TRADING PSYCHOLOGY, MEASURED.

Most writing about trading psychology is about feelings: fear, greed, patience, confidence.

The traders this blog draws on mostly talk about something else. They talk about rules, and about what happens to the numbers when the rules get broken. FOMO is an entry far above the trigger. Revenge trading is a burst of new positions after a loss. And fear is a stop moved lower so the trade can have "more room".

That matters, because a feeling is hard to fix and a number isn't. Every one of those habits shows up in a trade journal as a pattern you can count, and each one has a price you can measure.

Take the breakouts from our own backtest. Buying the same trades one day late, with the same stop and the same exits, turned +376.1R into −42.4R. And doubling size after every loss produced no extra return, while making the typical drawdown about 16% deeper.

Rules over feelings

Kristjan Kullamägi (Qullamaggie) is blunt about where discipline comes from. His advice is to "throw away your trading psychology books" and study a setup until you know it cold.

In his view, mastering one setup fixes the entries, the hesitation and the discipline all at once, because conviction comes from having seen hundreds of examples, not from willpower. He built his own conviction from a database of chart examples covering before, during and after each move.

The same idea runs through the other sources, and a simple rule sums it up: no setup, no trade. Write down your rules of engagement (what you buy and how you sell) before you build a single watchlist. Vague rules slow you down in the moment, while specific ones make decisions quick.

None of them claim that rules make the urges go away, though. Qullamaggie estimates he's put in more than 10,000 hours, and he still struggles with overtrading and overriding his own sell rules.

A rule isn't there because you're calm. It's there for the moments you aren't.

What they say about losses

Every trader in the notes treats the loss as part of the job, and decides it before entry.

If the stop is hit, sell. Qullamaggie's point is that hesitating turns a one-unit loss into two or three, and that not having to think about it frees up your attention for the next trade. He also says never to widen a stop. Dan Zanger's line, which he quotes, is "Love your stops, not your stocks."

Place it as a real order. Use hard stops placed at entry, not mental ones. A mental stop is the one you talk yourself out of, or forget to enter, on exactly the day the stock gaps through it.

Expect to be wrong most of the time. Win rates of 25–35% are normal for breakout trading, and Qullamaggie's view is that you spend most of your time in drawdown, even in a great year. One study of losing streaks puts the time in drawdown for low-win-rate trend systems at about 70–80%.

The market is always right. Exit now and evaluate later, with no ego about it. Marios Stamatoudis adds that you need to know what normal looks like for your own strategy, so a normal bad run doesn't feel like a broken one.

Streaks, and sizing down

Long losing streaks just come with this kind of trading, and they don't mean your strategy has stopped working. If you win about 30% of the time, you're more likely than not to hit ten losses in a row somewhere in 100 trades. In our 50-trade simulation at that win rate, the longest streak was usually around eight, and it ranged anywhere from five to fifteen. Those rough patches come from the same strategy and the same edge as the stretches that felt easy, so the streak on its own isn't really the problem. What does the damage is what you do with your position size while you're in one.

The math here is brutal. Eighteen full losses in a row would cost you about 4.4% of your account if you risk 0.25% per trade, but about 30.5% if you risk 2%. That's why sizing up to win back a drawdown backfires: you end up making your biggest bets right when the strategy is paying you the least. Marios Stamatoudis comes at the same idea from the other direction. It's the rare big mistake that really hurts you, not the frequent small one, and when you keep a small, fixed risk on every trade, no single loss can do much damage, so you can actually think through your decisions instead of reacting to them. So when you're in a streak, cut your size, start back small, and only build back up once your trades start working again. A monthly loss limit of around 5% of your account takes that decision out of your hands.

Mark Minervini, in notes on his 2021 interview, puts the rule simply: never double down on a loser and never size up after a loss. The trap that rule guards against is a death spiral: a losing streak plus idle cash pulls you into more trades and bigger ones, in exactly the market that is not paying. The fix is to watch your last ten or so trades as an early warning, and to trade smaller or not at all when conditions turn.

The opposite mistake is quieter: after a long run of losses, hesitating to take the next valid signal. Sitting out after a streak feels prudent, but the few large winners that pay for the losses can't be predicted in advance, and skipping setups is how traders miss the trade that pays for the streak. Sizing down keeps you in the game; stopping takes you out of it.

