Jonathan Stoop

4.3K posts

Jonathan Stoop banner
Jonathan Stoop

Jonathan Stoop

@jonathanstoop

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

Newport Beach, California Katılım Mart 2010
659 Takip Edilen5.8K Takipçiler
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
Karim Coulibaly will be a subject of interest in the summer window. I flagged him earlier in the season as he broke into the first team - my notes at the time: His standout quality is the huge action radius his large frame and long limbs provide as a central defender. – Extra reach to block and tackle – Long strides to cover space – Out-to-in diagonal passes on his left foot The developmental picture is more complicated. – His strength and coordination hasn't caught up with his frame - he's still learning to move his body in reactive situations which costs him in tight spaces or little time. – Signs of cognitive overload from a player making a significant jump in playing level - he's coped but hasn't figured out how to assert himself yet. Sort of reminds me of what a teenage Gabriel might have looked like. His market value will be priced based on what he could be rather than what he currently represents. Another season at Bremen would likely be best for his development but many top European clubs will see him as a market opportunity.
Jonathan Stoop tweet media
English
0
3
54
6.1K
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
The Zettelkasten design of my personal knowledge base allows me to scale the application of my ideas with AI because agents run on the same infrastructure. Most agents are novices because they run on a flat context file. An expert agent routes through a tiered knowledge base - index, sub-index, atomic ideas, wikilinks - that is far more efficient and nuanced. Expert: problem → directory → sub-directory → answer. Novice: problem → everything. The architecture only matters if what it routes through is worth routing through. The wikilinked connections tell an agent what matters and what threads should be followed for a given task. The necessary information is shaped by the connections I've personally made with my specific knowledge instead of LLM-generated generic knowledge that dilutes context. An agent with my Zettelkasten as its repository reasons from my perspective because the context is extremely potent - it's the processed thinking I've accumulated: decisions about what connects to what, what belongs in the same index, what's worth an atomic note.
Jonathan Stoop tweet media
Jonathan Stoop@jonathanstoop

My personal knowledge base in Obsidian is my attempt to deconstruct my intuition and deliberately improve it. My Zettelkasten mirrors same architecture as expert intuition, but made visible. Chunking through wikilinking - connecting ideas by their shared logic and substance. Indexes (topics) as the hierarchy. Creating and navigating the connections carves the navigational system. Writing is how I internalize my ideas and their connections to other ideas, rather than just recording them. The act of writing is the mechanism. The structure operates without the overhead of having to navigate through noise, similar to how intuition runs subconsciously. I generate ideas and connections consistently without brute force.

English
1
0
6
1.2K
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
My personal knowledge base in Obsidian is my attempt to deconstruct my intuition and deliberately improve it. My Zettelkasten mirrors same architecture as expert intuition, but made visible. Chunking through wikilinking - connecting ideas by their shared logic and substance. Indexes (topics) as the hierarchy. Creating and navigating the connections carves the navigational system. Writing is how I internalize my ideas and their connections to other ideas, rather than just recording them. The act of writing is the mechanism. The structure operates without the overhead of having to navigate through noise, similar to how intuition runs subconsciously. I generate ideas and connections consistently without brute force.
Jonathan Stoop tweet media
Jonathan Stoop@jonathanstoop

An expert's intuition is not just instinct but a structured hierarchy of information running below awareness that knows what to focus on and what to ignore. Novices in any domain lack this filter, which is why experts generally perceive less than novices. An elite midfielder checks his shoulder once for the same information that average midfielders struggle to see with multiple scans. A chess grandmaster isn't counting piece values mid-game. The pieces are felt intuitively rather than calculated. Experts have internalized which cues matter and narrow attention before anything irrelevant reaches conscious processing. Knowledge becomes a filtering system that runs without conscious overhead - we call this intuition. The mechanism is chunking. Clusters of related information bound together by principles and patterns. The formation of these networks and repeated navigation through them carve neural pathways that become unconscious pattern recognition. A decision tree, built below awareness. Intuition improves as you evolve your navigational systems. You see more with less. What once was information overload becomes a clear set of specific steps. Conscious bandwidth - finite by nature - is re-allocated toward what matters.

English
0
0
13
3.1K
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
@austriangooner1 I’ve done both my MCLs. The worse one happened the same way as Ben White - I was standing and the other player was running into the tackle. Difficult to brace for that impact when the opponent has far more momentum.
English
0
0
0
428
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
Metacognition is a powerful skill but first requires the externalization of thinking. Self-awareness is fundamental but you can't effectively think about your thinking without making your thinking visible. Producing and building things is how your thinking becomes legible, so you can observe the patterns in your own interpretive process with actual feedback from reality. Without feedback, the brain runs it loop unchecked. It's self-protective by design. The echo chamber is the default state when there's no external record to push back. The aim isn't endless self-analysis, rather it's developing the capacity to see your own process for generating your interpretation of reality.
All day Astronomy@forallcurious

🚨: Neuroscience considers metacognition the highest form of intelligence..... "the ability to think about your own thinking."

