Sioul: well-being gain as a budget to fill¶

Research report, started 2026-10-05 12:06. Builds on capacity-budget.md (cited as "capacity report §N") and time-estimation.md.

Questions (verbatim from the user):

1. Well-being score of tasks: can it be seen as a budget to fill?
2. What composes it?
3. How can we mix that with costs to plan for a balanced day?

Answers so far¶

(Updated 12:28, final. Entry numbers point to sections below; "§" = capacity report; "G" = design criteria in §5 of this file.)

1. Can well-being gain be a budget to fill? Yes — as a floor, kept separate from the costs; not as a currency that pays for costs, and not as a score to maximise.

• For a floor: scheduling chosen, valued or pleasant activity works as a repeated dose — behavioural activation trials, where activity scheduling alone matched full cognitive therapy (§6.2–6.3, 1.12), positive-activity meta-analyses (1.1), and a 649-patient RCT of a self-guided BA app (4.2). Detachment works better at higher dose (§6.6). On heavy days recovery is most needed and least likely to happen unaided (recovery paradox, 1.4), so the planner has to reserve it.
• Against a currency: good and bad events act on separate axes — good events raise positive mood but do not lower negative mood, while bad ones do both (3.1); positive and negative events add up independently on stress and health (3.10); losses weigh more than equal gains (1.5, 3.1); resources buffer only some demands (1.6). Gains soften the reaction to stress and speed recovery (3.2, §6.8) but no source shows they create same-day capacity (§6.12). Structured's "recharging" offset needed two arbitrary caps (4.1).
• Limits: free time helps up to about 2 h/day and declines beyond about 5 h when it feels unproductive (1.7); the same pleasure fades with repetition (1.3); measuring enjoyment lowers it (§7.6) and seeing leisure as waste lowers enjoyment (1.9), so the floor must not become a quota; meaning, connection and nature pay off over weeks, not days (2.1, 2.3, 2.9); in energy-limiting illness and burnout gain items also cost, sometimes hours later (§2.6, 1.11).

2. What composes it? Two families: restorative/hedonic (relaxation, switching off, pleasure) and eudaimonic (mastery, meaning), plus autonomy and connection; movement and nature are common sources. The best-supported vocabulary is DRAMMA — detachment, relaxation, autonomy, mastery, meaning, affiliation; within a person, relaxation, detachment, autonomy and mastery move well-being most from one period to the next, meaning and affiliation less (2.3). Pleasure and purpose diverge per activity (work and childcare: low pleasure, high purpose; 2.2), but at the level of a day or a person, multi-domain scales collapse onto one factor (PERMA vs SWB, latent r = .98; 2.5). How an activity is done (chosen, with whom, absorbed or ruminating) matters as much as what it is (2.2, 2.4). Satisfaction with leisure, not its quantity, carries the effect (2.3). So: one 0–10 gain rating per item, rated after, plus an optional kind tag (G1–G2); sleep, meals and medication stay needs, outside the floor (G4).

3. Mixing with costs for a balanced day. Constraints, not a net score: each cost scale under its budget, gain over its floor (G12). Place a low-effort gain of ≥15 min right after each heavy block; spread gains through the day because their effects are local (a 15-min lunch walk: d = 0.22–0.58, same afternoon; 3.7); keep the last hour or two low-arousal (evening detachment predicts next-morning freshness; 3.9); end the day light (G15). Protect gains more on heavy days (G16). Pairing something pleasant with a costly task raises follow-through and that task's own gain, but gains are never earned by finishing tasks (3.5, 1.9; G17). The planner proposes slots, the person fills them; an empty block counts as gain (G18). Learn the floor from days whose outcome was good — the evening "mix felt right" tap and the next-morning "rested" tap — and never learn it downward from a bad stretch (G7–G10). Suggest gains from the person's own well-rated items: advice built from a person's own ratings raised daily pleasure, modestly, in anhedonic young adults (4.6; G19). An optional U-index check (share of hours dominated by cost) can complement the totals (3.3, G14).

