The rundown
(in no particular order)
- 1.
The major news in Major League Soccer Soccer Land last week was that a new proposal for roster rules had been considered by MLS owners. The even bigger news this week is we now know who will succeed Don Garber as MLS commissioner: Larry Berg. The LAFC part-owner answered questions after being formally announced and ruled out any possibility of pro/rel in MLS for the foreseeable future (good, imo).
- 2.
A petition to change NYCFC's name to AC NY has registered a whopping 19 signatures as of time of writing, which is surprising, because I didn't know there were that many NYCFC supporters! (cue laugh track) All jokes aside the petition does have a point about the weirdness of American soccer team naming, but I'm not sure "AC NY" really does anything to fix that.
- 3.
In case you haven't been witnessing the collective madness unfolding on Bluesky as real people lose their lives to imaginary games (this is me I am real people), two new 82-0 style games dropped this week and have taken the sicko community by storm. The first is Centurions, by Futi, which has a Premier League and MLS mode and follows the traditional 82-0 format with randomized picks for every position. What's extra fun, though, is that they've added play styles and chemistry, which is activating the part of my brain that's been dormant since I last played FUT in Fifa 21. I finally managed a centurion season in MLS with not one, not two, but four 99-rated players, which was pretty cool. The second game to release was ASA's roadto75, which is similar to Centurions but instead focuses on maximizing cumulative g+ while having limits on what types of players can be drafted — in MLS mode, you can only have 3 DPs. There's also an NWSL mode, which is awesome and addicting. Overall, these games are awesome, just beware that you will be lost in a haze for the rest of your workday.
- 4.
After putting out a call for pitches a few weeks ago, SCOUTED will be running a Football Writing Festival at the end of this month, where the outlet will be publishing an article every weekday, mainly from new writers. SCOUTED will also be dropping paywalls across some of the site's most popular content and running a subscription drive. SCOUTED has some of the best analytically-informed content out there, and this is a great opportunity to get familiar with the type of pieces they produce.
- 5.
The Transfer Flow is back with Premier League previews, a few of which will be written by yours truly. Sometimes it can be difficult to read up on the data, learn all the storylines, and feel prepared to sound smart to your friends all at the same time, but these previews give you that all in one place. Obviously, I'm biased, but the £12 price tag for all 20 previews is a steal.
- 6.
Everything is becoming Big Tech and soccer clubs really want to be Google now. Mark Thompson over at Get Goalside wrote about the techification of soccer and I 100% agree with what he has to say. An example he used, that I noticed myself, was the "Data On The Lake" hackathon hosted jointly by Como 1907 and Google DeepMind. Como have been one of the more analytics-driven clubs in recent years, but the partnership with Google was surprising to me. Photos from the hackathon showed a slick event held in ornate rooms, and in every post about the conference, the word "AI" showed up more often than "analytics." I don't think this is a bad thing; I think it's great that this type of innovation is happening. But it is interesting to see the conversation around data in football shift from hobbyists, bloggers, and nerds with spreadsheets to a company that ranks #6 on Fortune 500. Don't get me wrong, the analytics community has long been driven by former big tech people that made career changes to work in football, but the worlds never really overlapped. Principles, concepts, and experience have been shared between disciplines, but for a long time, football analytics was seemingly untouched by the outside world1. Now, that's changing. In a way, it's the beginning of a gentrification. Will it work? Probably, and there are certainly things both industries can learn from each other. But all I could think about when looking at pictures from the Como hackathon was: "that's beautiful" and "I don't want to be there." As much as I'm sure I'd learn a lot from hearing talks on transformers and innovative dashboard methodologies, I'd much rather be at something like the American Soccer Insights Summit, surrounded by friends, colleagues, and soccer nerds. But change is happening. As Mark put it: "smart people who've been around different areas of this ecosystem should be in demand." I think that's true, but I also wouldn't be surprised if we see a wave of techification in this industry before a pendulum correction. In the end, good will come of it all, but for right now, it'll be interesting to watch how the out-of-towners adapt to soccer — and how soccer changes in the process.
The Post-Match Report
A closer look: AI, coding, and post-match reports
I started doing soccer analytics work around 2.5 years ago, midway through my freshman year of college. My first exposure to the industry was the post-match reports being posted to Twitter by and . These introduced me to many of the foundational graphics and metrics I still reference today; I learned about passing networks, PPDA, progressive passes, and so much more. After a bit of digging, I was inspired to make my own version of these graphics, so I created . My first attempt wasn't great, but it was original, and I kept working on it. Over the course of a year, my graphics went from humble beginnings to where they are now, and I learned so much in the process. Year 2 ended up being about ASA VizHub, an app I had dreamed up that could be used by anyone to create high-quality data visualizations based on ASA data. That released in September of last year and I made improvements around Christmas. If I were to put a label on this year, it would probably be writing. I've spent far less time on VS Code in the last few months and far more time on Google Docs, and the number of different places I've seen my words appear has increased. Each of these years taught me a host of different things, and each experience was built on the previous one — I would not be able to write now if it weren't for the work I'd done in the past, etc.
