“Getting this offer will make me happy for months.” “That meeting will ruin the whole day.” Both statements predict a feeling, but they also predict how far that feeling will reach into the future.
Psychologists call predictions about our future emotions affective forecasting. Research does not support the simple verdict that we are hopeless at it. Accuracy depends on the question: how strong a feeling will be, how often it will return, how long it will last, or whether one day will feel better than another.
One prediction can contain three different questions
Imagine a celebration you are looking forward to. “I will be happy” might mean:
- Intensity: How strongly will I feel happy when I think about the occasion?
- Frequency: How often will that happiness come to mind afterwards?
- Duration: How much of that day will I spend in a happy mood?
These are related, but a strong feeling need not occupy every hour of the next week.
In Carlson and colleagues’ 2023 research, participants reported using emotional forecasts in decisions about careers, education, voting and travel. Two of the four studies also compared forecasts with later feelings.
One followed 178 graduating medical students through the residency-matching process. Another involved 309 student voters around the 2020 US presidential election. In both, forecasts of emotional intensity had smaller absolute errors than forecasts of frequency or duration. In the medical-student study, average forecasts overestimated the frequency and duration of happiness, while intensity was not significantly overestimated.
The finding is narrower than “we know how happy something will make us.” These were particular events and samples. A small average error can also hide individuals whose predictions miss in opposite directions. Intensity is one part of emotional experience, not a substitute for the rest.
Being right about the direction is different from getting the level right
Consider a fictional example. Alex usually rates their positive feelings around 40 on a 100-point scale. They predict 80 for a celebration, then report 60 afterwards.
The exact level was overestimated by 20 points. Yet the prediction that this would be a more positive day than usual was right. These invented numbers illustrate two meanings of accuracy; they are not a recommended personal scoring system or study results.
Absolute accuracy concerns how closely the predicted and experienced levels match. Relative accuracy concerns whether higher or lower predictions correspond to higher or lower later experiences. Across repeated observations, someone can show useful relative accuracy while consistently overshooting the level.
That distinction helps explain apparently conflicting headlines. “People overestimated their feelings” and “people predicted their feelings reasonably well” can both describe the same dataset, if the claims concern different measures.
What researchers found in everyday life
Much forecasting research concerns consequential events. Moeck and colleagues’ 2026 study examined more ordinary predictions using two existing datasets.
In the first, 209 adults forecast their feelings for the following day during a seven-day study. They also rated feelings at nine prompts each day and in evening summaries. In the second, 69 adults nominated an upcoming unpleasant event each day over two weeks, predicted their emotional response and later rated how they had felt.
Daily and event forecasts showed relative accuracy for positive and negative feelings. When participants expected feelings to be higher than their usual level, later reports tended to be higher too. Some absolute errors remained, generally small in these analyses. Their direction depended on the forecast target and emotional measure.
The comparison mattered: a summary of how a day felt is different from an average of feelings reported during that day. Forecasts sometimes matched the retrospective summary more closely. Neither measure should silently stand in for the other.
These were observational analyses, not experiments teaching people to forecast better. The unpleasant-event sample was small, and its forecast and later-report wording did not match exactly. The findings qualify a sweeping pessimistic account of forecasting; they do not establish perfect self-knowledge.
Why the imagined future can become too narrow
When imagining a future event, it is easy to give it the whole stage. The celebration becomes the coming month; the difficult meeting becomes the whole working day. Other activities, changing attention and opportunities to adapt are harder to picture.
A systematic review by Taş and colleagues, published in the 2026 volume of Annual Review of Psychology, discusses these explanations alongside the influence of current feelings on future judgments. “Impact bias” refers to overestimating aspects of an event’s emotional impact. It is a pattern found under some conditions, not a rule that every prediction is exaggerated.
The review covers several biases in decisions across time. Its searches began in 2023–24, so its publication year does not mean it incorporates every later study. Proposed ways to reduce errors also vary in how well they have been tested.
A fuller picture of the celebration
Return to Alex. Imagining the moment of celebration may give a reasonable sense of its excitement. It supplies less information about the following Tuesday: travelling home, doing ordinary tasks, hearing from friends or dealing with an unrelated problem.
An everyday decision can therefore benefit from a more specific question: “Which feeling, at what time, and in response to what?” That is an editorial suggestion for making a prediction clearer. The studies above do not prove that asking it improves mental health or guarantees a better choice.
For a reversible leisure decision, Alex could consider both the occasion and its surrounding conditions: whether the plan is affordable, accessible and compatible with rest. A vivid emotional forecast is one piece of information. Practical constraints and personal values still matter. Our article on choice overload examines a different obstacle: comparing options when preferences are unclear.
This reasoning should never become “you will get over it” or “your dread is irrational.” The studies cannot set a deadline for grief, dismiss persistent distress or establish whether a workplace is safe. If the problem is recurring anticipation before work, our Sunday scaries article discusses uncertainty alongside real conditions that may need to change.
Future feelings are worth considering. They become more useful information when we specify what we are predicting and allow room for uncertainty.
About this article
Published by PsychPod. Evidence checked 24 October 2026. Prepared with AI-assisted writing and editorial source checking. Educational psychology, not an individual assessment or treatment guide. Alex and the celebration are fictional illustrations. No independent clinical review is claimed.