Course contents · Lesson IV of VI

Lesson IV of VI · 8 min · Intermediate

Overtrading: We Built a Machine to Make Our Mistakes Faster

We pointed a bot at every setup we could find. It lost 22.2R in sixty days while the filtered version of the same signals made money. Here's the arithmetic that kills challenge accounts, and what trading less actually looks like.

Updated 2026-08-24 · On the record

Sixty days. One hundred and thirty-five trades. Minus 22.2R.

That's our own bot, running on our own signals, in a stretch where the signals themselves were making money. It didn't malfunction. It didn't get bad fills — well, it got some, and we'll come back to those. It followed every rule we gave it, precisely, all day, without ever once getting bored or scared or hopeful.

That turned out to be the problem. We had built a machine for making our mistakes faster, and it was extremely good at its job.

What the machine did wrong (almost nothing)

Here's the uncomfortable part. Point-for-point, that bot was a better trader than most humans. It never widened a stop to avoid being wrong. It never doubled down after a loss. It never skipped a setup because the last one stung, and it never took a bigger size because it felt due.

It had exactly one flaw: it took everything.

Every setup our platform detected got traded, because that's what we told it to do. And the signal stream it was drinking from was genuinely profitable — the filtered version of those same weeks made money. The bot found the losing half of a winning stream and traded it with perfect discipline.

About 15R of the damage was execution drag — slippage, fills, the small tax you pay for being in the market constantly. Which is its own lesson: the cost of trading is charged per trade, not per good idea. But the other two-thirds was pure selection. It wasn't beaten by the market. It was beaten by its own appetite.

The part that should bother you

We went back through eleven weeks of our own breakout signals and sorted them by how selective we were being. Same market, same setups, same code — just four different bars for what got through.

Figure · fewer trades, more money
The same eleven weeks, filtered four ways
Every setup we detected213 trades-69.7R+ quality score floor68 trades+4.7R+ losing direction cut39 trades+25.9R+ our full filter stack19 trades+29.9RTRADES TAKENTOTAL RETURNBreakout M15, 8 Jun – 21 Aug 2026 · our own ledger, spread charged
One signal stream, four levels of selectivity. Every cut removes trades and every cut adds return — the whole set loses 69.7R, the smallest set makes 29.9R. Spread-charged, R at 1% risk. The trades thrown away weren't unlucky; they were the ones that never cleared the bar.

Read that top row again. Two hundred and thirteen trades, and the account is down 69.7R. Now read the bottom row. Nineteen trades, up 29.9R.

Those nineteen trades were inside the two hundred and thirteen the whole time. The 194 we threw away weren't unlucky, and they weren't a different strategy. They were simply the ones that never cleared the bar — and every single time we raised the bar, the number went up and the trade count went down.

That's the entire thesis of this course sitting in one picture. Trading less isn't a consolation prize you accept because you lack conviction. On our own ledger, it was the edge.

Why this is specifically fatal in a challenge

Out in the world, overtrading makes you poorer slowly. Inside a challenge, it makes you finished suddenly — and the mechanism is worth being precise about, because it's not the one most people picture.

Your losses aren't capped per trade. They're capped per day.

That distinction is everything. Taking more trades doesn't make any individual trade worse. What it does is raise the chance that several ordinary losses land between the same two midnights. Six losers spread over three weeks is a normal, survivable run for almost any strategy. The identical six losers on a Tuesday afternoon is a breached account and a dead challenge, and the strategy performed exactly the same in both versions.

So the thing that ends most challenges isn't a catastrophe. It's a completely unremarkable losing streak that got compressed into a single session by someone who was, in their own mind, working hard.

That's the joke the industry doesn't tell you at checkout. The evaluation isn't scored on how well you trade. It's scored on how well you ration — and effort, the one instinct everybody brings to something they paid for, is precisely the wrong reflex.

What trading less actually looks like

Not "be more patient." That's a mood, not a rule, and moods lose to a red afternoon every time. Concrete versions:

  • A written trade budget, set before the week opens. A number, on paper — trades per day and per week. Reaching it means you're done, whether the last one won or lost. The budget must exist before the temptation, because a limit invented at 2pm is just a negotiation.
  • A daily stop in trades, not just in money. Two losers and the platform closes for the day. Your daily cap will thank you, and you'll be amazed how many breaches live in the third and fourth trade of a bad session.
  • Filters you can check in under a minute. If your A+ definition takes five minutes of squinting, you'll skip it exactly when it matters. Ours is a scoring floor, a direction, and a couple of structural checks — boring enough to apply while annoyed.
  • Flat days on purpose. No trade cannot breach a daily loss limit. It's the only position in the market with a guaranteed floor, and most funded traders spend the majority of their days in it.

Next: the fine print that voids a pass — including the rules on copy trading and signal services that can disqualify an account after it's already cleared the target.

Check yourself

Quick quiz

  1. 1. Our unfiltered bot lost 22.2R over 135 trades. The main thing it did wrong was…
  2. 2. Why does overtrading breach a daily loss limit faster than it breaches a max drawdown?
  3. 3. In our filtered ledger, what happened to total return as the number of trades fell?

Frequently asked questions

Why is overtrading the main reason people fail prop firm challenges?

Because a challenge caps your losses per day, not your losses per trade. Taking more trades doesn't lower the quality of any single one — it raises the odds that several bad ones land in the same session, and a cluster of ordinary losses inside one day is what actually breaches a daily loss limit. Most failed challenges are a normal losing streak that happened to be compressed into an afternoon.

How many trades should I take during a prop firm challenge?

Fewer than feels productive. The honest answer is that it depends on your risk per trade and your daily cap — if you risk 0.5% against a 5% daily limit, you have ten losers of room in a day but should never plan to use more than about half of it. Most people who pass are taking a handful of trades a week, not a handful a day.

Does taking more trades increase my chances of hitting the profit target?

It increases your exposure to your edge and your exposure to your daily cap at the same time — and only one of those has a hard floor attached. Our own logs make the point: the same eleven weeks of signals lost 69.7R if you took all 213 of them and made 29.9R if you took the 19 that cleared every filter.

What counts as an A+ setup?

Something you defined before the week started, in writing, that you can check in under a minute. The specific rules matter less than the fact that they exist and predate the temptation — a filter invented mid-session to justify a trade you already want isn't a filter.

Is it normal to have days with no trades at all?

It's the majority of days for most consistently funded traders. A challenge rewards patience structurally: no trade cannot breach a daily loss limit, and flat days cost you nothing but time. The discomfort of sitting out is the actual skill being tested.

Share this lesson

We built a bot to take every setup we found. It lost 22.2R in sixty days — following our rules perfectly.

While you were reading
Our system watched every pair, every timeframe.

Levels only matter when price reaches them. Breakout Alerts tracks the levels across 12 instruments and alerts you the moment one is actually in play — with entry, stop, and target attached.