Data-Driven Analytics
Here's what the data says.
Prediction models, market analysis and scoring systems — built, tested and graded in public. Sport, property and everything in between.
Latest Articles
Seventeen Hours Ahead
The Rams and 49ers took opposite approaches to the NFL's first game in Australia. The research says both might be right. The market says it doesn't care.
Group Three
At what class does the median horse stop paying for its own keep? The answer survived every cost assumption I threw at it.
Six Point Nine Per Cent
The median horse breaks even at Group 3. This piece asks how many horses actually get there. The answer is 6.9%, and the ladder between maiden and Group 3 is steeper than the prizemoney tables suggest.
The Number I Nearly Published
The median foal returns 99 cents per service fee dollar. That was the headline, a day from publication. It was wrong, and the correction tells a better story.
What the Sires' Premiership Actually Measures
Five of seven new top-30 sires arrived on a single horse's season. The table adds up to the cent. It just doesn't answer the question its readers are asking.
The Horse Racing Money Problem: Who Gets Paid?
Australian racing paid out $996 million last season. 1,677 horses collected half of it. Here's the full distribution.
48 Groups, 104 Games, One Spreadsheet: Grading a World Cup Prediction Model
The model called 35 of 48 group positions and all four semi-finalists. It also picked the wrong winner. Here's the full scorecard.
How We Score Every Suburb in Australia (and Why Most Rankings Get It Wrong)
14,643 suburbs scored on ten dimensions. No listicles, no vibes, no real estate advertising — just data, weighted transparently.
What 1,054 NRL Matches Taught a Machine About Winning
A twenty-season statistical model picks NRL winners at 68 per cent. The secret isn't the forwards pack — it's the spine.
Super Bowl LX: Seahawks vs Patriots EPA Model Breakdown
An EPA model built on 48,578 plays found two points of value on the Patriots spread and a classic under spot. Here's the full breakdown.
Two From Three at the Top: Scoring a Melbourne Cup Ratings Model
The model's top three picks included the winner and the third placegetter. Its number one pick ran dead last. That's racing.