Not chasing

Qullamaggie's stream notes are clear on missed entries: accept the miss, and don't chase. In his telling, chasing out of FOMO is what drains accounts.

You can turn that into a number. Skip a trade when the low of the day is already more than 60% of the stock's ATR below the price, because a stop at the low would be too far away. And open at most three new positions per session.

That's different from a planned re-entry. Qullamaggie will buy a stock back, even higher, if it stops him out and then retakes its highs minutes later. That re-entry has a trigger and a stop. A revenge trade has neither. It just has a loss to win back.

A blunt guard against the second kind is to stop trading a ticker for the day after three stop-outs in it.

Why psychology shows up as data

An emotion you felt last Tuesday is gone. But the trade it produced is still in your journal, with a time, a price, a size and a stop.

That's why problems are easier to find in your log than in your memory of how you felt. A periodic review is non-negotiable, and one thing worth checking for is one-to-three-day bursts of hyperactive new positions, because that's where unnecessary trades hide.

Counting the trades that broke your own rules is often the most useful number in the whole review. You can go a step further, too, and journal your emotions and life events, so you can see a bad trading period coming.

Each common mistake leaves a specific trace:

• Revenge trading: more trades on the day after a loss than after a win, and re-entries in the same ticker within a day or two of a stop-out.
• Overconfidence: position size or risk that jumps after a winning run, without a written rule saying it should.
• Fear: losses beyond −1R that a gap doesn't explain (a moved stop), and winners sold long before the stock's move ended.
• FOMO: entries far above the trigger, with stops so wide that the trade was barely worth taking.

Once a habit is a pattern, it has a price. In the next section, we put a price on four of them using real trades.

Four mistakes, priced on 544 real trades

We took the 544 breakouts from our breakout backtest and replayed them with one mistake added at a time.

Here are the rules of the test. Buy as the price crosses the top of a tight base, and place the stop at the low of the day known at that moment (so many trades are stopped out the same day). Sell a third at the close of day 3, and trail the rest on the 21-day EMA.

The universe was about 600 liquid US stocks (today's S&P 500 plus well-known momentum names) from May 2025 to October 2026. That gave about 30 trades a month, with never more than ten open at once, using daily and hourly bars. Traded as planned, the 544 trades won 29% of the time and made +257.8R.

Results are in R, which means multiples of the amount risked (see R-multiples explained). And since rules run by code don't feel anything, these tests isolate the cost of each habit. They can't show how often you'd fall into it.

1. Chasing: the same trades, one day late

The FOMO version of each trade buys at the next morning's open, after the breakout is obvious, with the same stop and the same exits.

450 of the 544 trades could be chased. The other 94, mostly losers, had already opened below the stop the next morning, which is why the planned total for the chased set is higher than for the whole test.

The planned entries made +376.1R, and the chased ones lost 42.4R. The average trade fell from +0.84R to −0.09R, and the winners from 156 to 129.

The reason is arithmetic, not luck. The next open was usually only a little higher, a median 0.61% above the planned entry. But with a stop this tight, that was enough to make the distance to the same stop a median 1.32 times larger. For the same dollars at risk, that means fewer shares. So every winner shrinks in R, while every loser still costs a full unit.

The best trade in the test shows it at its starkest. TXG on May 28, 2026, bought at $25.58 with a stop at $24.99, made +66.1R. Bought at the next open, $27.82, the stop was almost five times further away, and the same trade made +13.0R.

2. Moving the stop: one more unit of room

This is the trader who doesn't sell when a stock hits their stop. Instead, they move the stop down by another full unit of risk to give the trade more "room". To keep it fair, we kept the risk the same: a stop twice as far away means buying half the shares, so a loss at the lower stop still costs one unit. Run that way, the same 544 trades made +161.8R instead of the planned +257.8R, 96R less.

The reason is that most of the extra room doesn't help. 364 of the trades hit their original stop. Only 81 of them bounced back to finish in profit, while 222 kept falling all the way to the lower stop. And because every trade now carries half the shares, the winners make half as much too. If you keep the full share count when you move the stop, the total can look better on paper (+323.5R in this test), but only because every trade that goes wrong now loses 2R instead of 1R, which is a bigger bet, not better trading. If normal noise keeps knocking you out of trades, pick a wider stop and a smaller position before you get in. And once you start moving stops, it's rarely just once.

3. Revenge re-entry: buying back the next morning

After each losing trade, the revenge version buys the same stock again at the next day's open, with a stop the same percentage below, and manages it with the same exits.

