Football intelligence & applied AI. Designing systems that turn information into decisions. Former @kaagent • @nkdomzale. MSc Sport Management. American.

Newport Beach, California
Jonathan Stoop retweeted
Please DM me if you know leaders, coaches who come to mind when you read this… Id like to see their work/meet some them
Football clubs can learn a lot about recruitment from tech companies. Player and staff recruitment should be people first, role second. As a club/organization, you want to recruit the best people (A players). But A players are A players because they are unique, multi-disciplinary, and hard to define. They are not cogs in a machine. They do not fit into a predetermined cookie-cutter shape. As a consequence, over-specifying a role filters out A players. Over-specifying a role adds (perceived) risk to hiring people who do not exactly fit the job specs. The odds of an A player emerging from the application process are low because the application process prioritizes conformity. Teams/organizations will have gaps that need to be filled, but roles (whether player or staff) should be a rough outline, not a strict shape. Criteria should shift away from specific skillsets to scalable, context-sensitive traits - low ego, adaptable, self-motivated. Criteria and specifications do not solve problems. People solve problems, often in ways that cannot be precisely predicted because they do so in a uniquely authentic manner. Identify the best people/players first, build roles/processes/systems second. This principle applies to player and staff recruitment. You get better returns organizing top people/players into roles and systems that suit them rather than settling for B/C players who fit into a rigid, predefined role or system.
1
4
1,263
The same thing happens in other industries with cookie-cutter job specs that eliminate anyone unique from the pool. The proof is the work, not the fit to the spec. Too tight a job spec is how you get button pushers instead of A-players, because A-players are not easily defined.
Many clubs fall into the trap of being too specific about the profiles they sign. It starts with being too prescriptive about a gap in the squad. The club decides what it needs, then defines as much criteria as it can around what the ideal profile looks like: age, position, footedness, physical profile, specific actions, experience, budget. Those criteria become filters that extend out into the market, through data and video, eliminating players who don't fit until the best fit is found. The player who remains is not necessarily the best player to solve the problem. He's the player who survived every filter while being available, affordable and interested. Data accelerates the trap. It makes the process feel rigorous and the criteria precise, but every criterion you add makes the search more dependent on the assumptions you started with: that you identified the right problem, that you correctly figured out the solution, and that neither will change over time. The mistake is thinking the solution is a specific set of player characteristics you can plug and play into a complex system. Rarely is there one specific solution to a problem like this, and the more specific your solution, the lower your margin for error. It's a tricky trap because it's usually the product of logical reasoning and objective data. We think we know what the solution looks like ahead of time, and data disguises an educated guess as objective truth. Players are capable of solving problems in unexpected ways, but if you narrow the range of profiles you consider based on specific criteria, those players are often removed from the process before they are ever seriously evaluated. You end up selecting for conformity, not quality. The better process is identifying the best players, the ones who solve many problems in many ways, then building specific roles and systems around their qualities.
1
10
3,433
Are strikers the most team-dependent position in football? To score goals effectively as a team, the striker needs the rest of the team to play towards his strengths. If that isn't happening consistently, the team is probably dysfunctional at some level and likely not scoring many goals. Responding to that dysfunction tends to require 1) adjustments to the team's tactical approach, and/or 2) playing a different striker. A problem somewhere else in the team can end with the striker being replaced. It's also what makes scouting and recruiting strikers difficult. It's hard to remove them from their team context and evaluate them in isolation.
1
4
703
Jonathan Stoop retweeted
How cool is this?
167
3,830
49,938
3,224,400
A team is not the sum of its parts. Models that treat it as such are missing the layer of how teams perform above or below the additive value of its players: synergy. From my notes:
3
16
1,312
I was talking to someone recently who builds financial models for trading. He says each model follows one basic principle: past performance is the best predictor of future performance. That's correct when the data going into the model is available and validated. I challenged it by asking what the dataset was missing. In finance, this is the obvious model because you have decades of clean, audited data. You don't have that in football. Event data is a proxy for performance, and we apply value judgments to it. You can't let past performance predict future performance when you can't measure the data with a high degree of trust. Every model inherits the limits of its dataset. That doesn't mean models aren't effective in football. It makes it vitally important to know what your data can't see.
