Library Guide

Welcome to DataScouteR

DataScouteR is an R package for football scouting and player evaluation, providing league-adjusted, position-specific ratings derived from publicly available data and scaled for intuitive interpretation.

The package is designed to offer a clean, ready-to-use interface for analysts, researchers, and practitioners who want access to processed scouting outputs, without needing to run complex models themselves.

Installation

install.packages("remotes")
remotes::install_github("machoude/DataScouteR", subdir = "DataScouteR")

Players

library(DataScouteR)
gk <- get_gk()
head(gk)

df <- get_df()
head(df)

mf <- get_mf()
head(mf)

fw <- get_fw()
head(fw)

Data Source

All underlying match and event-level statistics are sourced from:

  • FBref

FBref provides high-quality, publicly accessible football statistics.
DataScouteR uses these data as inputs but does not redistribute raw FBref datasets.

Methodology

The workflow behind the package follows these principles:

  • Player performance metrics are modeled using z-scores and logistic mapping
  • Player-level effects are filtered to reduce noise from small samples
  • League-level effects are included to implicitly account for competition strength
  • Final latent ability estimates are transformed to a 1–20 scale using a smooth, monotonic mapping

This scaling is not intended to replicate Football Manager internally, but rather to:

  • Provide a familiar reference frame
  • Improve interpretability for practitioners

Why a 1–20 Scale?

The 1–20 scale:

  • Preserves relative differences between players
  • Compresses extreme values smoothly
  • Allows quick qualitative interpretation (e.g. elite, above average, poor)

This makes the results more accessible for:

  • Scouting discussions
  • Player profiling
  • Comparative analysis across leagues

Interpreting the Metrics

The ratings in DataScouteR are scaled from 1 to 20 FM-style for easier comparison. A player’s score shows how well they performed this season, relative to other players in the dataset.

For example, if a goalkeeper has a 16 in ShotStopping while Alisson has 8, it does not mean the player is generally better than Alisson. It only indicates that, in the current season, this player performed better in shot-stopping relative to the other goalkeepers in the dataset.

Also get_fw() and get_md() use similar metrics, the ratings are position-specific. A Scoring rating of 15 for a midfielder does not carry the same meaning as a 15 for a forward, since all scores are calculated relative to players in the same position group.

This scale is designed for seasonal comparison, not historical ranking. High numbers mean above-average performance this year, while lower numbers indicate below-average performance for the same season.