Main uncertainties: duration weighting and the per-item cap, the floor estimator (25th percentile is a guess), daily vs weekly floors, how rarely to ask for ratings, and whether a non-gamified app keeps people engaged (open questions 1–6). No operations-research model of a personal day with well-being terms was found (3.11).

1. Can gain be a budget to fill?¶

Already covered, not repeated here: effort-recovery model (capacity report §4.6 Meijman & Mulder), behavioural activation trials and meta-analyses (capacity report §6.1–6.3), Beck's mastery and pleasure 0–10 ratings (§6.4), Sherdell 2012 anticipatory pleasure (§6.5), recovery experiences and detachment (§6.6, §4.8–4.9), Sonnentag 2001 and ten Brummelhuis & Bakker 2012 (§6.7), Fredrickson's undoing effect (§6.8), autonomy at breaks (§6.9), restorative environments (§6.10), allostatic load (§6.11), Etkin 2016 (§7.6), NICE "all activity costs energy" (§2.6). Strength scale as in the capacity report.

1.1 Positive activity interventions (PAIs): dose and duration matter. Sin NL, Lyubomirsky S. 2009. Enhancing well-being and alleviating depressive symptoms with positive psychology interventions: a practice-friendly meta-analysis. J Clin Psychol 65(5):467–487. doi:10.1002/jclp.20593 — verified (Crossref). Bolier L, Haverman M, Westerhof GJ, et al. 2013. Positive psychology interventions: a meta-analysis of randomized controlled studies. BMC Public Health 13:119. doi:10.1186/1471-2458-13-119 — verified (Crossref).

• Sin & Lyubomirsky: 51 interventions, N = 4266: well-being up (mean r = .29), depressive symptoms down (r = .31); longer interventions worked better. Bolier: 39 RCTs, N = 6139: small effects (SMD 0.34 subjective well-being, 0.20 psychological well-being, 0.23 depression), still small but significant at 3–6 months; high heterogeneity.
• Meta-analyses. Strong for "deliberate pleasant/meaningful activity raises well-being"; effects small; Bolier, restricted to randomised trials, found smaller effects than Sin & Lyubomirsky.
• Relevance: a recurring daily/weekly practice of chosen positive activities works; it is a dose that must be repeated, not a one-off.

1.2 Lyubomirsky S, Layous K. 2013. How do simple positive activities increase well-being? Curr Dir Psychol Sci 22(1):57–62. doi:10.1177/0963721412469809 — verified (Crossref). Lyubomirsky S, Sheldon KM, Schkade D. 2005. Pursuing happiness: the architecture of sustainable change. Rev Gen Psychol 9(2):111–131. doi:10.1037/1089-2680.9.2.111 — verified (Crossref).

• Positive-activity model: the effect depends on the activity's dosage and variety, on the person's motivation and effort, and on person–activity fit; it works through positive emotions, positive thoughts, positive behaviours and need satisfaction. The 2005 paper argues that intentional activities, not circumstances, are the lever that escapes hedonic adaptation, and reports two small pilot interventions in which timing/frequency changed the effect (students doing five acts of kindness a week for six weeks became happier only when they did all five on one day rather than spreading them — confirmed via the Greater Good Science Center summary, https://ggia.berkeley.edu/practice/random_acts_of_kindness; a parallel gratitude finding, once a week beating three times a week, is commonly reported but was not confirmed today).
• Review/theory with small experiments. Moderate for the model, weak for the specific dose findings.
• Relevance: "more" is not automatically better; fit and variety matter more than volume. A floor should be a small chosen dose, varied.

1.3 Sheldon KM, Lyubomirsky S. 2012. The challenge of staying happier: testing the Hedonic Adaptation Prevention model. Pers Soc Psychol Bull 38(5):670–680. doi:10.1177/0146167212436400 — verified (Crossref).

