Day Trading Tips That Survive Contact With Data

Most day trading tips are folklore with a confident tone. Cut your losses. Trade with the trend. Don't overtrade. All true, all useless, because none of them tells you what to do differently tomorrow morning.

What follows is drawn from a study we ran on our own data: 142,348 simulated opening range breakout trades across 104 liquid US stocks and ETFs, with 3,106 of the fills measured at one-second resolution to see where orders actually filled rather than where a backtest assumed they did. The strategy itself is not the point here. The point is that a large honest sample keeps contradicting the advice everyone repeats, and the contradictions are specific enough to act on.

1. Your win rate is not your edge

In that sample the canonical five-minute opening range breakout won 53.9% of the time. It also made almost nothing, about +0.011R per trade before the fill work.

Then we added a filter that everybody recommends, a 5% penetration buffer so you only take breaks that clear the level convincingly. The win rate went up to 56.0%. Expectancy did not move at all.

That is the whole lesson in two numbers. The filter removed some losers and an almost identical share of winners. You felt better and earned the same. Most "improvements" to a strategy do this, which is why you should measure changes in expectancy and never in win rate. A tip that raises your win rate and says nothing about your average win is selling you a feeling.

2. Find out where your orders actually fill

This is the single highest-value hour of work available to most traders, and almost nobody does it.

When we zoomed 3,106 real breakout triggers to one-second resolution, the median slippage against the level was zero. Half the fills came in at the level or better. That is the good news, and it is the opposite of what most traders assume.

The bad news is the tail. The 95th percentile was +0.13R, and those bad fills clustered on exactly the fastest, best-looking breaks, the ones you most want to be in. Averaged out, the honest cost was +0.0066R per trade, which converted an apparent edge of +0.023R into a real one of +0.016R. Roughly a 29% haircut, and every cent of it came from a minority of trades.

Go and pull your own last fifty fills. Compare each one to the price that triggered your entry. If the gap is materially worse than zero on your best setups, your problem is order type and venue, not analysis.

3. Know which side your edge is on

The same mechanical setup, taken long, produced a profit factor of 1.05. Taken short, it produced 1.009 across 71,108 trades, which is nothing at all.

Mechanically the two are identical. Economically they are not, and any strategy you run has a version of this asymmetry hiding in it. Break your own results down by direction before you conclude the strategy works. A blended number can be a real edge on one side and dead weight on the other, and the blend hides both.

4. Trade the stock, not the index

SPY was net negative in the same study, a profit factor of 0.94, while 68% of individual symbols were positive.

The index is the average of thousands of decisions and it mean-reverts intraday far more readily than a single name with its own catalyst. Strategies built on breaks of a level tend to work where there is a reason for the level to matter to somebody. That is much more often true of one company on a day it has news than of a broad basket on an ordinary Tuesday.

5. Size from the stop, never from conviction

Position size should be a calculation, not a decision. Take a fixed fraction of equity as the amount you are willing to lose, measure the distance from entry to the point where the idea is wrong, and divide. The number that comes out is the size.

The order matters more than it sounds. Traders who pick a size first and then look for a stop that fits end up placing stops at the distance their size can tolerate rather than at the distance the instrument respects, and a stop at the wrong level gets hit by noise that has nothing to do with their idea.

If the arithmetic returns an uncomfortably small number, that is information. Either the stop is too wide for the account or the trade is too expensive to take. Both answers are better than the third option, which is taking it anyway.

6. Stop looking for a better entry and look at your exits

Almost all published trading content is about entries, and almost all of the variance in outcomes sits in exits.

In our sample, trades that reached the first target and carried the remainder to the close averaged +0.32R. Trades flattened at the end of the day without reaching a target averaged −0.21R. Same entries. The difference is entirely in what happened afterwards, and specifically in whether there was a rule at all.

Write the exit down before you enter. Not "I'll see how it acts", which is the phrase that precedes most large losses, but the actual price at which you are wrong and the actual conditions under which you take something off.

7. Keep a journal that can prove you wrong

Most journals are written after the trade closes, which makes them a record of your explanation rather than of your reasoning. The version that works records the reason for entry before the fill, along with the level, the invalidation and the size. Then a week later you can compare what you thought with what happened, which is the only mechanism by which anyone actually improves.

Grade decisions, not outcomes. A correct decision that lost money is a good trade and it will pay over a hundred repetitions. A reckless one that won is the most expensive thing that can happen to a new trader, because it gets repeated.

8. Expect the tail, because it is where the damage is

The through-line of everything above is that trading outcomes are dominated by a minority of events. The fill cost lives in a fat tail rather than in the typical trade. The strategy's edge lives on one side and one instrument type. The account damage lives in the run of six or eight consecutive losers that is entirely unremarkable in any method winning fewer than 60% of the time.

Plan for the tail and the average takes care of itself. Plan for the average and the tail removes your account.

What this does not tell you

None of the above says what to trade tomorrow. It is calibration, not a system, and a study of one strategy on historical data cannot tell you what this morning's auction is doing.

The part that is genuinely hard to learn from an article is judgement under time pressure: deciding whether a level matters while it is being tested, with money on it. If you want to watch that happen with the reasoning said out loud, our desk runs live every session and you can sit in for three trading days.