Quantitative Analysis vs Price Action
Quantitative analysis applies the same specified rule to every occurrence, so its behaviour is identical whatever the trader is feeling. Price action reads the bars in the moment and can respond to circumstances no rule anticipated, at the cost of doing something different each time.
Both of these read price. One decides in advance exactly what it will do and then does that; the other decides in the moment. Neither is better in the abstract, and which one you need depends on how your results actually go wrong.
What each one is
Quantitative analysis specifies a rule precisely enough to apply mechanically, then measures the result on data the rule was not built from. Quantitative analysis covers the method.
Price action reads the bars in the moment — rejections, closes, structure — without a fixed specification. Price action covers it, and technical analysis covers the wider tradition.
One is fixed and the other is responsive. Whereas a rule cannot notice that today is unusual, a person can — and a person also cannot help behaving differently after three losses, which the rule is immune to.
Where they differ
Whether behaviour varies with your state. A rule executes the same after a winning week and a losing one. A discretionary reader is measurably different when tired, frustrated or recently stopped out, and usually cannot tell from the inside.
What happens in a situation nobody anticipated. A rule applies itself confidently to circumstances it was never designed for. A person can recognise that something is different and stand aside, which is a genuine advantage and the strongest argument for judgement.
Whether a record means anything. A rule’s history is reproducible — anyone can run it and get the same result. A discretionary record mixes the method with the person’s state at each decision, so it cannot be separated into what worked and what was a good day.
How each handles noise. Direction runs on this site’s shared series average 2.01 bars with a longest of 11. A rule filters that by whatever threshold it specifies. A reader interprets it, and there is always something to interpret.
Where they agree
Both read past price and neither has access to anything the market has not printed.
Both can be fitted to history. A rule can be tuned until the backtest looks good; a reading can be constructed to match what already happened.
Both need position sizing from elsewhere, since neither supplies one.
And both cost a round trip when acted on — 0.0098 here, about 2% of the median bar range of 0.493.
Which one to use
Use a rule when your results vary more than your ideas do. If two similar setups produced different behaviour because of how the week had gone, that is a consistency problem and a specification solves it directly.
Use judgement when conditions are outside anything you specified. A market behaving in a way you have not seen is exactly where a rule is dangerous and a person is useful.
Use a rule as the default and judgement as an override you must justify. Writing down why you overrode it turns discretion into something reviewable rather than invisible.
And when you cannot state your reading precisely, notice that. It may still be good; it cannot be checked, and that is worth knowing before sizing a position on it.
Why consistency is usually the binding constraint
Because the same method applied inconsistently produces a record that means nothing. If the rules change with your mood, then a bad stretch cannot be attributed to the method or to the execution, and there is nothing to improve from.
And because the adaptability argument is usually theoretical. Genuinely unprecedented conditions are rare; the daily experience of discretionary trading is mostly ordinary conditions being interpreted slightly differently each time.
The original data
Of the 24,971 videos in the search corpus, no title compares these two directly. Price action appears in 479 videos at a median of 21,273 views across 288 channels. Quantitative analysis appears in 1 video, at 1,456 views.
Four hundred and seventy-nine videos against one. The most discretionary approach on this site is among its largest subjects and the discipline of specification has a single video in the entire corpus — which is a fair description of what is easy to teach and what is not.
On the chart above the honest answer for most people is no, and that inconsistency is invisible in every record that does not track it.
When it fails
The characteristic failure in discretionary reading is that the record cannot be audited. Every decision blends the method with the reader’s state at that moment, so a losing quarter has no attributable cause — the setups may have been poor, or fine and executed differently on bad days, and nothing in the trade log distinguishes those. The trader responds by adjusting the method, which was possibly not the problem, and the cycle repeats. A specification is what makes the two separable, which is most of what it is for.
A second failure in quantitative work is over-fitting, producing a rule that describes history and predicts nothing.
A third is a rule applied in conditions it was never specified for, where it acts confidently and without basis.
A fourth is overriding a rule without recording why, which converts a testable system back into discretion while keeping the appearance of one.
And a fifth is treating either as complete, since neither supplies position sizing or a risk rule. Both answer only the question of whether to act and in which direction, and the question of how much is left entirely open — which is the one that determines whether a run of ordinary losses is survivable, and the one neither approach will prompt you to answer.
Related
Quantitative analysis covers specification and testing. Price action covers reading the bars in the moment. And technical analysis covers the wider tradition.
The honest question is which failure you are more likely to have. If your losses come from doing something different every time, a rule fixes that directly. If they come from a rule firing in conditions it was never meant for, judgement is what you need — and far more people have the first problem.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.