• 481 students, 3 waves over 3 months: gains from a positive change erode through fading positive emotions and rising aspirations; continued appreciation and variety slow the erosion.
• Longitudinal study. Moderate.
• Relevance: the same pleasant activity repeated daily yields less over time; a planner proposing "your usual gain items" should rotate them and let the person vary them; the learned gain value of an item may drift down with repetition, which is adaptation, not a fault of the person.

1.4 Sonnentag S. 2018. The recovery paradox: portraying the complex interplay between job stressors, lack of recovery, and poor well-being. Res Organ Behav 38:169–185. doi:10.1016/j.riob.2018.11.002 — verified (Crossref).

• Review: detachment, exercise and sleep protect well-being, but they are impaired exactly when stressors are high (high-stress days are followed by worse detachment, less exercise, worse sleep).
• Narrative review of diary research. Moderate.
• Relevance: on heavy days the gain floor is both most needed and least likely to happen spontaneously; the planner should protect gain slots more, not less, on heavy days, and keep them low-effort (a high-effort gain will be dropped).

1.5 Conservation of resources (COR) theory. Hobfoll SE. 1989. Conservation of resources: a new attempt at conceptualizing stress. Am Psychol 44(3):513–524. doi:10.1037/0003-066X.44.3.513 — verified (Crossref). Hobfoll SE, Halbesleben J, Neveu J-P, Westman M. 2018. Conservation of resources in the organizational context: the reality of resources and their consequences. Annu Rev Organ Psychol Organ Behav 5:103–128. doi:10.1146/annurev-orgpsych-032117-104640 — verified (Crossref).

• Principles as stated in the 2018 review: resource loss weighs disproportionately more than equal gain (primacy of loss); people must invest resources to gain or protect resources; gains become more important when losses are high (gain paradox); losses and gains form spirals, loss spirals faster.
• Theory with a large supporting literature, mostly correlational. Moderate as theory.
• Relevance: a point of gain does not cancel a point of cost (loss weighs more), so a net "gain − cost" score would understate bad days. Supports a separate floor. "Gain spirals" give a reason to protect gains during recovery periods, but no planning rule.

1.6 Job demands–resources (JD-R) model: buffering is an interaction, not a subtraction. Bakker AB, Demerouti E. 2007. The Job Demands-Resources model: state of the art. J Manag Psychol 22(3):309–328. doi:10.1108/02683940710733115 — verified (Crossref). Bakker AB, Demerouti E, Euwema MC. 2005. Job resources buffer the impact of job demands on burnout. J Occup Health Psychol 10(2):170–180. doi:10.1037/1076-8998.10.2.170 — verified (Crossref).

• Demands drive exhaustion (health-impairment path); resources (autonomy, support, feedback) drive engagement (motivational path) and soften the effect of demands. In 1012 employees, the buffering interaction held for exhaustion and cynicism in only 18 of 32 demand × resource combinations, and not for professional efficacy.
• Cross-sectional survey; model review. Moderate for two separate paths; weak-to-moderate for buffering (partial, specific pairs).
• Relevance: gains and costs act on different paths (two outcomes, not one balance); resources reduce the harm of some demands but do not remove them. Matches capacity report §6.12: keep two quantities.

1.7 Discretionary time: an inverted U. Sharif MA, Mogilner C, Hershfield HE. 2021. Having too little or too much time is linked to lower subjective well-being. J Pers Soc Psychol 121(4):933–947. doi:10.1037/pspp0000391 — verified (Crossref).

• Two large data sets (35,375 Americans) and two experiments: well-being rises with discretionary time, then flattens and, at high amounts, falls, because of a lacking sense of productivity; the drop disappears when the free time is spent on productive activities (the authors' school summary, titled "blame solitude or lack of productive activity", indicates social use of the time helps too). Thresholds from the paper as summarised by the authors' school (https://anderson-review.ucla.edu/too-much-free-time-blame-solitude-or-lack-of-productive-activity): below about 2 h/day of free time people feel stressed; well-being does not rise beyond about 2 h and starts to fall beyond about 5 h.
• Large cross-sectional surveys + experiments. Moderate-to-strong for the shape; thresholds are population averages.
• Relevance: there is a "too much" for idle free time, but not for meaningful or social gain. The gain floor is a floor, not a target to maximise; the ceiling problem applies to unstructured time, which a planner rarely creates.

