Brief: failure-preview

Research brief for failure-preview · how it was made.

Research brief - `failure-preview`

Field	Value
Slug	failure-preview
Working title	Failure Preview
Article shape	Specimen Record
Category	Judgment Errors (`je`), next to Regression to the Mean
Intent tags	decision-biases, anxiety, overthinking
AI authors credited	Grok 4.7
Reviewer or checker	none
Date	2026-10-06

1. Concept gate

Reader's real question:

Why does one unlucky episode feel like a trailer for a more important event that has not happened yet?
When is that earlier miss a real clue, and when is it a sample of one?

Why this belongs in the catalog:

No page describes the pair: a smaller or earlier miss, then a more important event expected to fail because the first one did.
Fortune telling and overgeneralization are classical CBT entries and stay on that list. This page does not extend that list.
The hot-hand page is a success streak continuing into the next similar try. The gambler's fallacy expects a chance streak to reverse. Regression to the mean is the ordinary sequel after an extreme, misread as a cause.
The correction is specific: ask what actually travels from the first event to the second.

Bug claim in one sentence:

The bug is failure preview, where one unlucky outcome is treated as a sample of a more important event still ahead, which makes it harder to see whether any cause from the first event will still be there.

What this bug is not:

Not a classical CBT distortion, and not an addition to that list.
Not the claim that an earlier miss never carries information. A shared cause can be a real clue.
Not the gambler's fallacy. Expecting a magical reversal is a different mistake.
Not the hot-hand fallacy. That page is a run of success, continued into the next similar try.
Not a published effect under this name. The name is the catalog's. The sources are the law of small numbers and positive recency for a sequence filed as "how I am doing."

Nearby bugs to check:

Nearby bug	Difference
Regression to the Mean	An extreme is often followed by a more ordinary result. Failure preview expects the miss to repeat in a higher-stakes sequel.
Hot-Hand Fallacy	A success streak is expected to continue into the next similar try.
Gambler's Fallacy	A chance streak is expected to reverse.
Fortune Telling	A bad future is treated as decided, with or without a recent sample.
Overgeneralization	One bad event becomes an always-or-never rule. Failure preview can stay on one pair of events.

Concept gate: specific, distinct, grounded, and correctable - yes

2. Sources and research

Source / concept	Year	What it supports	URL / DOI / note
Tversky, A., & Kahneman, D. "Belief in the law of small numbers." Psychological Bulletin, 76(2), 105-110.	1971	A small sample is treated as highly representative, so a second sample is expected to resemble it more than sampling theory allows. Undue confidence in early trends. The bias runs whether the conclusion is welcome or not. Cases are research samples, not everyday rehearsals.	https://doi.org/10.1037/h0031322 - PDF checked 2026-10-06. Do not cite the questionnaire medians; the first-item OCR is easy to misread. Use the paper's own summary: representative samples, early trends, stability of observed patterns.
Ayton, P., & Fischer, I. "The hot hand fallacy and the gambler's fallacy: Two faces of subjective randomness?" Memory & Cognition, 32, 1369-1378.	2004	Experiment 1: negative recency for roulette colors, and positive recency for the hits and misses of the person's own predictions, on two statistically identical chance sequences. Filing the sequence as "how I am doing" is what makes a recent result expected to continue. They did not test a rehearsal against a final performance.	https://doi.org/10.3758/BF03206327 - abstract and experiment description checked 2026-10-06. No effect sizes in the article.

Do not invent a study of "failure preview." Say what each paper studied, then say which everyday step the page adds.

3. Article plan

Explained boundary:

One unlucky outcome is used as the trailer for a more important event.
A cause that will still be there is a clue. The bug is skipping from the outcome to the forecast.
A magical reversal is not the correction.

Examples planned (4-8 quote examples):

Practice interview, real interview.
Jammed test print, client meeting.
Small favor stumbled, funding conversation.
Awkward first date, meeting the family.
Blanked quiz, exam.
Hallway demo crash, launch review.
Wrong time on a minor appointment, important trip.
One pitch came back no, the big pitch.

Real-world scenarios, if useful:

The rehearsal, the morning of, the early no, the permit before the hearing.

Impact:

The gap between the two events is spent inside an ending already written.
Preparation becomes damage control, or it stops.
A real transferable cause gets buried under "this is how it goes."

Causes:

A sample of one feels like a picture of the process.
Once the miss is filed as a fact about performance, it is expected to continue.
The important event wants an ending, and the smaller one is an ending already in hand.

Self-spotting signs, if useful:

The feared event matters more than the one that already happened.
The link is "because that one failed," with no cause named that will still be there.
A smooth small version would not have settled the big one the way the rough one unsettled it.
Replaying the miss feels like it makes the forecast more solid.

Prevention moves (3-6):

Name the two events separately.
Ask what travels. If something does, fix that.
If nothing travels, call it a sample of one.
Write two other plausible endings before rating the repeat.
Do one concrete preparation step, then rate the event again.

Questions & Answers, if genuinely sharp:

What if the earlier failure shares a real cause?
Should I expect the opposite, since I am "due"?

Reframing examples (2-3):

Label	Original thought	Reframed thought
The practice interview	The practice was a mess, so the real one is already lost.	The practice showed an unfinished opening. The real room is a different draw. Fix the opening.
The jammed morning	The test print jammed, so the client meeting will fall apart.	Check whether anything about the jam will be in that room. If it will not, it is not a preview.
The early no	One pitch came back no, so the big pitch is a no.	Ask what in that no still applies. The rest does not write the ending.

Related thinking bugs:

Bug	Difference note
Regression to the Mean	Ordinary sequel after an extreme, versus expecting the miss to repeat at higher stakes.
Hot-Hand Fallacy	Success streak into the next similar try.
Gambler's Fallacy	Chance is expected to reverse.
Fortune Telling	Bad future decided without needing this sample.
Overgeneralization	Always-or-never rule, versus one pair of events.

4. Rewrite inventory (only for existing pages)

Not a rewrite.

5. Sign-off

☑ Belonging test passed
☑ Nearby bugs checked
☑ Sources verified or uncertainty labeled
☑ Examples are concrete
☑ Reframes are realistic, not forced optimism
☑ Credits model/person named