Seeing all the CAW debate in horse racing, but not many solutions being shared about what to do. From a technical lens, here are some known strategies for slowing CAWs the tracks could try to implement.

So @stoolpresidente you hate the betting bots in horse racing too? We’ve got ideas and skills to stop them and put horse racing back in the hands of the people. We’d love an El Presidente for our business. Want to build this together? @barstoolsports @stoolgambling #ForThePeople

This is a graph database, and we’re confident it’s the future of thoroughbred intelligence! Graphs are widely used in global supply chains and social media networks, and we see huge value for everyone in building a graph for this industry. The image on the left shows the

We’re not trying to point out mistakes — we want to help you get this right for the whole industry @Equibase @chancemoquett . Here’s a simple and practical way to help identify and fix these issues: plot purse earnings (or other success metrics) on the X-axis and rating score on

Forever Young shows two @Equibase ratings (106 and 67) on Oct 17. Even if ratings update daily, two outputs on the same date signal a database bug and lack of proper testing. While you’re at it, please fix the bug causing Ushba Tesoro (57) — a horse with almost $19M in earnings

We have contacted Equibase and offered assistance in correcting issues with their new rating system. Our recommendation calls for implementing ratings similar to chess Elo scores that clearly categorize horses based on their distance, surface, and running style.

Given our background in mathematics — and seeing the backlash @Equibase is getting — we felt compelled to share our findings from stress-testing the math behind their new rating system. To our own surprise the math checks out — but there are clear gaps that explain how

Today we officially open-sourced all the code for the AI Pivot Table we use inside ThoroBid Moments — free for anyone to use, build on, or adapt. This tool lets you slice and explore any complex dataset with AI-powered grouping, filtering, and aggregation. And this is just the

We took a still race photo and used AI to keep @jose93_ortiz & Classic Q running 👀 this is amazing

$0 for AI analytics at the 2025 Keeneland Yearling Sale. 👉 https://t.co/Pie06nsqhK Use it, break it, tell us what you want and we’ll update it. Others will charge you for weak AI and analytics. We built this in 2 days and made it free for everyone. Data access is the change

LFG 🚨 unreal purchase at #KEESEP for @lmgriffjohnson & @westpointtbred with Hip 100 — Flightline x Into Mischief filly for $350k. Our machine learning models flagged this as such a large outlier we thought it was a data error — that’s how far off the projection was 🤯 Great

Real-time bloodstock analytics in action 👊 Our computer vision model for bid momentum is dialed in — scanning hip #'s + final bids directly from Keeneland’s YouTube feed. Frames are converted to high-contrast black & white so numbers pop for maximum accuracy. As soon as the

F*ck PDFs! 👊 Thoroughbred data is paywalled or buried in unusable formats — hips, pedigrees, PPs, Auction Edge. We’re building a platform to make research & auction management simple at any scale with ML, generative AI + computer vision, just like other pro sports. Great to see

Toughest part of @keenelandsales -- spotting the next stakes winner… or spotting an outlet for your iPads and laptops?

We flagged this to Keeneland and told them how to fix it — but don’t hold your breath. We’ve been running a computer vision model on their YouTube stream to track bid momentum. YouTube’s pipeline adds 15–30s of latency (buffering, bitrate, etc.), while their site feed runs

SF/Starlight/Madaket buys 18 colts so far at #KEESEP while doubling down on proven lines. • 4 Gun Runners (1/3 of spend on $11.9M) • Big stretch buys on Good Magic & Nyquist • Zero first-crop punts That’s not diversification — that’s conviction. 👊 @DonatoLanni