How to stop second-guessing yourself is to stop treating the replay as missing research. The choice is already stamped. What remains is a regret loop that compares the discarded option and asks “what if.” A new fact can reopen a call. A feeling of doubt, by itself, usually cannot.
Key takeaways
- How to stop second-guessing yourself starts after the stamp: college seniors who maximised their job search landed 20% higher starting pay and still liked the job less (Iyengar, Wells & Schwartz, 2006).
- People overestimate how long a disappointing outcome will hurt — the durability bias in affective forecasting (Gilbert, Pinel, Wilson, Blumberg & Wheatley, 1998).
- After a choice, people spread the alternatives: the picked option looks better and the rejected one worse (Brehm, 1956), unless the leftover option stays on the desk.
- Anticipated regret changes what people pick and whether they pick at all (Zeelenberg, 1999); treating that forecast as a fact is how a dated call un-decides.
- Derek Sivers’ Hell Yeah or No filter is a practitioner rule: if the yes was not a clear pull, the leftover comparison was already built in.
What second-guessing actually is
Second-guessing is reopening a choice you already made, then scoring the discarded option as if the file were still open. The search phrase “how to stop second-guessing yourself” sounds like a how-to. The useful answer is a recognition page: name the loop, then stop feeding it. The stamp has already happened. What you are doing now is post-decision work — comparing, forecasting regret, and asking whether a different version of you would have picked the other card.
This page hangs off the Decisions parent. The freeze that happens while you are still comparing options is covered in analysis paralysis. The sister page on decision paralysis is the freeze at the stamp: the shortlist is there, and the hand will not come down. This article is the loop after the stamp. The rest of the productivity hub is where the cluster sits.
Jack Brehm asked people to rate household items, then let them choose one as a gift. After the choice, the picked item rose in rated desirability and the rejected one fell (Brehm, 1956). That spreading of alternatives is how a finished choice usually settles. Second-guessing is the opposite move: you keep both cards on the desk, so the spread never happens and the comparison stays live. In our experience, the tell is not “I am still thinking.” The tell is “I already sent the email, and I am still ranking the draft I did not send.”
The practitioner version of a seal sits in Derek Sivers’ Hell Yeah or No: a yes has to be a clear pull, or it is a no. That rule is not a trial. It is a filter that leaves less leftover comparison after the call — which is the part this page is about.
How is second-guessing different from analysis paralysis, decision paralysis, or a real review?
Second-guessing starts after the pick; analysis paralysis starts before it; decision paralysis sits on the stamp; a review looks for a new fact. Those four get used as synonyms in search, and they are not. Task paralysis is a fifth neighbour — the freeze before starting a task you have already chosen — and it is not this loop either.
| Analysis paralysis | Decision paralysis | Second-guessing (this page) | Genuine review | |
|---|---|---|---|---|
| When it fires | Before any option is stamped | At the moment of commit | After the call is dated | At a review date, or when a new fact arrives |
| What it feels like | One more tab will settle it | The shortlist is there; the hand will not come down | The discarded option is still scoring | The original bar is checked against new information |
| What it is not | Missing a load-bearing fact | A hard trade-off you have already named | A safety-critical error you can still undo | Relitigating on a feeling |
| Parent / sister | Analysis paralysis | Decision paralysis | This article | A dated check-in, not an open door |
Why second-guessing happens
Second-guessing happens because the brain treats a finished choice as still reversible, then overestimates how much the discarded option would have healed. The mechanism is not a character flaw. It is a stack of ordinary biases that stay quiet until the stamp, then get loud.
Barry Schwartz and colleagues built the Maximization Scale and found that people who chronically hunt the best option report more regret and less happiness with the outcome, even when the objective pick looks fine (Schwartz et al., 2002). Sheena Iyengar, Rachael Wells and Schwartz later followed college seniors through the job search. Maximizers accepted jobs with starting salaries about 20% higher than satisficers, and still felt worse about the job they took (Iyengar, Wells & Schwartz, 2006). That is the after-stamp pattern in one line: a better-on-paper result, a worse felt result, because the scoring never closed.
Daniel Gilbert, Elizabeth Pinel, Timothy Wilson, Stephen Blumberg and Thalia Wheatley showed that people overestimate the duration of future pain — “immune neglect,” a durability bias in affective forecasting (Gilbert et al., 1998). Wilson and Gilbert later summarised the same family: people know the valence of a future feeling better than they know how long it will last (Wilson & Gilbert, 2005). If you forecast a long ache, reopening the file looks like self-defence. The forecast is usually long. The ache is usually shorter.