English
0
0
8
1.2K
Jonathan Stoop retweetledi
Training Ground Guru
Training Ground Guru@ground_guru·
Bukayo Saka got 4 A*s and 3 As in his GCSEs in 2018. “He always put education first and epitomised what we think a lifelong learner is. Having a thirst for personal development is crucial for us. “He has made himself a better player because he’s a better person, a better student. And he will continue to go on that journey, he’s not someone to think ‘this is me done.’” @mertesacker “It would have been nice to see him do a degree and a very hard degree, he was that level.” @DeasunO archive.trainingground.guru/articles/merte…
Football on TNT Sports@footballontnt

“I don’t think he’s going to need those GCSEs to be honest with you” Max Dowman has a lot to balance with playing for Arsenal and doing his GCSEs 😅 Watch the full episode of The Breakdown, now live on YouTube!

English
1
17
114
65.9K
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
An expert's intuition is not just instinct but a structured hierarchy of information running below awareness that knows what to focus on and what to ignore. Novices in any domain lack this filter, which is why experts generally perceive less than novices. An elite midfielder checks his shoulder once for the same information that average midfielders struggle to see with multiple scans. A chess grandmaster isn't counting piece values mid-game. The pieces are felt intuitively rather than calculated. Experts have internalized which cues matter and narrow attention before anything irrelevant reaches conscious processing. Knowledge becomes a filtering system that runs without conscious overhead - we call this intuition. The mechanism is chunking. Clusters of related information bound together by principles and patterns. The formation of these networks and repeated navigation through them carve neural pathways that become unconscious pattern recognition. A decision tree, built below awareness. Intuition improves as you evolve your navigational systems. You see more with less. What once was information overload becomes a clear set of specific steps. Conscious bandwidth - finite by nature - is re-allocated toward what matters.
Jonathan Stoop tweet media
English
0
6
36
5K
Jonathan Stoop retweetledi
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
Here is Brian Brobbey posting up against the giant that is Jonathan Tah. He's such an effective backboard. Few CFs cause as much discomfort for opposing CBs as Brian Brobbey.
Jonathan Stoop@jonathanstoop

I love these passes. Streamlined access to the box. It's uncomfortable for defenders to deal with and plenty of attacking ideas and patterns can emerge. This picture isn't executed enough. Center-forwards that can post up and be a backboard for teammates are highly effective.

English
6
8
82
32.9K
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
Commentators should have better knowledge of which players have been teammates at previous clubs. “Lisandro Martinez hasn’t dealt with many strikers like Brian Brobbey” They spent years at Ajax together. He’s defended against him 100s of times.
English
0
1
13
1.2K
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
In football, we generally assume being "data-driven" is being "objective" in analysis and decision-making. The irony is that assumption itself is an aesthetic preference - what others would call being subjective. I think many people are allured by the aesthetic of objectivity, but as with most allures, the allure itself is a preference. The decision to weigh what's measurable over what's observable is really a philosophical commitment to a specific position (some would call this an opinion). The mistake is thinking objective = right and subjective = wrong. Downstream of this assumption is something I call "objective bias": assuming something is correct because you've deemed it objective. In the near future, data will be commoditized and therefore "objectivity" will be commoditized. When every clubs runs the same datasets, the outputs naturally converge. The competitive advantage disappears and "objectivity" becomes table stakes. What's left is taste - how you select and filter what you pay attention to, based on what you actually care about. Taste is typically dismissed as being subjective, but that's exactly why it can't be commoditized. It is built from your specific experience, your specific curiosity, your specific judgments accumulated over time. No two people have the same taste, that's the edge.
English
0
2
17
2.1K
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
Sitting above the data layer means you've stopped trying to hold every bit of information in your head and started making better decisions with better information, not more information. Most decision-makers are either inside the data themselves, or rely on someone who is. Attempting to capture and process as much information as possible through sheer cognitive effort leads to diminishing returns because the volume limits the quality of thinking and deciding. AI removes the bottleneck by handling the synthesis of information at scale, enabling decision-makers to move up a layer. AI changes the human job to be done - it shifts from manually producing patterns from the data to selecting the patterns that AI automatically produces from the data: which ones are most important and - equally - which ones aren't. That selection is judgment. AI can surface patterns at scale but it will struggle to rank them in context. Knowing which patterns matter most is the human ability that cannot be automated. Taste is the edge.
Jonathan Stoop@jonathanstoop

Headline: Getafe's points per game rose from 0.89 to 1.37 since working directly with an AI model in January 2025. Let's break down what's actually happening, without assuming the stat proves causation. A large language model works alongside the coaching staff, synthesizing data at scale to surface relationships and patterns across variables that human attention can't hold simultaneously. The density of information available to a coaching staff - especially on matchday - is enormous. The application of AI can synthesize the data in real-time and identify patterns that are predictive, not just descriptive. The final decision is still human. Thinking can be outsourced with AI but understanding cannot - the staff are still the understanders. But the human decision-maker now sits above the data layer with an information edge that improves the raw material used to make decisions. I'm bullish on AI in football for its fundamental ability to surface patterns at scale. Not because it replaces judgment, but because it improves the quality of information judgment works from.