The 388 re-entries made +64.5R in total. That's an average of +0.17R per trade, against +0.47R for the planned trades. 275 of them lost a full 1R or more, and the median was −1.0R. Three trades made +133.9R between them, which is more than the whole total.

So in this test, revenge re-entries didn't lose money. What they did was take 388 extra trades, each with a full unit at risk, for less than half the planned trades' average. And even that depended on three lucky outliers.

That dilution is the real cost: more trades, more screen time and more losses to sit through, for very little. Compare that with Qullamaggie's planned re-entry, which needs the stock to retake its highs first.

4. Revenge sizing: doubling up after a loss

This version doubles the risk after every loss, up to four times normal, and drops back to normal after a win.

At a 29% win rate, that means it spent much of its time at the top size, and it ended up risking 2.80 units per trade on average. So the fair comparison is flat sizing at 2.80 units:

R measured in units of the normal risk per trade, trades in the order they closed. The 10,000 orders reshuffle the same 544 results (seed 7). Across them the median totals were about the same: +702.7R flat, +709.7R doubling.

Doubling after losses added no return on average. What it added was drawdown: about 16% deeper in a typical order and 18% deeper in a bad one, because the extra size lands right in the middle of losing streaks. In the actual order, it also made less.

The systematic way to change size, raising it after wins by a written rule, is progressive exposure. And that's a different thing from sizing up because you feel sure.

The four side by side

*Planned here means flat sizing at the same average size, 2.80 units, with trades in the order they closed. 544-trade breakout test, May 2025 to October 2026. The universe is today's S&P 500 plus well-known momentum names and the period was mostly rising, which flatters every row, and the plan itself lost money without its 10 best trades; the results are illustrative, not a forecast.

Chasing was by far the most expensive mistake here: it turned a winning strategy into a losing one. Moving the stop cost 96R once the risk was kept the same, each revenge re-entry earned well under the planned trades, and doubling up made less while adding drawdown. That is a reason to measure your own, because your list will not look like anyone else's.

A mistake list you can tag

A mistake tag is only useful if the list is short and you can check each tag from the trade itself. Here's a starting list, grouped by where the decision went wrong:

Leave the field blank when a trade followed the plan, because that blank is information too. Then, once a month, add up the R on each tag.

A line like "moved stop, five trades, −6R" is something you can act on, in a way that "be more disciplined" never is. The journal template covers the other fields worth keeping, and how to review your trades covers the monthly routine.

A pre-trade checklist

A checklist moves the decision to before the order, when it's cheap. You should be able to answer every item yes or no:

1. Is it one of my setups? If you can't name the setup, it isn't one.
2. Has it triggered? Price through the level, not just near it.
3. Is the stop at a real level, and close enough? For example, the low of the day and no more than one ADR away, as Qullamaggie sets it, or a 60%-of-ATR limit.
4. Is the size from the formula? Shares = risk in dollars ÷ (entry − stop), inside the position cap. (See how much to risk per trade.)
5. Any earnings or news before my exit? Qullamaggie's stream notes say not to buy within three days of earnings.
6. Is the market filter green? (See market environment and your edge.)
7. Am I under my limits? New positions today, open risk and this month's loss cap.
8. Would I take this trade if my last trade had won? If the honest answer is no, it's a revenge trade.

Rules for drawdowns

The time to write these is now, while the account is calm. (Drawdowns and losing streaks covers the maths behind them.)

• Step size down by rule. For example, drop one tier after two losses in a row, or once the account is a set percentage below its high, and only go back up by the same rule. One ladder steps between 0.12%, 0.25%, 0.5% and 1% risk on pairs of wins and losses.
• Never raise size to win it back. The doubling test above is why.
• Keep taking valid setups, just smaller. Trading them at reduced size keeps you in the trade that ends the streak, without risking much to get there.
• Cap the month. 5% of equity is a common cap. At 0.5% risk per trade, that's about ten full losses, and when it's hit, there are no new trades until next month.
• Cap new positions per day. Three per session is a sensible limit. But remember that trading less pauses a drawdown. It doesn't fix the cause.
• Read the last ten trades. It's the best early warning you have. Were they your setups, at your size, with your stops? If yes, it's the market. If no, it's you, and the tags will tell you which mistake.
• Restart small. After a deep drawdown, cut size and rebuild from small positions.