Marcel Zeelenberg’s review of anticipated regret is the other half: people change what they pick, and whether they pick, to avoid a regret they have not felt yet (Zeelenberg, 1999). That is useful before a one-way door. After a two-way door has already been walked through, the same forecast becomes fuel for the replay. Christopher Anderson’s review of decision avoidance names the family this sits in — postponing, status-quo, and omission — without claiming that every pause is a freeze (Anderson, 2003).
The four triggers that keep the file open
Four ordinary conditions keep a dated choice alive: a leftover option in view, a long regret forecast, a maximising scoreboard, and no written seal. You do not need all four. One is enough.
- Leftover option. Brehm’s spreading of alternatives needs the rejected item to recede. If the other tool, the other hire, or the other draft stays on the desktop, comparison never closes (Brehm, 1956).
- Anticipated regret. You forecast a long ache and reopen to avoid a pain that usually fades (Gilbert et al., 1998; Zeelenberg, 1999).
- Maximising. You keep scoring the pick against an unseen better version. Iyengar, Wells and Schwartz’s job-search cohort is the cleanest field picture of that move (2006).
- No seal. No written “good enough” bar, no dated review, no Hell Yeah filter. Doubt then counts as new information. Anderson’s decision-avoidance review is the map for that stall (2003). Sivers’ filter in Hell Yeah or No is the practitioner seal — not a trial, a stopping rule.
Why the replay costs more than it feels like
The replay spends the hours that should have gone into the pick you already made, and it can drop satisfaction even when the objective result is fine. That is the cost that does not show up on a to-do list.
Iyengar, Wells and Schwartz’s maximizers did not fail the job search. They won on paper and lost on feeling (2006). Schwartz’s earlier scale work already tied maximising to more regret and less happiness with the outcome (Schwartz et al., 2002). If you treat “I still feel unsure” as proof the pick was wrong, you will keep paying for a result that is already good enough on the bar you would have written yesterday.
Thomas Gilovich and Victoria Medvec’s review of regret is the time-shape most people get backwards: in the short run, people regret actions; over time, they regret inactions more (Gilovich & Medvec, 1995). Neal Roese and Amy Summerville later found that the biggest life-domain regrets cluster around education, career, and romance — paths not taken more than paths taken (Roese & Summerville, 2005). Reopening a dated, reversible work pick to avoid a short-run action regret is how you manufacture a longer inaction regret: the week you spent re-ranking instead of shipping.
Daniel Kahneman and Amos Tversky’s prospect-theory work is the loss frame underneath (Kahneman & Tversky, 1979; Kahneman, Thinking, Fast and Slow). A discarded option that “might have been better” is coded as a loss. Losses weigh more than equivalent gains. The replay then looks rational: you are trying not to lose. You are actually spending the gain you already have. Annie Duke’s Thinking in Bets is the practitioner translation: judge the quality of the bet at the time you placed it, not by the next card that turned. That is a decision-quality rule, not a trial that one named journal “proves” poker thinking.
In our experience the cheap version of the cost is a Tuesday afternoon: the brief is sent, the tool is bought, the hire is offered — and the next two hours go to a comparison spreadsheet that cannot change the email. The expensive version is a month of half-commit, which is how prioritising turns back into ranking theatre.
What second-guessing looks like in a normal week
Second-guessing looks like extra research after a dated call, a leftover option left in view, and a review that has no new fact. The list below is recognition, not a diagnosis and not a six-step system. If you want the freeze-before-start version, that is task paralysis. If you want the freeze-at-stamp version, that is decision paralysis.
| Where it shows up | What it looks like | What it is not |
|---|---|---|
| A tool you already bought | You keep the competitor’s tab open “just to check” after the invoice is paid | A refund window with a load-bearing bug you did not know about |
| A hire or vendor you already chose | You re-read the runner-up’s portfolio the night after the offer went out | A reference call that surfaces a fact the interview missed |
| A brief you already sent | You rewrite the unsent draft and score it against the sent one | The client named a constraint that was not in the brief |
| A weekly ranking | Monday’s #1 is still being compared to the task you parked | A date you set to drop anything that has rolled three days — see the Ivy Lee method |
| A yes that was not a pull | You said yes to be polite, then spend the week building the case for no | Sivers’ filter applied before the yes — Hell Yeah or No |
Barry Schwartz’s The Paradox of Choice is the popular book under the maximising finding, not a second URL for this keyword. The book’s claim is that more options can raise expected regret. The careful read, after Scheibehenne and Chernev, is narrower: more options are a risk when the task is hard and the bar was never written. Second-guessing does not need a supermarket aisle. It needs one leftover card.
