First, I'd trade micro futures or very small position sizes, so no single trade could seriously damage the account. With $1,000, survival is the whole game. One oversized loss and you're done before you've learned anything.

Second, I'd treat that account as tuition, not as an income source. The goal wouldn't be to double it. It would be to collect data about myself: which setups I execute well, where I hesitate, when I break my rules. I'd journal every trade and review it honestly every week.

Third, I'd automate early. Not necessarily full automation, but at least stops, targets and trade logging, so emotion has fewer chances to interfere. And I'd test every idea on historical data before risking a dollar on it. Knowing what I know now, I'd spend far more time testing and far less time trading in the first year. Patience at the beginning pays for itself many times over.

The biggest factor was the improvements I made to my system. After the 2024 championships, I went back through my trades in detail: what worked, what didn't, where the system left money on the table, and where it took losses it didn't need to take. That review led to a series of refinements to the system's logic and how it manages trades.

None of these were dramatic overhauls. The core approach stayed the same. But in trading, small improvements compound. Cutting out a category of weak trades, managing winners a little better, tightening the conditions for entry: each change adds a few percent, and together they make a big difference. The 2024 result was solid but cautious. By 2025 the system was sharper, and I trusted it more, which mattered just as much. Confidence in your system lets you execute it fully instead of second-guessing it at the worst moments.

I'd add a risk-adjusted ranking alongside the raw-return leaderboard. Something like return divided by maximum drawdown, or a measure of consistency across the quarter.

Right now, pure return rewards whoever takes the biggest swing at the right moment. Sometimes that's real skill, and sometimes it's a lucky bet with oversized leverage. The trouble is that newer traders look at the leaderboard and copy the behavior that got someone to the top, even when that behavior would destroy an account over a year. Showing both numbers side by side would tell a much more honest story. It would teach people that how you made the return matters as much as the return itself, and it would recognize the traders who manage risk well, which is ultimately what keeps you in this business.

For me, almost always from watching the market. After enough screen time, you start noticing things that repeat: a certain behavior around the open, a typical reaction at a prior day's high or low, how a market acts after a news spike. That observation is the seed.

Then the data takes over. I write the idea into code and test it across years of history and different market conditions. Most ideas die at this stage, and that's healthy. What I'm looking for is an idea that holds up across many market periods and parameter settings, not one that looks spectacular in a single window.

I'm wary of ideas that start purely in the data. If you search through enough combinations, you'll always find something that looked profitable in the past just by chance. An idea that comes from real market behavior and then survives rigorous testing has a much better chance of having a genuine reason behind it, and a reason is what keeps a system working after you go live.

Over-optimization, without question. A new system developer runs a back test, sees mediocre results, and starts tweaking: change the moving average from 20 to 23, the stop from 10 ticks to 12, add a filter for Tuesdays. Eventually the equity curve looks beautiful. What they've actually built is a perfect description of the past, which has very little to do with the future.

My rule of thumb is that if a system only works with one exact parameter setting, it doesn't work. A robust system should stay profitable, maybe less so, when you nudge its settings around. I also test on data the system has never seen before trusting it.

The second common mistake is ignoring real-world costs: slippage, commissions and fills. Many beautiful back tests assume perfect fills that never happen in live markets. And the third is assuming automation removes the need for supervision. It doesn't. Code can fail, connections can drop, and markets can do things your system was never designed for, which I learned the hard way, as I'll explain later.

The principles transfer, but the details don't. Trend, mean reversion, volatility expansion and contraction show up in every market, so the core logic of a system can often be shared. Human behavior drives all of them, and fear and greed look similar on any chart.

But each market has its own structure. Futures have defined sessions, contract rolls and specific tick sizes. Stocks have earnings, gaps and sector effects. FX trades nearly around the clock and responds heavily to central banks and macro data. Crypto never closes and has volatility in a completely different league. Options add time decay and implied volatility, which change the entire nature of a position.

So I use a common framework and testing process, then adapt and calibrate each system to the market it trades. Taking a system built for one market and dropping it unchanged into another is one of the quickest ways to lose money.

