MartinOnData

What would your bankroll look like if you had followed BBC pundits over the past 16 years? Find out below 👇 Forecast data by @Soccer_Stats Odds data by @12Xpert

We’ve quietly built something pretty useful 👇 • 106 clubs tracked • 120+ Companies House entities mapped • Auto-refresh every 30 mins with the latest filings No more manual digging. No more email alerts. It’s all just… there. Free for everyone.

@EBL2017 MLS - The way he dribbles past opponents once inverted is astounding. I would love to keep him.

Lewis-Skelly should not be sold. Special footballer. Just needs to A) be patient and B) grab a hold of the shirt and not let it go when he gets a chance. Nwaneri has Foden-esque qualities. Just turned 19. But is further away than MLS. Sell wisely, but with a buyback clause.

reep updates: - the reep register has a new home! reep.football - we're almost at 500k ⚽ entities across 44 providers. please keep sending any mappings or datasets you have - we now have competition and season ids across several providers, coverage will increase slowly - icymi: reep_id is a new stable identifier for all entities across providers. so one id -> any player, team, coach, competition, season, etc if you have any mappings or IDs to share (statsbomb needed, anything football related, even games, are useful), please use the anonymous upload on the site to share them. you can also DM me links

@EBL2017 The most underrated Arsenal player and one of the best right backs in the world.

Now you see why Timber is a bonafide starter.

@tropianhs Let’s goo 🚀🚀

March Monthly Target excedeed $1k month coming soon!

matching player IDs across opta, transfermarkt, fbref, fotmob and 25+ other football data providers sucks. spreadsheets, fuzzy matching, hallucinated IDs. the reep register fixes this: an open source crosswalk of 430k players and 45k teams across all of them. updated weekly. github.com/withqwerty/reep

Just over half of the 92 clubs in top 4 divisions in England have published 2024/25 accounts. Only 4 are profitable: Liverpool, Bournemouth, Plymouth Argyle and Peterborough United. Average pre-tax losses: PL £31.1m, Championship £15.1m, League One £5.0m, League Two £2.6m.

@tropianhs I guess it’s a different game when you are allowed to advertise your product 😅

I have already broken my ambitious target for March And there are still 10 days left until the end of the month Could maybe break the $900 mark which would be huge Let's see!

@Marauderrr_ @GoonerballFC @scottjwillis Bullying defenders.

@GoonerballFC @scottjwillis What does gyokeres do better than Jesus?

Massive appreciation for Havertz today. His sub made a massive difference towards Arsenal's ability to grab a foothold in this match.

@xGPhilosophy Game state matters.

HT: Brighton (0.43) 0-1 (0.01) Arsenal

@tropianhs Love them. It would be interesting to see if this also leads to some spikes in the KDP sales (given that not all traffic would be physics-related😅)?

@MartinOnData My specialty 🤣

10 years ago I was working in London Decided to move back to Italy I was in the best position, London salary, Italian cost of living Right? Huge mistake - wife could not find a job - my career done - stopped saving (one salary and taxes too high) - lost 20k when my company exited If you feel homesick, don't give up There is a reason you moved abroad

@tropianhs I miss the constant KDP royalty estimator page refreshes once you are on an upward trend 😅

Broke the $350 mark in royalties already Already my best month ever and there is one week left Probably will pass $400 royalties this month 🤞

@tropianhs Why nations fail 🔥

Big book order just arrived I don't know where to start from but probably will read 2 or 3 in parallel Btw already checked the Space Math one Dover never fails to deliver