English
0
1
21
6.2K
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
Early adoption is useful now but table stakes once AI fully diffuses and everyone has the same information edge - therefore no longer an edge. The edge is the judgment that exists above the tool, specifically to know what information matters and what doesn't in an era of unlimited info. The mistake will be assuming the information itself is the edge.
English
0
0
0
97
jucojames
jucojames@jucojames·
@jonathanstoop Early adopters may enjoy some predictive edge but there is a reflexivity to the dynamic as adoption increases - basically same as humans trying to anticipate the other anticipator. This often causes what were relatively stable ex-post statistical correlations to fail badly.
English
1
0
2
482
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
Headline: Getafe's points per game rose from 0.89 to 1.37 since working directly with an AI model in January 2025. Let's break down what's actually happening, without assuming the stat proves causation. A large language model works alongside the coaching staff, synthesizing data at scale to surface relationships and patterns across variables that human attention can't hold simultaneously. The density of information available to a coaching staff - especially on matchday - is enormous. The application of AI can synthesize the data in real-time and identify patterns that are predictive, not just descriptive. The final decision is still human. Thinking can be outsourced with AI but understanding cannot - the staff are still the understanders. But the human decision-maker now sits above the data layer with an information edge that improves the raw material used to make decisions. I'm bullish on AI in football for its fundamental ability to surface patterns at scale. Not because it replaces judgment, but because it improves the quality of information judgment works from.
Jonathan Stoop tweet media
English
5
2
26
13.9K
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
@ForestHills1903 I think the right framing is AI as a tool that improves human judgment. It is an extremely powerful tool because it is highly adaptive, but ultimately it is a tool. The skill is knowing how to use it.
English
1
0
1
101
Keyser Söze
Keyser Söze@ForestHills1903·
@jonathanstoop I use AI a lot at work but Im concerned about AI in sport. definitely see big increases in top level quality as AI could help identify those millions of kids good enough but left behind by coaches with little talent ID, but the same time you want humans deciding teams/tactcis.
English
2
0
2
148
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
@jason_g4 I think something like "good understanding of LLMs and AI tools" might eventually appear in job descriptions as a nice-to-have. But the fundamental requirement (if I were hiring) would be the curiosity to explore new tools and technologies.
English
0
0
1
80
Jason Gonsalves
Jason Gonsalves@jason_g4·
@jonathanstoop Do you think this could have an impact on how clubs hire coaches moving forward? I.e. their willingness to utilize an AI model in this fashion?
English
1
0
1
119
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
@JPT5_ Can't emphasize this enough, and I've been on both sides. Most don't realize how many people in the football industry (even at the top level) are lurking around X reading and sharing work, and the work they're sharing is rarely the posts with the most likes and reposts.
English
0
0
2
211
JT
JT@JPT5_·
Adding to this… your work online is always seen by some professionals in the game. It is also sometimes discussed behind closed doors. This can both be positive and negative, it all depends on how you present yourself and your work online.
Jonathan Stoop@jonathanstoop

@umirf1 The key is playing the long game. Consistent quality compounds in ways that can't be anticipated, the cost is the delayed gratification.

English
2
0
8
1.4K
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
This is a skill anyone can develop with enough practice. One of the most important for anyone in the knowledge arena. Two training methods I'd recommend: 1. Capture ideas as they arise. Voice notes, handwritten notes - I voice dictate into a daily note in Obsidian. Try this and you'll quickly realize how difficult it is to express ideas in real time. 2. Write every day. Sit with an important idea and think deeply about it. The words will not appear immediately, the resistance is part of the process. If you try to articulate a complex idea - in speech or writing - you'll notice how unclear it is compared to how the original thought felt in your head. Thoughts are brittle and original ideas are never as complete as they seem. When you write, the gaps in your thinking appear immediately and you are forced to resolve them. That is when the idea actually forms. The resistance you feel when writing is the generative act, that's where the actual thinking happens. Train this articulation muscle consistently and your thinking becomes clearer even outside of writing. Ideas start arriving with more completeness. It's a beautiful feedback loop - you learn thinking through writing and you learn writing through thinking.
yimika|@yimikaaaa

Being able to articulate your thoughts into writing is a crazy gift.

English
1
1
17
1.6K
Jonathan Stoop retweetledi
Jonathan Stoop
Jonathan Stoop@jonathanstoop·
@umirf1 The key is playing the long game. Consistent quality compounds in ways that can't be anticipated, the cost is the delayed gratification.
English
1
3
24
8.1K