilities above the graph (as well as the number of games that go into that calculation). To make the trend lines less noisy, only scenarios that occurred in at least 5 different games are plotted (e.g., there were not 5 games when the home team had a 3 goal advantage within the first 17 minutes).
A few observations:
If the score is tied and your team is playing at home, you may want to start lowering your expectations for a victory around the 70 minute mark - this is when a draw becomes more likely than a win.
How comfortable is a one goal advantage at home? Pretty comfortable actually - teams go on to win over 70% of the time even if that lead came in the 2nd minute. That said, the win likelihood doesn’t cross 80% until the 70th minute and it doesn’t hit 90% until the 85th minute.
A one goal lead is not nearly as cushy when you’re away - a 70% win rate doesn’t happen until the 68th minute for away sides that go up by one.
What about a two goal advantage? Home or away, a two advantage in the second half leads to a victory over 90% of the time.
It’s interesting to see some of the rare and crazy comebacks that happened over the years, like the home team that was down by 3 in the 80th minute and went on to secure a draw (this was West Brom (h) vs. Man U (a) in 2013, summary here).
Feel free to play around with the inputs and hover over the chart to explore the scenarios you’re curious about, or maybe the one your team happens to be facing during a match.
Oh, and in case you missed the Manchester derby earlier today, enjoy :)