An if-then line for the reopen moment belongs to the same family as Peter Gollwitzer and Paschal Sheeran’s implementation-intention meta-analysis: a written plan for a specific cue raises follow-through, with a medium-to-large average effect across 94 tests (Gollwitzer & Sheeran, 2006). The form on this page is not a step ladder. It is one sentence: “If I want to unpick this, I will first write the new fact.” The longer treatment of that form lives on implementation intentions.
Common mistakes
The usual miss is treating the replay as diligence. Extra tabs after the stamp look like care. They are how a dated choice un-decides.
- Opening more research after the email went out. That is the analysis-paralysis move, applied too late. The parent page on analysis paralysis is the before-stamp version of the same habit.
- Treating doubt as data. Gilbert’s durability bias says the ache is usually shorter than the forecast (Gilbert et al., 1998). A feeling is a signal to check for a new fact, not a fact.
- Keeping the leftover option on the desk. Brehm’s spread needs the rejected card to recede (1956). A pinned runner-up is a live comparison.
- Confusing this loop with the freeze at commit. If the hand has not come down yet, you want decision paralysis, not a second-guessing page.
- Confusing this loop with the freeze before start. If the task is chosen and the body will not move, that is task paralysis.
- Using a grit speech as the seal. The seal is a written bar and a review date, or Sivers’ filter in Hell Yeah or No. A pep talk does not close a file.
- Re-ranking instead of choosing the next physical action. That is how how to prioritise tasks turns back into a list of thirty. One essential for the day is enough; reopening yesterday’s essential every morning is the replay in ranking clothes.
When reopening is the right call
Reopening is the right call when a load-bearing fact arrives, or when the original pick is still cheap to undo and the original bar was wrong. This page would be confidence theatre if it pretended every replay is a waste. The evidence above is about forecasts and leftover comparisons. It is not a rule that you never reverse.
Scheibehenne, Greifeneder and Todd’s near-zero mean for choice overload is the first limit: “too many options” is not a universal law (2010). Chernev, Böckenholt and Goodman’s conditions are the second: overload shows up in a subset of hard, unclear, complex choices (2015). A safety-critical call, a legal constraint you missed, a number in the contract that was not in the deck — those are facts. They reopen the file on purpose.
Gilovich and Medvec’s inaction-regret curve is the third limit (1995). If the pick was a polite yes you already know was not a pull, the cheap moment to undo it is now, while the door is still two-way. Sivers’ filter is for the next yes. It does not require you to honour a yes you already regret for a reason you can name. Name the reason. If the reason is a feeling with no fact underneath, hold. If the reason is a fact, reopen and date the new call.
This is a work-and-habits page. It does not diagnose anxiety, depression, or OCD-spectrum checking. Persistent doubt that does not attach to a specific decision is outside this article. Nothing above is medical advice.
What the evidence actually shows
The evidence shows a post-decision loop that can drop satisfaction after an objectively fine pick, plus a regret forecast that usually runs long — not a law that every extra option freezes people.
| Study | Design & size | Finding | What we used it for |
|---|---|---|---|
| Brehm, 1956 | Lab choice after rating household items; classic post-decision dissonance design | The chosen item rose in rated desirability; the rejected item fell | Why a leftover option on the desk blocks the usual settle |
| Kahneman & Tversky, 1979 | Prospect-theory paper; choice problems, not a field trial of second-guessing | Losses weigh more than equivalent gains | Why a discarded “maybe better” option is coded as a loss |
| Gilovich & Medvec, 1995 | Review of regret studies across time horizons | Short-run regrets cluster on actions; long-run regrets cluster on inactions | Why reopening a reversible pick can manufacture a later inaction regret |
| Gilbert, Pinel, Wilson, Blumberg & Wheatley, 1998 | Affective-forecasting experiments on durability bias / immune neglect | People overestimate how long a bad feeling will last | Why a regret forecast is a poor reason to un-decide |
| Zeelenberg, 1999 | Conceptual review of anticipated regret and expected feedback | Anticipated regret changes choices and whether people choose | The forecast that keeps the file open after the stamp |
| Schwartz et al., 2002 | Scale development plus correlates of maximising vs satisficing | Maximising ties to more regret and less happiness with outcomes | The scoreboard personality of the replay |
| Anderson, 2003 | Psychological Bulletin review of decision avoidance | Postponing, status-quo, and omission are distinct forms of doing nothing | Where second-guessing sits in the avoidance family |
| Roese & Summerville, 2005 | Meta-analysis of life-domain regrets | Education, career, and romance dominate what people regret most | Which leftover paths stay loud — and why a work pick is a common replay |
| Wilson & Gilbert, 2005 | Review of affective forecasting | People miss duration more than valence | The “feeling usually cannot” half of the answer box |
| Iyengar, Wells & Schwartz, 2006 | Field study of college seniors’ job search; maximisers vs satisficers | About 20% higher starting pay, lower satisfaction with the job taken | The after-stamp cost: better on paper, worse as felt |
| Gollwitzer & Sheeran, 2006 | Meta-analysis, 94 tests of implementation intentions | Medium-to-large average lift in goal attainment | The “if I want to unpick this, write the new fact” line |
| Scheibehenne, Greifeneder & Todd, 2010; Chernev, Böckenholt & Goodman, 2015 | Two meta-analyses of choice overload | Mean effect near zero (2010); overload under hard / complex / unclear conditions (2015) | The counter-case: more options are not a universal freeze, and second-guessing does not need a long list |
Frequently asked questions
How do you stop second-guessing yourself?