1
2
34
5,922
Jonathan Stoop retweeted
Football clubs can learn a lot about recruitment from tech companies. Player and staff recruitment should be people first, role second. As a club/organization, you want to recruit the best people (A players). But A players are A players because they are unique, multi-disciplinary, and hard to define. They are not cogs in a machine. They do not fit into a predetermined cookie-cutter shape. As a consequence, over-specifying a role filters out A players. Over-specifying a role adds (perceived) risk to hiring people who do not exactly fit the job specs. The odds of an A player emerging from the application process are low because the application process prioritizes conformity. Teams/organizations will have gaps that need to be filled, but roles (whether player or staff) should be a rough outline, not a strict shape. Criteria should shift away from specific skillsets to scalable, context-sensitive traits - low ego, adaptable, self-motivated. Criteria and specifications do not solve problems. People solve problems, often in ways that cannot be precisely predicted because they do so in a uniquely authentic manner. Identify the best people/players first, build roles/processes/systems second. This principle applies to player and staff recruitment. You get better returns organizing top people/players into roles and systems that suit them rather than settling for B/C players who fit into a rigid, predefined role or system.
4
11
102
31,937
Many clubs fall into the trap of being too specific about the profiles they sign. It starts with being too prescriptive about a gap in the squad. The club decides what it needs, then defines as much criteria as it can around what the ideal profile looks like: age, position, footedness, physical profile, specific actions, experience, budget. Those criteria become filters that extend out into the market, through data and video, eliminating players who don't fit until the best fit is found. The player who remains is not necessarily the best player to solve the problem. He's the player who survived every filter while being available, affordable and interested. Data accelerates the trap. It makes the process feel rigorous and the criteria precise, but every criterion you add makes the search more dependent on the assumptions you started with: that you identified the right problem, that you correctly figured out the solution, and that neither will change over time. The mistake is thinking the solution is a specific set of player characteristics you can plug and play into a complex system. Rarely is there one specific solution to a problem like this, and the more specific your solution, the lower your margin for error. It's a tricky trap because it's usually the product of logical reasoning and objective data. We think we know what the solution looks like ahead of time, and data disguises an educated guess as objective truth. Players are capable of solving problems in unexpected ways, but if you narrow the range of profiles you consider based on specific criteria, those players are often removed from the process before they are ever seriously evaluated. You end up selecting for conformity, not quality. The better process is identifying the best players, the ones who solve many problems in many ways, then building specific roles and systems around their qualities.
3
5
51
8,116
From my notes on what a scouting report is actually for:
1
4
40
3,052
I saved this quote from a Mourinho press conference last week. Defensive tactics are a calibration of the players' capacities, spacing and distances against an opponent. Stamina is a key factor in that formula. You have to design team intentions that account for energy declining across the 90 minutes. It is so difficult to maintain the same level of aggression and coordination in a low block or a high press across 90 minutes. The best way is to do a bit of both. It's the coach's responsibility to know enough about their players' capacities to to find the right calibration and to train players on how to adjust this calibration during the game as energy declines and momentum shifts.
12
1,324
Jonathan Stoop retweeted
I've written about the scouting principle Luis Campos is referring to in the quote below: "Scouts should form opinions in isolation to prevent groupthink"
🚨 𝗡𝗘𝗪: Luis Campos has revealed one of his unique scouting methods. 🗣️ "Sometimes I send two scouts to the same match without them knowing they're both watching it. "I love comparing their reports. I don't let them share information because I don't want them to influence each other. I want their opinions to be as honest and pure as possible. "I call them my super scouts. I also scout the scouts by choosing them. They are special people."
1
8
96
12,950
Another fascinating idea from the documentary: you can't evaluate a player from across the ocean if you don't know what players on both sides of the ocean look like.
Some time ago, I saw a YouTube documentary on the Brooklyn Nets scouting department and saved this image of the mantra on the wall during meetings. "DISAGREE AND COMMIT" I'm fascinated by how organizations like this make decisions: who has the final say and whose input carries the most weight. The mantra sparks this question: does debate lead to the best idea or the best argument? Those aren't the same thing. It's easy to mistake clear reasoning for correct reasoning. Imagine a group of individuals with specialized knowledge who are paid for their expert opinion, all with different versions of the truth and strong arguments. What happens when the highest-level decision-maker has a specific opinion and nobody below wants to disagree with the person responsible for their job security? The best thinking lives in the tension between opposing ideas, and the decision hierarchy should respect that and optimize for it. That's why scouts should form opinions in isolation, and why the most senior person in the room shouldn't speak first. The meaning of disagree and commit is that disagreement is healthy, not a red flag. It's important to dig into disagreement to understand why, rather than agreeing to disagree. The commit part is integrating multiple perspectives and selecting one angle from many. That's far easier when one person is responsible for the final decision rather than needing consensus from the group. There's a balance between working with people who think similarly to you and people who challenge your thinking. What is a healthy agreeableness ratio? 80/20? How deep should differences in perspective go?