I never revenge trade. When I take a loss, even a frustrating one I feel I didn't deserve, I don't jump back in to win it back. Revenge trading is one of the most destructive things a trader can do, because it replaces your plan with emotion at exactly the moment your judgment is weakest. One bad loss is a normal cost of doing business. A string of angry trades trying to recover it is how small losses become account-threatening ones.

When a trade goes against me, I accept it, step back, and let the next trade come from the system or the plan, not from the need to be right. The market doesn't know or care that I just lost money, and it doesn't owe me anything. The traders who survive for decades aren't the ones who never lose. They're the ones who never let a loss make the next decision for them.

Curiosity, at first. Crypto was a brand-new market forming in real time, and for someone who had traded traditional markets for decades it was fascinating to watch. It had the kind of volatility and inefficiency that traditional markets had long since lost, and it raised big questions about money, technology and trust. CryptoCritic grew out of wanting to apply a skeptical, data-driven trader's perspective to a space full of hype.

Over time, though, I became very disenchanted with the crypto markets and with most of the companies in them. The longer I looked, the more I saw projects with little or no real utility, tokens that existed only to be sold to the next buyer, and an alarming number of outright scams and frauds. Some of the biggest names in the industry turned out to be the worst offenders. There's interesting technology underneath it all, but the gap between the promises and reality was enormous. That experience reinforced something I've believed my whole career: always ask what's really behind the price, and be most skeptical when everyone around you is most excited.

AI will transform how systems are built. Within a decade, a trader will be able to describe an idea in plain language and have it coded, back tested and stress-tested in minutes. Work that took me weeks in the early days will take moments. That's an enormous opportunity, especially for traders without a programming background.

But it cuts both ways. When everyone can generate and test strategies instantly, the market will be flooded with them, and edges will be discovered and competed away faster than ever. Over-optimization will also become easier, because a machine can search millions of combinations and find patterns that are pure noise.

The traders who succeed will be the ones with good judgment: knowing which ideas make sense, testing them honestly, managing risk strictly, and having the discipline to turn a system off when it stops working. The tools will keep getting smarter, but the human will still make the decisions that matter most.

A bug I didn't catch. Some years ago I was running an automated FX system, and it had a bug in the code that I hadn't noticed. Under a certain set of conditions, the bug caused the system to buy and sell over and over, hundreds of times in rapid succession. Each round trip cost spread and commission, and the system just kept going. It happened overnight while I was asleep. I woke up to find my account down 50%.

It was a painful and humbling lesson, because the trading idea itself wasn't the problem. The code was. And I'm a programmer by background, so I should have known better than anyone that software has bugs.

That experience changed how I build and run every system. Now I build in safeguards at multiple levels: maximum number of orders per period, maximum daily loss limits that shut everything down, position size caps, and alerts when something behaves unusually. I test new code in simulation far longer before it touches real money, and I never assume an automated system is fine just because it's running. Automation removes emotion from execution, but it doesn't remove the need for oversight. If anything, it raises the stakes, because a machine can make the same mistake hundreds of times before a human would make it twice.

The market is the one problem that never gets solved. Every time you think you understand it, it changes. A system that worked beautifully for years stops working, a new market emerges, technology reshapes everything. For someone who has spent a career writing code and solving problems, that's endlessly engaging. There's always something new to test, learn or improve.

I also value how honest it is. The market doesn't care about credentials, opinions or reputation. The P&L tells you exactly whether you were right and whether you managed your risk properly. That honesty is humbling, and it keeps you sharp.

And frankly, I've learned to respect it. I've made mistakes, some expensive, and I've learned from every one. Survival in this business comes from respecting risk, staying curious, and being willing to keep learning. After 30 years, I still feel like a student, and I think that's exactly why I'm still here.

Trade your normal plan, not a lottery ticket. In a competition it's very tempting to size up and swing for first place from day one. Most traders who do that are out of contention within the first few weeks, and some damage their accounts badly in the process.

Treat your first championship as practice under pressure. Learn how it feels to have your results visible on a leaderboard, notice how that pressure affects your decisions, and focus on executing well rather than chasing a ranking. Keep a journal during the competition and review it afterward. You'll learn more about your trading in one quarter of competition than in months of trading alone.

And remember, there's always another quarter. The traders who reach the podium consistently are the ones who show up, manage risk and improve each time. Consistency builds results. Survive first, and the podium will follow.