How a modern professional club's scouting department works

Drop the radar on the left side of the penalty area for the U-18 Portuguese league and filter by progressive carries > 7.5 per 90. That single query, run at 09:07 on a Tuesday, returned three names; the third belonged to a 17-year-old still playing for Portimonense B. By 09:21 the analytics chief had tagged 38 video snippets, calculated a 0.73 xG-adjusted conversion rate against top-tier youth defenders, and fired the file to the head of recruitment. At 10:04 the player’s estimated market value jumped from €1.4 m to €2.1 m inside the internal pricing model because his sprint profile matched the gap left by the departing first-team wide man.

Scouts now sit down with Bayesian dashboards that refresh every four minutes while matches are live. The algorithm weighs acceleration peaks, defensive regains within eight seconds of a loss, and pass reception angles against historical benchmarks for 14 000 wingers. If a teenager’s three-game rolling score crosses the 82nd percentile of the squad’s current starters, an alert hits the sporting director’s phone before half-time. Last season this happened 61 times; seven visits were arranged, four contracts offered, two signatures secured, generating a combined resale surplus of £38 m within 18 months.

Contracts are no longer gambles. They are receipts built on 312 discrete performance markers, medical cartilage scans, and psychological resilience indices benchmarked against 1 200 European first-team regulars. When the Lisbon teenager arrived at Carrington for a 48-hour assessment, he wore a GPS vest during a behind-closed-doors friendly. His high-speed distance (5.3 km) and number of decelerations > 3 m/s² (46) beat the club average by 11 % and 18 % respectively. The deal sheet was printed before he boarded the return flight.

Inside a Pro Club Scouting Office: From Data to Signed Player

Filter the Wyscout export for U-21 midfielders with >7.0 defensive duels won per 90 and >0.28 expected assists; anything lower hits the trash folder. Manchester City’s recruitment wing deletes 83 % of names at this stage, leaving roughly 120 global targets for human eye checks.

Each survivor gets a 12-match video pack clipped by three analysts working in four-hour shifts; they tag every misplaced first-touch, sprint after turnover, and off-ball run, then feed the tags to a Python model that spits out a 0-100 adaptation score. Last winter, Bayer Leverkusen binned a 19-year-old Brazilian who topped the statistical sheet because his score stuck at 61; the threshold for Bundesliga gametime is 74.

Live trips come next: a two-man team flies economy, sits on the uncovered stands, and times the prospect’s recovery runs with a $199 laser gun; they also record heart-rate via the WHOOP strap loaned to the player’s agent the night before. If the average sprint recovery exceeds 43 seconds in the final quarter-hour, the report flashes red and the head of talent drops the dossier to monitor for six months.

Contract: the analytics chief presents a 20-slide deck to the board-wage curve, resale probability, injury algorithm-then opens a sealed envelope holding the single number the negotiation cell is authorized to offer. When Frankfurt closed for Hugo Ekitike, that envelope read €24.5 m base plus €9 m after ten league goals; the deal passed 11-0 in six minutes.

Triggering the Search: How the Shortlist Gets Built from 48-Hour Injury Alerts

Scan the league-wide fitness feed at 06:00 every morning; any first-team regular forecasted to miss ≥14 days automatically flags the position in the recruitment matrix. Immediately bump the replacement threshold down by 8 % ball-progression P90 and 0.12 xGContribution P90, then pull every eligible athlete aged 18-27 with ≥600 senior minutes in the last calendar year. Cross-check against wage ceiling (€35 k pw), passport quota (two non-EU slots left) and medical red-flags (previous hamstring ≥20 days). The algorithm spits out 11 names in 90 seconds; export to the 48h folder and push Slack alert to the technical director, physio lead and video analyst before breakfast.

