Data analysis definitions

Total
Total refers to the overall sum of certain metrics, such as total goals scored, indicating a player’s or team’s cumulative performance in those areas across a specific period. For example, a team scored 10 goals in total at a tournament.

Ranking
Ranking assigns players or teams a position based on their performance in a particular metric, indicating their standing relative to other players / teams. For example, if a team scored more goals than any other, they would be ranked number one. Metrics are ranked in either ascending or descending order depending on their nature. 

Average (mean)
Average refers to the standardised measure of a player or team’s performance, calculated by summing up the values of a specific metrics and dividing by how many values there are. An example of this is goals per game, if a team scores 18 goals in 6 games at a tournament, it means that on average, they scored 3 goals per game.

Difference from the average (mean)
Difference from the average (mean) measures the extent to which a player or team’s performances are different from the average (mean), indicating the diversity and distribution of performances against the rest of dataset. For example, if the average (mean) goals scored at a tournament is 4 and team A scored 6, then they scored 2 more goals than the average. 

Median
Median refers to the middle value of a set of numerical data, such as goals scored, if the numerical data was perfectly ordered. Using the median value instead of the average value is useful when there are exceptionally high or low values (like a particularly high-scoring game), because it reflects the exact midpoint, meaning the value is not distorted by outliers. For example, if a team played 5 games and scored 1, 2, 2, 2 and 8 goals in those matches the median value would be 2 goals. In this example the average would be 3 goals which is skewed by the 8-goal anomaly.

Percentile
The percentile of a particular performance metric refers to the percentage of other performances in the same sample / dataset that are below or equal to the measured value. This provides a comparative understanding of how a performance compares to other similar performances. For example, if a player’s goal-scoring is in the 90^th  percentile, it means their goal tally is equal to or better than 90% of the other players in the same category. 

Standard deviation
Standard deviation refers to the extent to which player or team performance metrics deviate from the average value in a dataset. A higher standard deviation indicates that values are spread out over a wider range, suggesting a greater difference in player or team performance. In contrast, a lower standard deviation indicates that values tend to be close to the average, suggesting consistency in their performances. For example, player passes in a match will have a high standard deviation, as some players will attempt significantly more passes depending on their position and how much possession their team has compared to others.

Per match
Per match refers to team statistics that are normalised to reflect their performance over a match, providing a fair comparison of team’s contributions regardless of the number of matches they’ve played. For example, if a team scores 12 goals and plays 6 matches, they have scored 2 goals per match.

Per 90 minutes
Per 90 minutes refers to player or team statistics that are normalised to reflect their performance over a standard 90-minute match, providing a fair comparison of players’ or team’s contributions regardless of the actual minutes they played / the different lengths of matches. It gives more context than per match metrics, as it accounts for the different lengths of matches (e.g., matches that go to extra time). For example, if a player scores 4 goals and plays for 180 minutes across 7 matches, they have scored 1.8 goals per 90 minutes, but 0.6 goals per match. 

Per 30 minutes in / out of possession
Per 30 minutes in / out of possession refers to player or team statistics that are normalised to reflect their performance when their team is in / out of possession. The ball is in play for around 60 minutes of a standard 90-minute match. If possession is normalised to be a 50:50 split, this gives an approximate time in and out of possession of 30 minutes. This can provide additional context over per 90-minute metrics as it accounts for the actual time able to perform actions (e.g. a team can only pass the ball when they are in possession). This enables stylistic comparisons between teams’ in and out of possession performances / strategies. For example, if a player presses their opponents 15 times in a match but their team was only out of possession for 20 minutes, they pressed their opponents 22.5 times per 30 minutes out of possession. 

Completion rate
Completion rate represents the percentage of successful actions completed by a player or team, indicating their accuracy and effectiveness. For example, if a team attempts 400 passes in a match and 340 of those reach their intended destination, the pass completion rate is 85%. 

Proportion
Proportion represents the ratio of specific event categories compared to the overall number of instances. For example, if a player performed 1 driven pass, 6 push passes and 3 lofted passes (10 total passes), the proportion of their passes that were lofted is 30%. 

Percent / proportion of time
Percent / proportion of time represents the share of a match / period in which a player or team spends in a given phase of play or possession state. This can help suggest a team’s style of play or indicate how possession was distributed during a match. For example, if a team was in possession for 20 minutes, out of possession for 25 minutes and the ball in contest for 5 minutes during a match, the percentage / proportion of time in possession, out of possession and in contest would be 40%, 50% and 10% respectively.