How to stop second-guessing yourself is to treat the replay as a different job from the original choice. Write whether a new fact arrived. If it did not, the file stays closed until the review date you already set. Iyengar, Wells and Schwartz (2006) showed that a better-on-paper pick can still feel worse if the scoring stays open. Gilbert, Pinel, Wilson, Blumberg and Wheatley (1998) showed that the ache you are forecasting usually runs long. A feeling of doubt is a prompt to look for a fact, not a reason to un-stamp the call.
Is second-guessing the same as analysis paralysis?
No. Analysis paralysis is the freeze before any option is stamped — more tabs, more comparison, no pick. Second-guessing starts after the pick. The parent page on analysis paralysis is the before-stamp system. This page is the after-stamp loop: the email is sent, the tool is bought, the offer is out, and the discarded option is still on the desk. If the hand has not come down yet, you are not second-guessing. You are still in the comparison.
Why do I second-guess every decision I make?
The common stack is a leftover option in view, a long regret forecast, and a maximising scoreboard with no written seal. Schwartz and colleagues (2002) tied maximising to more regret. Gilbert’s group (1998) tied the forecast to a durability bias. Brehm (1956) showed that a finished choice usually spreads the alternatives — unless both cards stay visible. “Every decision” is also how people describe a confidence story. This page only covers the work loop after a dated call. Persistent doubt that does not attach to a specific pick is outside it.
Is second-guessing a sign of low confidence?
Not as a finding. The papers above measure regret, maximising, and forecasting error, not a confidence trait you can read off a replay. A maximiser can look confident in the search and still feel worse after the stamp (Iyengar, Wells & Schwartz, 2006). A satisficer can look unsure and still close the file. Treat the loop as a procedure — leftover option, forecast, no seal — not as a verdict on character. Confidence talk is how the replay turns into a personality story.
Should I ever reopen a decision?
Yes, when a load-bearing fact arrives or when the original bar was wrong and the door is still cheap to undo. The evidence on this page is about feelings that run long and leftover options that stay in view. It is not a rule that you never reverse. Scheibehenne, Greifeneder and Todd (2010) also warn against a cartoon “too many options” law. Name the fact. If you cannot name one, hold until the review date. If you can, date the new call and close the old comparison.
What is the difference between second-guessing and reviewing a decision?
A review has a date and a question: did a new fact arrive, or did the original bar fail? Second-guessing has an open door and a feeling. Annie Duke’s Thinking in Bets frame is useful here as practice, not as a trial: judge the quality of the bet at the time you placed it. A review looks at information. A replay looks at the discarded card. Write the review date when you stamp. An undated “I’ll keep thinking” is the loop.
Does more research stop second-guessing?
After the stamp, more research is usually how the choice un-decides. Before the stamp, a short research window can be the right move — that is the analysis-paralysis page, not this one. Choice-overload findings are mixed: the 2010 meta-analysis sits near zero, and the 2015 paper finds overload only in harder, less clear sets. Extra tabs do not fix a regret forecast. They feed it. If you want a stopping rule for the next yes, Sivers’ Hell Yeah or No filter is the practitioner version.
How is second-guessing different from decision paralysis?
Decision paralysis is the freeze at commit: the shortlist is already there, and the hand will not come down. Second-guessing is the loop after the hand came down. The sister page on decision paralysis covers the stamp. This page covers the eraser. If you are still choosing between A and B, you are not second-guessing yet. If A is sent and B is still open in a tab, you are.
