With event level metrics, you might analyse players regaining possession of the ball with metrics such as interceptions, ball recoveries or tackles. Player sequence metrics aren’t just valuable for attackers. Sequence involvements allow us to value a player’s importance to the team even if they didn’t perform strongly in their traditional numbers. Even if we remove the times where Sterling played the pass immediately before the shot or took the shot himself, we find that he ranks in the top five percent of players. Despite only registering one assist in the Premier League 2019-20 season, his teammate Kevin De Bruyne was the only player to outperform him in this metric (291). Manchester City’s Raheem Sterling was involved in 221 unique sequences that ended in shots. Sequence Metrics for PlayersĪs we alluded to earlier, we can quantify a player’s involvement in sequences that end in shots. We can aggregate this context for a team or player and define even more metrics upon this information to give us new ways to evaluate their performances. We have seen that the sequence framework can provide context about an event that isn’t possible through event data alone. Within this sequence, there were six players involved and we can identify the number of passes in the build-up to this shot (5). The sequence traced between all events covered 142 metres, or 72 metres if we consider only the distance progressed directly up-field. We can see that Bernd Leno (number one) initiates the sequence in his own area and that nearly 22 seconds elapsed in the build-up to Ødegaard’s shot (number eleven). After a succession of short carries and direct passes through midfield, Kieran Tierney beats Matt Doherty and squares it for Ødegaard to score. The sequence starts with Arsenal’s goalkeeper, Bernd Leno, whose pass is a controlled action that initiates the sequence following an unsuccessful flick on by Harry Kane. Let’s consider the passage of play leading to Arsenal’s Martin Ødegaard’s first Premier League goal in the North London derby against Tottenham. We will focus on sequences for our examples given that they allow for increased granularity in our analysis and derive more intuitive metrics for players and teams.
#The sequence series#
A possession is ended by the opposition gaining control of the ball.Ī series of uninterrupted passes leading to a shot would be counted as a sequence.
Possessions are defined as one or more sequences in a row belonging to the same team. This includes passes but not defensive events such as tackles and interceptions, unless these events are followed by a controlled action such as a pass or dribble. A sequence starts with a player making a controlled action on the ball. Sequences are defined as passages of play which belong to one team and are ended by defensive actions, stoppages in play or a shot. While secondary assists (the pass before an assist) provide some of this context, we can now go even further back with sequences and see who is starting these moves or who is most frequently involved before these final actions. By analysing the sequences of events that constitute a period of possession, we can create and utilise a whole suite of new metrics to gain insight into a team’s playing style or an individual player’s contribution.īy stringing events together, we are able to value the contributions of the players who don’t necessarily score or assist goals but are still integral in the build-up. For more than 20 years Stats Perform has recorded detailed event data to provide a snapshot of every action at any moment during a football match.