Filter	Cut-off	Names removed
Ball-progression P90	≥1.10	27
Expected goals assisted P90	≥0.18	14
Days injured (prev. 24 mo)	<30	9
Contract expiry	≤18 months	5
Final shortlist	-	11

By 09:00 the analyst has clipped each target’s last four matches into a 6-minute reel tagged by situation: first-phase build, counter-press, set-piece delivery. The manager only wants to see clips where the player receives under pressure and still advances the ball 8+ metres. Meanwhile, the negotiator phones the agent of the top name to sound out salary split and release clause; if the gap is >15 % of budget, drop to the next on the list. When the injury window stretches past three weeks, the committee green-lights a medical and schedules a plane for the same evening; recent example: the Rangers swoop for a €330 M-rated winger after a 21-day ACL setback, detailed here: https://sports24.club/articles/proposed-blockbuster-rangers-trade-sees-texas-land-330-million-two-t-and-more.html. The whole cycle, from alert to signed paperwork, averages 41 hours.

Filtering 2,300 Reports: SQL Queries that Isolate the Final 12 Names

Run a single PostgreSQL CTE that flags 17-year-old midfielders with ≥1 800 successful passes in domestic U19 leagues, non-penalty xG+xA per 90 above 0.48, sprint repeatability index ≥7.1, and hip-rotation flexibility within top-15 % of positional peers; the same statement joins three winter-tracking tables to discard anyone who missed >12 days to soft-tissue complaints or recorded a yoy VO₂-max drop >4 %. Sorting by a composite score that weights 45 % on-ball efficiency, 30 % defensive counter-pressing actions, 15 % aerial duel success, and 10 % psychological resilience rating trims the pile from 2 300 to 34; a second pass cross-checks agent asking price against salary cap headroom and deletes 22 more, leaving 12 names tagged green-light in the dashboard.

The last filter is a window function: rank() over (partition by nationality order by adjusted transfer fee asc) keeps only the cheapest candidate from each passport bloc, preventing squad quotas from clogging up non-EU slots. The query returns player_id, short_name, expected resale value, and a JSONB column with 38 KPI arrows; the recruitment head exports the result set, locks the row-level access until the January window closes, and forwards the list to the sporting director with a one-line memo: Sign top three, monitor rest in satellite club.

Live Coding a Wyscout Clip into a 35-Second Highlight for the Head Coach

Set timeline resolution to 1080p 25 fps, drop the Wyscout MP4 on track 1, scrub to the first frame the striker loses the marker at 17'43", hit "I", jump to 18'18" where he finishes with left foot, hit "O", drag the 35-second slice to track 2, apply 110 % speed to squeeze it to 30 s, add 4-frame fade-in/out, export H.264 3 Mb/s, upload to the coach's Telegram bot.

• Keep the freeze-frame at the exact touch: arrow-key to frame 423, right-click "Add Marker", label it "1v1 trigger".
• Overlay heat-map: Window > Generator > "Gradient", set opacity 35 %, duration 12 s, align with the dribble start.
• Zoom 150 % on the ball carrier: click "Crop", hold Shift, drag corner until the winger fills 70 % of width.
• Duplicate audio, high-pass at 200 Hz on the copy, duck original to -18 LUFS so crowd noise stays but commentaries vanish.
• Colour-grade: lift shadows +5, gamma -10, push reds in mids to make the kit pop against grass.

Coach wants three versions: full 30 s, 12 s vertical for phone, and 6 s GIF for WhatsApp status; queue Media Encoder with three presets named "First-Team", "Mobile", "GIF", hit "Start" once, walk away.

If the clip lacks broadcaster angle, stitch Wyscout wide with the club's drone footage: align both at the pass release frame using audio spike at 6.7 s, blend 50 % for two seconds, then cut to drone for the finish.

Save project template "35s_winger" so next time only the source file and timecodes change; last season this trimmed 11 minutes per player to 90 seconds.

Negotiation Spreadsheet: How €2.3 m Becomes €1.6 m with Three Add-on Clauses

Drop the headline fee to €1.6 m cash and load the rest on conditional rows: €250 k after 50 senior appearances, €300 k if he starts 30 league matches in a single season, €150 k upon Champions-League qualification. Each line is time-boxed (36, 24, 18 months) and capped at one payout only; the seller keeps skin in the game while the buyer caps downside at 70 % of the original ask.

