From “I Think This Works” To “I KNOW It Works”

2001 – I think this works.

Charts on a wall. Handwritten notes. A setup I’d seen a few hundred times that seemed to pay more than it lost.

Pattern recognition dressed up as an edge.

The confidence was faith, not proof. I was right about the setup – but if you’d asked me to explain why, I’d have shown you a chart and said something like “it just works.” Not exactly rigorous.

That’s how most trading careers start. You spot something. It works a few times. You do it again. You build a mental model, but the model lives in your head. You can’t hand it to anyone else and get the same result.

I was already sitting on the setup that would become the spine of the next two decades. I just didn’t know it yet.

2015 – I know this works.

Fourteen years later. Every trade logged. Every win, every loss, every deviation from the rule.

The setup that started as a hunch now had a track record I could stack on a desk.

When I placed a trade, I placed it with muscle memory and a spreadsheet behind me. If someone asked why this setup and not another, I had numbers. Not enormous numbers – a few hundred trades logged manually is nowhere near what a modern backtest gives you – but enough that “I think” had graduated to “I know.”

Still, “I know” had a limit.

I knew because I’d lived it. That’s real, but it doesn’t scale. You can’t hand your muscle memory to another trader. You can hand them your rules and your log – but the rules stay a little fuzzy at the edges because you’re the one filtering them through experience.

The setup was 80% mechanical, 20% discretionary. And the 20% was where I lived.

2022 – AI and a shit tonne of data recalibrate everything.

The same setup, run through 21 years of historical data. Backtested a dozen different ways. Forward-tested. Stress-tested against every market condition since 2001.

Here’s what I expected: the data would either confirm the setup or blow it up.

Here’s what actually happened: the data agreed with the setup – and refined it.

The gut feeling from 2001 was right. But the numbers tightened the entry criteria. Sharpened the exit rules. Filtered out false positives I’d been eating for years without realising they weren’t part of the edge.

The 80% mechanical stayed. The 20% discretionary got broken down into rules I could see, test, and remove.

That’s the piece most traders miss about “trusting the data.” The data doesn’t overturn a real edge. It refines the last 20% you couldn’t quite articulate. Everything you were doing on instinct becomes something you can hand to another trader and expect similar results.

Now – I KNOW it works. The data says so.

Not because I feel it. Because 2,000+ verified trades, run across two decades of market regimes, keep confirming the same 72% win rate.

The math holds through 2008. Through 2020. Through 2022. Through whatever we’re in now.

The setup fires. The software flags it. The rules execute. The result posts. Win or loss, the record grows.

I don’t need to defend the setup any more. The record defends it.

What actually changed across the three eras

The setup itself? Barely changed. The mechanics I trade today are recognisably the ones I stumbled into in 2001.

What changed is how much of it lives outside my head.

• 2001: 0% quantified. 100% in my head.
• 2015: 50% quantified in a manual log. 50% in muscle memory.
• 2022: 80% quantified in code and data. 20% still in judgement.
• Now: 95% quantified. The remaining 5% is knowing when to sit out entirely – and even that is getting rules defined.

Each era stood on the previous one.

Discretion built the hypothesis. Time built the log. Data proved the hypothesis. Software scaled the proof.

Skip any one of those and the whole thing collapses. A gut feeling with no data is just a story. Data with no gut is a backtest that overfits to history. Software with no rules is a random number generator.

You need all three. In that order.

The question I’d ask a newer trader

Do you trust the setup because it’s yours, or because it’s testable?

Both answers are valid at different stages. But if you’re still on “because it’s mine” ten years in – and you can’t hand the setup to someone else and get similar results – then the setup lives in you, not in the market. That’s a career limit disguised as an edge.

The work of the next decade is turning “mine” into “testable.”

If you want to see what testable looks like in practice – the actual setup, the actual rules, the actual verified track record – the collect morning income here.

If you’d like to know a little more about credit spreads first – click here for credits spreads