Hello, this is very interesting
Could we set this up as a model to see the results needed for 80 points? For example, Tottenham are on 40pts with 17 games to go and have won over half their games so far; that ratio would give them at least another 27 points plus a few draws, which makes me think 72 might not be enough to finish ahead of them. Everton have only lost twice, so even if they only win 50% of the remaining games and draw most of the rest, they will probably be above 72. On the plus side our GD is far superior so counts as an extra point.
I wouldn't be surprised if MU went on a run but they might just be too far back. However, despite scoring a lot we have conceded more goals than all our rivals, which shows some vulnerability in defence as well as an attacking mindset.
We have to consistently do better than the three teams below us, so another algorithm would take their matches and remaining fixtures and model our required results based on the latest scores.
We might need at least 12 more wins - I think there are 10 or so "should-win" games left for us both home and away, so the crunch matches against the rivals are key. If we beat Spurs, Man City, Everton, Arsenal and Chelsea then we make up ground. I also don't know if I want the other teams to progress in Europe - might put strain on their squads, but at the same time give them confidence and momentum.
Nice statistical approach here!