From Chaos to Clarity: The Breakthrough in Filtering Races

It’s been a while since my last blog post. That doesn’t mean I’ve been idle — far from it. Behind the scenes, I’ve been working harder than ever, testing, rebuilding, and refining systems in search of something truly sustainable.

If you’ve followed my journey in sports trading, you’ll know the path is rarely straight. What starts as a promising idea can quickly descend into chaos when faced with the reality of variance, drawdowns, and the unpredictable nature of greyhound markets.

Over the past year, I’ve been testing, discarding, rebuilding, and refining strategies at a furious pace. Some ideas that looked brilliant on paper collapsed in live testing. Others held up, but lacked the scalability I needed. It has been a grind — but also an education in patience, discipline, and letting data do the talking.

The Turning Point

The breakthrough came when I started asking a simple question: Do all races really behave the same way?

The short answer? Absolutely not.

At first, I focused mainly on Open Races (OR) because of their liquidity and visibility. They seemed like the obvious place to hunt for an edge. And while ORs did show potential, the real power came when I went a step further — refining the filters to include only the race types that consistently outperformed the rest.

When I filtered down to this smaller, sharper set, everything changed:

• ROI jumped to nearly 50% after commission.

• Monthly results flipped from inconsistent swings to consistent profits (with the occasional flat month — part of the game).

• Drawdowns became more manageable, even while targeting long odds where variance normally eats traders alive.

The Monthly Reality Check

This wasn’t just a lucky streak. Looking back month by month, the pattern held:

• Some months were spectacular, adding well over +150 units.

• Others were steady builders, adding +20 to +60 units.

• Only one month was close to flat — proof that variance still exists, but also that the edge holds up over time.

You can see the full breakdown here on my Strategies page.

Why Filtering Works

Filtering isn’t about cherry-picking past winners — it’s about identifying repeatable conditions where the market consistently misprices outcomes. By cutting out the noise, the edge became crystal clear.

It sounds simple, but it took thousands of bets, painful drawdowns, and a lot of trial and error before the data finally revealed the path.

How Many Bets Do We Need for Confidence?

One question that always comes up is: “How many bets do you need before you can be sure the edge is real?”

Right now, the Greyhound Filtered Strategy (aka Dog Pound) has been tested on over 1,100 bets — enough to reveal a very strong pattern. But to reach the kind of statistical confidence where the numbers are beyond doubt, we’ll need to keep building.

• Around 5,000 bets will give a very high degree of confidence that the edge is structural.

• By 10,000 bets, the results should be undeniable.

That’s the journey we’re on now — adding more data month by month, and letting the numbers do the talking.

What This Means Going Forward

I’m not about to publish the exact race types here — that’s the “secret sauce” I’ll keep under wraps for now. But the broader lesson applies to any trader: your edge is rarely universal. If you’re grinding away with flat ROI across all markets, there’s a good chance certain conditions are carrying the profits while others are dragging you down.

The challenge — and the opportunity — is to find them.

For me, this filtering breakthrough has given the system a new identity: the Greyhound Filtered Strategy (aka Dog Pound). The next step is more live validation, bigger data samples, and eventually scaling with confidence.

It’s been a long road, but finally, the fog is lifting.