11
2,189
Some time ago, I saw a YouTube documentary on the Brooklyn Nets scouting department and saved this image of the mantra on the wall during meetings. "DISAGREE AND COMMIT" I'm fascinated by how organizations like this make decisions: who has the final say and whose input carries the most weight. The mantra sparks this question: does debate lead to the best idea or the best argument? Those aren't the same thing. It's easy to mistake clear reasoning for correct reasoning. Imagine a group of individuals with specialized knowledge who are paid for their expert opinion, all with different versions of the truth and strong arguments. What happens when the highest-level decision-maker has a specific opinion and nobody below wants to disagree with the person responsible for their job security? The best thinking lives in the tension between opposing ideas, and the decision hierarchy should respect that and optimize for it. That's why scouts should form opinions in isolation, and why the most senior person in the room shouldn't speak first. The meaning of disagree and commit is that disagreement is healthy, not a red flag. It's important to dig into disagreement to understand why, rather than agreeing to disagree. The commit part is integrating multiple perspectives and selecting one angle from many. That's far easier when one person is responsible for the final decision rather than needing consensus from the group. There's a balance between working with people who think similarly to you and people who challenge your thinking. What is a healthy agreeableness ratio? 80/20? How deep should differences in perspective go?
3
2
34
4,645
Jonathan Stoop retweeted
The Zadok Yohanna deal from AIK to Brighton is a clean case study for two recruitment principles: 1. Clubs that make confident decisions from small sample sizes have an edge ~50% of Yohanna's total professional minutes (580') came in the last two months. Most clubs would prefer to wait for the sample to grow before making a move, but by the time it does, the price has already moved to reflect the larger sample size. Brighton (and other clubs, possibly because they trust Brighton's strategy) act earlier because they bet on their ability to analyze sparse information and make projections from a small sample size. The uniqueness of each player makes this principle a case-by-case application, but after watching Yohanna on Wyscout, this is a player where a 30 game sample size would only confirm - not deny - what the first 10 already show. Paying a fee that looks inflated relative to the player's experience and current level is rational when you're confident in a sample (or your analysis of that sample) that everyone else considers too thin. 2. There is lucrative business in being the middleman AIK signed Yohanna out of Nigeria, gave him a platform in European football, and sold the proof of concept at a significant multiple. Scandinavia has become the connective tissue (middleman) between the West African market and bigger European leagues - sourcing the talent and letting buyers pay for the network connection rather than build those relationships themselves. This type of arrangement carries a better risk profile for the buyer and an enormous return for the seller, which is why it keeps repeating. The middleman's real product is their leveraged position within the supply chain that connects two markets who otherwise don't interact with each other. Source where the information is this, decide before the market adjusts, and price on your own conviction.
6
10
118
23,874
Early-stage investing has a lot in common with player recruitment: - Making projections under uncertainty - Working with small sample sizes - Paying a premium for an unfinished product Both require skill in selecting talent.
2
25
3,228
I saved these clips many years ago of James Garner at U18 level. I remember being so impressed by how he strikes the ball so purely. Certainly thought he'd break into United's first team. Now on a trajectory to becoming a regular for England.
8
1,276
"We want to be unpredictable" This is optimal from a game theory perspective but there is a balance to strike. Identity is a constraint in both a positive and negative sense. A team that can shape-shift and show different identities in different moments is difficult to predict. Unpredictability is the highest form of competitive advantage from an attacking standpoint. But a well organized team has to attune to the same references. An identity is how you get players to narrow their focus on what matters most to the team. The players need clear team intentions so they can collectively recognize what the team is trying to do in real time. Team intentions can be executed with infinite tactics, and that's where the unpredictability should come from. It's the coach's responsibility to decide those team intentions and apply them top-down. Team intentions are fixed but the tactics are variable. The range of tactical solutions comes from the players. The more range, the more adaptable the team. Ideally, you want to adapt to the opponent, not because of them. Unpredictability only works as an advantage when your own players can still predict each other.
26
1,692
Jonathan Stoop retweeted
Current events at Man United remind me of a few Jason Wilcox podcast quotes from Nov 2025 that now seems completely incoherent: "Ruben's got a clear idea. It's a lot more flexible than people give it credit for" "We're really clear. Myself, Ruben, Omar, the ownership. Really clear on the direction of travel" "We're very clear on the profiles... the brief will come from myself and Ruben... a lot of debate and discussion" My read: The friction has always been there. A recruitment strategy built on the (supposed) alignment between two people who see the game differently is a fundamental misalignment. JW is skeptical of system-specific team-building. RA is uncompromising. That's the bottleneck. Everything downstream is corrupted. My assumption: JW has always pushed for a more flexible approach. RA perhaps verbally agreed (hence the JW quotes) but never followed through. Finally RA gives in and his pride is damaged (hence the MANAGER not coach quote). It is untenable now.
8
24
190
53,422