Every time I scroll on LinkedIn, I'll see someone just starting out. Maybe they're building their first public project or maybe they're sharing a link to a substack article. When I see these posts, I think back to what it was like when I began working on soccer data. How it felt to throw stuff out into the world, spend hours trying to make something look just right, and constantly refreshing TeamworkOnline to see if there were any new internship postings. It's a weird and exciting state to be in.
Plenty of people will tell you what you need to do to get into soccer analytics. Maybe they'll sell you a course. Maybe they'll offer a report template. Maybe they'll promote a networking event. These things are all useful, in one way or another. My belief is that at the end of the day, it all comes down to putting yourself out there and publishing work you're proud of so that others can see what you can do. That's at least what I've done, and so far, I think it's worked somewhat well2. Uniqueness is a core aspect of that though, and unfortunately, it's sometimes forgotten. In February of 2025, I wrote a piece on Substack called "In defense of beauty." I still stand by pretty much everything that I said in that piece, but it was born out of a frustration with sameness.
There's a certain type of post-match report graphic that has been widely shared across LinkedIn. There must be a GitHub repo somewhere, I'm sure, because countless different people have made their own copies and written their own posts about building an end-to-end data reporting pipeline. I think this is awesome; the more people getting their hands dirty with soccer data and post-match reports, the better. But something about that graphic has always irked me. It's generic, hardly changes from post to post, and my eyes glaze over as soon as I see it. I'll wonder "why did they choose this plot or this color" and know that there isn't an answer. Because to use that GitHub code, to create a project simply for the sake of having "done a project," is to surrender your choice. By choosing the easy route, you blend in.
I see this type of graphic less and less these days. Instead, I see Claude.
AI use for coding is an interesting question which deserves far more thought than I can fit into this paragraph, but I'm probably more on the anti-AI side of the spectrum than most. I think it's a useful tool for coding, but I've always objected to it's use for creative pursuits — writing, image generation, and yeah, UI design. When I built VizHub, I specifically wanted it to not look AI-generated. Since the advent of coding agents, hallmarks of artificially constructed UI designs have come and gone. Services have popped up to make your page look less AI-generated, only for those same design styles to become associated with vibe coding. It is an endless loop, with really only one way out: doing it yourself. The feeling I get when I look at an AI-generated dashboard, report, or website is the same feeling I get when I look at that generic post-match report. It feels empty. Bland. And it tells me that whoever created this did not put much thought into how things looked. Maybe they asked Claude to execute a certain design style or iteratively prompted the agent to make something look a certain way. This is thought, yes. But if, at the end of the day, your dashboard has the same tells as a dashboard made by a completely different person, then the point is moot. You may have created something from scratch, but it is nothing more than an aggregation. It is, by virtue of its creation, unoriginal.
Most people will care far less than I do about this, and that's okay. But mostly, when I think about the question of breaking into the industry, I think about originality. Most teams in MLS have at least one data analyst. Every team has a league-provided set of data, and most have external contracts as well. The resources are largely similar. The question is what you do with it. The point of data analytics is and always has been competitive advantage.
Originality is what gives you that advantage. Claude Opus or ChatGPT 5 might give you the edge in creating whatever it is that you're building, but as soon as you start outsourcing thought and creative decision making to the model, you are ceding control to a publicly available resource. Who's to say some other analyst at some other team couldn't use that model to build the exact same thing you did? There is no "we thought of it and they didn't" advantage when a shared model is doing the thinking.
I have been on all sides of this. I have lost myself to the models and gone completely cold turkey. I have tried local LLMs3 and signed up for IDEs simply to use free trial credits. What I come back to, time and time again, is that coding agents are dangerous, especially when you're trying to create something meaningful and original.
I'm sure that if I wanted to, I could paste an image of that generic post-match report into Claude and ask it to recreate the graphic, create a data ingestion pipeline, set up automatic reporting, and configure automated uploads to a Twitter bot. If I had done that 2.5 years ago, I wouldn't be here now, writing this newsletter.
So, I don't know. Maybe this is a plea to stop using AI creatively because I'm tired of seeing Claude-generated UI4. Maybe it's a warning to newer people in the industry. Maybe it's a reflection on the techification of soccer — if every team is chasing "perfect process" and integrating AI into decision-making, is the industry destined to converge on sameness? Maybe it's just me ranting because I want to.
As always, discussion is encouraged, so I'm interested to see what others think.