Sources
- Brehm, J. W. (1956). Postdecision changes in the desirability of alternatives. Journal of Abnormal and Social Psychology, 52(3), 384–389. https://doi.org/10.1037/h0041006
- Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47(2), 263–291. https://doi.org/10.2307/1914185
- Gilovich, T., & Medvec, V. H. (1995). The experience of regret: What, when, and why. Psychological Review, 102(2), 379–395. https://doi.org/10.1037/0033-295X.102.2.379
- Gilbert, D. T., Pinel, E. C., Wilson, T. D., Blumberg, S. J., & Wheatley, T. P. (1998). Immune neglect: A source of durability bias in affective forecasting. Journal of Personality and Social Psychology, 75(3), 617–638. https://doi.org/10.1037/0022-3514.75.3.617
- Zeelenberg, M. (1999). Anticipated regret, expected feedback and behavioral decision making. Journal of Behavioral Decision Making, 12(2), 93–106. https://doi.org/10.1002/(SICI)1099-0771(199906)12:2<93::AID-BDM322>3.0.CO;2-S
- Schwartz, B., Ward, A., Monterosso, J., Lyubomirsky, S., White, K., & Lehman, D. R. (2002). Maximizing versus satisficing: Happiness is a matter of choice. Journal of Personality and Social Psychology, 83(5), 1178–1197. https://doi.org/10.1037/0022-3514.83.5.1178
- Anderson, C. J. (2003). The psychology of doing nothing: Forms of decision avoidance result from reason and emotion. Psychological Bulletin, 129(1), 139–167. https://doi.org/10.1037/0033-2909.129.1.139
- Roese, N. J., & Summerville, A. (2005). What we regret most… and why. Personality and Social Psychology Bulletin, 31(9), 1273–1285. https://doi.org/10.1177/0146167205274693
- Wilson, T. D., & Gilbert, D. T. (2005). Affective forecasting: Knowing what to want. Current Directions in Psychological Science, 14(3), 131–134. https://doi.org/10.1111/j.0963-7214.2005.00355.x
- Iyengar, S. S., Wells, R. E., & Schwartz, B. (2006). Doing better but feeling worse: The paradox of choice and the job search. Psychological Science, 17(2), 143–150. https://doi.org/10.1111/j.1467-9280.2006.01677.x
- Gollwitzer, P. M., & Sheeran, P. (2006). Implementation intentions and goal achievement: A meta-analysis of effects and processes. Advances in Experimental Social Psychology, 38, 69–119. https://doi.org/10.1016/S0065-2601(06)38002-1
- Scheibehenne, B., Greifeneder, R., & Todd, P. M. (2010). Can there ever be too many options? A meta-analytic review of choice overload. Journal of Consumer Research, 37(3), 409–425. https://doi.org/10.1086/651235
- Chernev, A., Böckenholt, U., & Goodman, J. (2015). Choice overload: A conceptual review and meta-analysis. Journal of Consumer Psychology, 25(2), 333–358. https://doi.org/10.1016/j.jcps.2014.08.002
Related reading on The Growth Reads
The books behind this
- Hell Yeah or No summary & review — Sivers’ filter for a yes that leaves less leftover comparison
- The Paradox of Choice summary & review — Schwartz’s popular case for maximising and regret
- Thinking in Bets summary & review — Duke’s decision-quality frame at the time you placed the bet
- Thinking, Fast and Slow summary & review — Kahneman’s loss frame under a discarded “maybe better”
Related articles
- Analysis paralysis — the Decisions parent this article links up to
- Decision paralysis — the freeze-at-stamp sister in the same cluster
- Task paralysis — the freeze before starting a task you have already chosen
- How to prioritise tasks — choosing the next action, not re-ranking the pile
- Implementation intentions — the if-then form for the reopen moment
Go deeper
How this article was researched
The Growth Reads editorial team writes from primary sources: the books themselves and the peer-reviewed literature behind the claims. The post-decision spreading, prospect-theory, regret, affective-forecasting, maximising, job-search, implementation-intention, decision-avoidance, and choice-overload findings above were checked against the original papers or their published abstracts and are quoted with study design and, where the source supports it, the direction and size of the effect. Sivers’ Hell Yeah filter and Duke’s bet-quality frame are labelled as practitioner syntheses, not as trials that one named journal causes life outcomes.
This article is a work-and-habits page. It does not diagnose, treat, or list clinical symptoms, and nothing above is intended as medical or mental-health advice.
How we work: every article on The Growth Reads is built from the books in our library plus the research they rest on, and is reviewed and date-stamped when the evidence changes. Read our full methodology.