10 repos that mass replace a $100,000/year football analytics department. all free. all open source.
https://t.co/f88gBwzDuG -> replaces Hawkeye and Second Spectrum YOLO tracks every player and ball from any broadcast. assigns teams by jersey color. calculates speed, distance, possession. from a TV feed. no sensors.
https://t.co/It40HRL56i -> replaces entire quant sports desk stacked ensemble: LightGBM + XGBoost + Neural Networks + Random Forest. scrapes FBRef automatically. ELO with dynamic K-factor. Poisson xG. MongoDB backend. the most complete open-source football prediction pipeline on GitHub.
https://t.co/HBy8KHg5Vq -> replaces paid prediction platforms ($30/mo) full GUI app. 7 ML algorithms. downloads data from football-data. co. uk. predicts upcoming fixtures. exports to Excel. one click.
https://t.co/fbX02JGfYL -> replaces manual feature engineering XGBoost with 354 hand-crafted features. works for any European league. data straight from football-data. co. uk. plug and predict.
https://t.co/N4NBKWJ8yr -> replaces value bet scanners ($50/mo) ELO + expected goals + offensive/defensive ratings. compares model probability vs Vegas lines. flags when you have edge.
https://t.co/FEGRoaWs0x -> replaces bookmaker calibration tools Gradient Boosting tuned to output probabilities that match real bookmaker odds. not just accuracy - calibrated confidence.
https://t.co/LFzoWs4dcD -> replaces StatsBomb xG subscription xG model from KU Leuven researchers. LogReg + XGBoost pipelines. supports Wyscout, StatsBomb, Opta data. academic grade.
https://t.co/ZPQMuCwcAZ -> replaces xG analytics dashboards xG on StatsBomb open data. SHAP explanations for every prediction. proper calibration. tested on FIFA World Cup 2022.
https://t.co/tKj2sBOVzc -> replaces basic prediction models Poisson distribution for goal simulation. the classical approach that still beats most ML models on draw prediction.
https://t.co/Kp046GL1KL -> replaces Premier League prediction services XGBoost + AdaBoost + SVM on EPL data. detailed EDA. confusion matrices. honest 56% accuracy - because football is hard.
like + bookmark you'll need this when you build your first football prediction bot
60+ years of licensed collectibles go fully onchain 🤝
@PaniniAmerica named OpenSea as their exclusive marketplace.
Starting March 30, Panini collectors can bridge digital cards to self-custody wallets and trade directly on OS, keeping transaction history and verified status.
Ownership, your way.
Manchester United and Chelsea both being without head coaches when they are basically on the expected path from the summer is kinda weird.
Tough jobs? Tough politics? Impossible expectations?
I've definitely been harsher on Maresca than Amorim, but neither has been bad!
What an incredible World darts championships from @bettingemporium & @stattobets who is up350 pts on the tournament - truly remarkable tipping performance over the last month. Well done and by far the best 🎯 tipster in the business.
Panini Blockchain is by far the greatest NFT project in the world right now. Legit value exchange for consumers. Robust community. Transparent product roadmap. Qualified leadership. Great product. This is how it's done.
ICYMI: Crypto . com and Underdog have partnered to offer sports prediction markets, per announcement earlier today; Underdog has a sports betting license in North Carolina and is applying for one in Missouri
A "natural extension" as I've always said about digital sports cards.
-no grading fees
-dual use (gaming & collecting)
-no taking pictures or mailing
-no counterfeits
-100% serialized
-burning to reduce supply
-frictionless 24/7 market
-Airdrops. Everyone loves airdrops
-no searching for comp prices. Entire Tx history at fingertips.
-Easy for groups to build on top of the product w/ digital. Not happening with physical.
As part of a recent 'upgrade' it appears @SkyBet are now offering different odds to different customers.
A low value customer might well see for example 80/1 First 6 on Si Woo Kim for the US Open
A high value customer when logged in might well see 125/1 first 8 on the same bet
Hard Rock Sportsbook generated $881 million in gross gaming revenue in Florida in 2024, per a report from analysis firm Yield Sec; that Florida total alone would make Hard Rock the fourth-highest grossing legal sportsbook nationwide, behind only FanDuel, DraftKings and BetMGM
FanDuel, DraftKings, BetMGM & Fanatics are among the likeliest sportsbooks to go live in Hawaii should the betting bill pass; a group representing the quartet lobbied for the bill; proposed low tax rate and HI's high per capita income could attract others, despite low population
You can learn a lot in sports betting just watching the odds screen move. Given enough hours of thoughtful focus you will be able to reverse engineer some good stuff from it.
I've recommended that on several podcasts and the comments always include things like "That's boring! I don't have time for that!"
It gets more interesting when you remember those aren't just numbers -- those are people taking positions.
When I got into financial markets I did the same thing - I spent hours watching the order book.
Guess what, it works there too.
Stop being lazy.
All time high month for #PaniniBlockchain per API data at @cryptoslamio, beating crypto/NFT peak in August 2021.
Blockchain collectibles with fundamentals over utility are here to stay.
@PaniniAmerica
https://t.co/k97f5ZFiJN
Hard Rock Sportsbook, the only legal book in Florida, only has 20% market share in the state: 80% of sports bets placed in FL are with offshore books, per research cited by Next Io's Pierre Lindh at a company conference today
@spinotron It’s getting more difficult to buy the packs. I’ve tried for WNBA and Flawless but not a sniff despite trying to pay from the release time. Hopefully going to the right people and not the bots!
Global offer is a huge innovation on #PaniniBlockchain. 5 before-vs-after’s and a remark:
(1) Before, buyers see certain seller names and might mix offer with feeling/strategy.
Now it’s almost like stock market, within population for each, sales happen when demand & supply meet.
(2) Before hard to chase a card with multiple listings:
-bidding tends to happen under new names where less hold out and more tendency to recycle money
-buyers might end up with 0 or unwanted multiple.
Now buyers can reach all the sellers and reach the market floor with one offer, and bidding is accessible to all sellers (older users are not excluded).
(3) Before sellers have all the leverage, as offers are binding and last for 24 hours
-sellers can shop and wait until last moment
-buyers budget can lock up, not as nimble
-bidding wars can last for days and weeks even.
Now buyers are as empowered as sellers, buyers can revoke offers and can set offer length, putting urgency on seller and can shorten the extended bidding interval (from 1 day to 1-4 hours).
(4) Before one cannot make your bounty official and guys with wrong time zone is at a disadvantage.
Now you can put bounty via global offer before cards get pulled. At least sellers will not dump unknowingly and your interest is manifested and seller can reach out and hear from you before selling.
(5) Before each card is their own being, and unlisted cards won’t affect supply.
Now co-owner risk is more prominent, as print run = supply, each owner is put in seller’s position, and the only way to stop a seller caving is his understanding of his asset, thinking of trajectory and his cash flow need. This makes meaningful 1/1 and coffin cards more rubust and stable.
Remark: Initially owners without proper knowledge, understanding and assistance/guardrails of their PaniniBC assets, when all their cards are susceptible to be reached by others, some guys sold very good cards at abysmally low prices at the cusp of a great run, not only miss out FMV but also miss out being part of Blockchain revolution in the hobby. But after the initial shake out and pain from a peer’s perspective, market has become respectable in multiple fronts, overall stable and strong, and will only get better and more robust.