Give me 5 minutes to reframe what you think about young wingers. The TLDR: physical capacity to run declines consistently pretty much from the early 20s to the end of a players career, while 'end product' related metrics at best have a slightly upwards trajectory, and at worst are wildly erratic. If you want a truly 'game breaking' winger, you need to target hyper-athletic players and hope their 'end product' reaches a high level before their physical level deteriorates too far. This is why wingers like Yan Diomande and Yankuba Minteh are wildly valuable. The longer explanation: We're looking at all wingers in the Premier League w/ 900+ minutes in a season since the 18/19 campaign; metrics below: - "End Product Rating": (npxG+xA) per (shot + key pass) multiplied by npG+A p90... quality of action x realized outcome - Dribble Actions: Prg Runs + Dribbles Won p90 The first graph is all wingers with the age on the x axis (19 is 19 and below, 31 is 31 and older), and the y axis is the average z-score of each player at that age for end product rating and dribble actions. The takeaway here is the pretty obvious decline for wingers on the physical side, but largely variable status of end product rating. The second table shows the year over year changes of individual players as averages for carry actions, and the end product metrics split out separately. Carry actions show a YoY decline and much smaller standard deviation than the end product metrics. End product metrics all show YoY gains and wide variance from season to season, with the least variance being expected goals and assists (no shit). If carrying actions decline at a more predictable rate starting from the early 20s, and end product, as elusive as it is to quantify, has little meaningful pattern except that it is highly variable (with the exception of npxG+xA), the answer is pretty clear to me: To get game-breaking wingers with maximum impact, target younger players with the highest starting physical ceiling, as this will behave in the most predictable pattern, provided they have at least some semblance of an expected output floor. Then, hope (or genuinely develop the player) that the end product improves or has a couple season spike while the physical element is still elite. This is how you get wingers like Bukayo Saka, Antoine Semenyo, and Rafael Leao: currently in their physical prime still, but end product has come around, despite not being fully developed in their u21 stages. But what about young wingers with ready-made end product but a lower physical starting point? They're not wingers, they're a dying breed of no. 10 that require very specific platforming to maximize their genuinely big talents... think Phil Foden, Dani Olmo, etc. Nothing against the players, but they're not wingers for 99.9% of teams. What about young wingers with both? Kylian Mbappe's and Lamine Yamal's don't grow on trees. Is this an exact science? Absolutely not. Sometimes you'll end up with Sadio Mane, but sometimes you'll end up with Adama Traore. Conclusion: If you have data or general conjecture to the contrary of what I've said, my opinions are not allergic to new information. There are plenty of holes in the data I've presented, this is just what I could cook up to roughly demonstrate an idea that's been bouncing around in my head for some time. This is simply a framework to understand that, for true wide area wingers, considering the knowns of physical deterioration with age/minutes, the variance of "end product", and the slight signal that it improves with age, your best chance at getting a winger that truly elevates a top level team is to find the best athlete that meets a certain xG+xA level and do everything you can to maximize the chances that the spike output seasons happen before their legs leave them.

@EmmanuelMacron For sure ❤️

“This clown wants to make France an AI leader with €30M.” €30 million → to attract and support around forty top-tier international researchers. They chose France for its values and its commitment to science. Sometimes it’s too slow… €54 billion → mobilized as early as 2022 to build France 2030. A France that advances health, climate, and fundamental science by investing boldly in artificial intelligence. Over €100 billion → in private investment announced at the Paris AI Summit by French and international companies to develop AI in France. That’s it? #1 → In 2025, France ranked first among countries attracting foreign investment to build data centers. More billions invested in our regions and in our talent. Yes. Here in France, we believe in science. And AI is here to elevate, to build, to move the world forward, not to insult. 🙊 #ForSure

This is actually insane. If Liverpool, Man Utd or Spurs were achieving this the media would be salivating. As it's Arsenal they are thinking of any excuse to downplay it. Always been the way! #Arsenal

You’d never know this from the way the team is talked about but Arsenal rank 2nd in goals scored in the Premier League and just finished the League Phase of the Champions League with the most goals scored.

📊 Why Robust Z-Score Matters - A standardization method designed to reduce the impact of skewed distributions and outliers - It can be seen as a middle ground between Percentile and Z-Score 🧵Thread.

xG doesn't determine your league position but can it predict the future? We found that across 5 Premier League seasons, the bigger the gap between results and underlying performance, the harder teams regress the following season Walk with me [A THREAD] 🧵 ✍️ @sjatfkb