Skip to main content
Back to work
RecruitmentDashboard

Case study

Midfielder Scout 2025/26

A role-based Streamlit scouting dashboard for shortlisting midfielders across Europe’s top five leagues.

Evidence structure

Midfielder Scout 2025/26

01

Problem / Question

How can a recruitment analyst move from a broad midfielder market to a role-specific, financially realistic shortlist before committing time to video review?

02

Context

Independent recruitment analytics project built around Europe’s top five leagues for the 2025/26 season, focused on chance creation and midfield role fit.

03

Dataset / Video Sources

WhoScored match event data transformed into player-season metrics; Transfermarkt squad pages for market value, contract expiry, and transfer-feasibility tiers.

04

Methodology

Built an offline Python pipeline for fixture extraction, event parsing, player/match/team table construction, per-90 normalization, percentile ranking, role scoring, cross-league merging, and Transfermarkt enrichment.

05

Outputs / Deliverables

Interactive Streamlit dashboard with global filters, ranked shortlist, role taxonomy, player scout reports, radar charts, similar-player matching, match logs, comparison views, and league identity analysis.

06

Decision Supported

Supports recruitment analysts, scouts, and sporting directors in deciding which midfielders deserve deeper video scouting, live observation, or transfer-feasibility checks.

07

Limitations / Assumptions

Public event data lacks full tactical context, opponent adjustment, tracking data, and off-ball movement. Carry metrics are inferred from event sequences rather than native carry events, so outputs should guide shortlist generation rather than replace video validation.

08

Next Step

Add opponent-quality adjustment, pass-value or xA modelling, and a video annotation layer for shortlisted players.

Narrative notes

Traditional scouting often starts with a name: someone watches a match, likes a player, and the analysis begins from there. This project reverses that flow. It starts with a recruitment question: what kind of midfielder do we need, what constraints matter, and which players fit before video time is spent?

Midfielder Scout 2025/26 profiles qualified midfielders from Europe’s top five leagues using WhoScored match event data. The pipeline turns raw match events into per-90 metrics, percentile ranks, and four role scores: Creator, Ball Progressor, Box Threat, and Deep Builder. Transfermarkt enrichment adds market value, contract expiry, and a simple transfer-feasibility layer, so the shortlist can reflect football fit and market realism at the same time.

The Streamlit app is designed as a recruitment workflow rather than a static ranking. Filters narrow the pool by minutes, age, position, league, team, role, market value, percentile mode, and feasibility. The Shortlist tab surfaces ranked candidates; Role Map explains the role taxonomy; Scout Report turns one player into a profile with radar, role ratings, similar players, and match log; Compare and Explore support head-to-head review and deeper metric lenses.

The result is a decision-support tool for the stage before full video scouting: define the role, filter the market, inspect the profile, then decide which names deserve deeper review.

StreamlitShortlistEvent dataRole scoringMidfielder scouting