Manchester United 4 Newcastle 3 | Match Review using shot strength stats

There will be hundreds of reviews of the Manchester United v Newcastle game discussing formations and showing how exciting the game. I would like, then, to show the raw data, mapping the game, to show that it is possible to predict the expectation of goals during a game by looking at shots as time decays.

Background data before the game started

1. Manchester United won 7 and lost 3 when conceding first, so we know they have strength of character to not lose discipline or concentration when the game does not start well for them. They have already won 70% of their games when conceding first.
2. An early away goal in a game is a significant event. When an away team scores first between kick of and 35′, the expectation of further goals increases to a level greater than the initial pre-game prediction.
3. One should always expect a late goal, even if there is little evidence in a game to suggest that this will happen, as more late goals are scored than during any other time band. In terms of time decay, goal production increases and spikes at 45-HT  and again at 9o-FT, but if you look at a graph you will see a steady increase  as time decays.

There is no expectation that Newcastle would score the early goal, but once they did it mapped out expectation of further goals. Manchester United started slowly and Newcastle dominated early exchanges. Evans’ equaliser to make it 1-1, then, was against the run of play. It was no surprise that Manchester United conceded again so soon after the initial “fight back”.

The second Newcastle goal, which was given as a Evans own goal, was controversial and, although Newcastle were still the better side and deserved to be in front, it shows the need for technology in football as the linesman flagged for offside and the referee Mike Dean decided to take control and award a goal.