• Appearance trigger: 1 min on the pitch counts; injury-time sub included; U-23 fixtures excluded.
• Starts metric: only XI at kick-off; red-card suspension games deducted; season ends on final whistle of last domestic match.
• Qualification clause: domestic top-four finish or playoff victory; if postponed by appeal, payment due within 14 days of UEFA confirmation.

Book the €700 k contingent liability off-budget; it hits cash only if all three conditions mature, reducing the recognised transfer spend in Year-1 accounts to €1.6 m and lifting operating profit by 0.4 % under league FFP calculation.

FAQ:

How do scouts decide which numbers matter most when the spreadsheet has 300 columns?

They start by throwing out anything the manager will never ask about. If the coach refuses to press high, sprint counts in the final third get deleted. After that, each remaining stat is run through a coach filter: does it predict minutes next season? If the correlation is below 0.35, it’s binned. Normally 12-15 metrics survive. Those are turned into percentile ranks so a 1.8-metre centre-back from Norway can be lined up against a 1.9-metre centre-back from Brazil without currency trouble.

Do clubs still send a live scout after the data lights up, or is video enough?

Video wins the first knockout round; feet win the second. Analysts tag every clip with context—tired opponent, rainy pitch, red-card scoreline—then the scout books a ticket only if the player keeps showing the same behaviour in three different situations. One Premier-League side worked it out: 80 % of eventual signings passed the video test, but only 40 % of those survived the live night game in February with a hostile crowd. The travel bill is tiny compared to a misfire worth 7 million in wages.

What does a red flag look like in the personality report?

A single argument with a referee is ignored; five yellows for dissent followed by a social-media rant is not. Scouts ask the youth coach, the physio, the kit man: Would you drive 300 km with this guy? If two say no, the code turns amber. If his last club froze him out after a row over fines, the code turns red. One Dutch club keeps a shortlist of 200 targets; last summer 14 names were scratched after WhatsApp voice notes leaked showing the player mocking teammates.

How much does the final decision weigh between analyst, scout, and manager?

Most offices print the same pie-chart: analyst 30 %, area scout 30 %, first-team coach 40 %. The numbers shift when the fee crosses 10 million; then the sporting director gets a 15 % slice and the coach drops to 25 %. If the manager still says no, the deal dies even if the algorithm screams yes. The only thing that overrules everyone is the owner’s banker, and that conversation happens in a corridor, not in the war room.

Can a rejected player re-enter the same club’s radar a year later?

Yes, but only if the thing that sank him changes. A Championship club turned down a winger because he tracked back like a retired tourist. Twelve months later the same scout watched him cover 11.4 km per match for a relegated side fighting a man down. The code was reset, the fee rose from 800 k to 3.2 m, and the player now starts most Saturdays. The folder never closes; it just waits for new numbers or a new haircut—sometimes both.

How do scouts decide which numbers actually matter when every player now has thousands of data points?

They start by throwing most of them away. Before a single decimal reaches the first-team recruiter, the club’s head of analytics runs what staff call a noise sweep: every metric is checked against three seasons of video clips to see if it repeats under pressure, against weak opponents, or only on perfect pitches. Anything that wobbles more than 12 % game-to-game is deleted. What survives is then weighted by age and league strength—so a 19-year-old pressing 11 times per 90 in the Championship keeps the same value as a 25-year-old pressing 13 times in the Bundesliga. Once the list is down to roughly 35 KPIs, the scouting group votes with red, amber, green stickers on the wall: red if the stat has never translated to their own squad, green if it matches a non-negotiable club principle (for example, full-backs who hit the blind-side channel within two touches). Only green stickers advance; everything else is archived. The result is a living shortlist of about 180 players worldwide that the office genuinely trusts, and from that point they ignore the spreadsheet until the live eye-check confirms what the trimmed-down numbers hinted at.