Neurotactic specializes in data-driven football analysis, offering insights for individual players, clubs, and national teams. Our expertise helps optimize performance, strategy, and decision-making by leveraging advanced analytics to unlock the full potential of players and teams. Earlier, we used to make player reports using Tableau which took atleast1-2 days to make. I was tasked with created an quick automated data report generator that can generate the report in 3-4 mins by simply running some code.
- Used for ad-hoc analysis as required by our team's player performance analyst.
- Increased efficiency by reducing the time needed to create reports compared to manually building them in Tableau.
- Made using Python and libraries such as Pandas and Matplotlib.
The data used in from Wyscout. It is tabular with multiple columns containing different type of player stats.
Used to identify key player strengths, and visualize their standout attributes.
-
Key Features
Filtering: Select player positions and apply filters for age and minutes played.
Statistical Analysis: Compute percentile and z-scores for each statistic.
Visualization: Generate bar and radar plots for highlighting a player's best stats. 4 colors are used to represent the percentile of the player stats. Easy to understand for our clients. Radar shows comparision between player and average.
Can find player strengths, calculates aggregate statistics, and finds similar players represented by a similarity score.
-
Key Features
Statistical Analysis: Compute percentile ranks, z-scores, and aggregate stats using custom-weighted combinations of individual stats. Aggregate stats are calculated differently for each position using different weights. Weights are assigned based on our knowlegde of the game.
Player Comparison: Use cosine similarity to find players with similar performance profiles. Top 5 similar players are selected.
Visualization: Generate table and radar plots for comparision of stats. Table contains the aggregate stats and radar compares the main player and another player from the table.