1.8 Time affluence. Whillans AV, Dunn EW, Smeets P, Bekkers R, Norton MI. 2017. Buying time promotes happiness. PNAS 114(32):8523–8527. doi:10.1073/pnas.1706541114 — verified (Crossref). Kasser T, Sheldon KM. 2009. Time affluence as a path toward personal happiness and ethical business practice: empirical evidence from four studies. J Bus Ethics 84(Suppl 2):243–255. doi:10.1007/s10551-008-9696-1 — verified (Crossref; no abstract read).

• N = 6271 across four countries: people who spend money on time-saving services report higher life satisfaction; a field experiment found working adults happier after a time-saving purchase than after a material one. Kasser & Sheldon: feeling one has enough time relates to well-being across four studies.
• Cross-sectional + one small field experiment. Moderate.
• Relevance: the sense of unhurried time is itself a gain; slack in the plan (capacity criterion 27) contributes to well-being, not only to safety. Removing chores (delegation, batching) counts as gain.

1.9 Tonietto GN, Malkoc SA, Reczek RW, Norton MI. 2021. Viewing leisure as wasteful undermines enjoyment. J Exp Soc Psychol 97:104198. doi:10.1016/j.jesp.2021.104198 — verified (Crossref; no abstract indexed; findings from university press summaries, e.g. https://www.uni-mannheim.de/forschung-erleben/artikel/spass-an-der-freude/ ; preregistration https://aspredicted.org/jk6pm.pdf).

• Over 300 people: those who believed leisure is wasteful enjoyed their Halloween activities less. In an experiment, reading an article calling free time wasteful lowered enjoyment of funny videos watched afterwards. The authors suggest that for people holding this belief, seeing how a leisure activity serves their longer-term goals restores enjoyment.
• Several studies, one experimental manipulation, preregistered. Moderate.
• Relevance: an app that frames gain as productive "recharging so you can work more" may help people who see leisure as waste, but framing leisure as a duty or a score to reach (a quota) risks the same loss of enjoyment as Etkin's measurement effect (§7.6). Wording matters: "time that is yours", not "recharge target".

1.10 Anticipated vs experienced pleasure in depression. Wu H, Mata J, Furman DJ, Whitmer AJ, Gotlib IH, Thompson RJ. 2017. Anticipatory and consummatory pleasure and displeasure in major depressive disorder: an experience sampling study. J Abnorm Psychol 126(2):149–159. doi:10.1037/abn0000244 — verified (Crossref).

• 41 adults with MDD, 39 controls, one week, 8 prompts/day: depressed people had blunted anticipated AND experienced pleasure and raised anticipated and experienced displeasure. In both groups people predicted pleasure accurately but overestimated displeasure.
• Experience sampling, small. Moderate.
• Relevance and conflict with capacity report §6.5 (Sherdell 2012, lab): in daily life the forecast of pleasure was not too low relative to the experience; what was wrong was the forecast of displeasure (too high). So the bias to correct is on the cost/anxiety side (capacity criterion 7 already does this), and the gain side is "low both before and after" in depression. A low gain level is a symptom to plan around, not a forecasting error to correct upward.

1.11 Pleasant now, tired later. Leikas S, Ilmarinen V-J. 2017. Happy now, tired later? Extraverted and conscientious behavior are related to immediate mood gains, but to later fatigue. J Pers 85(5):603–615. doi:10.1111/jopy.12264 — verified (Crossref).

• 48 people, 12 days, 5 prompts/day (2328 observations): sociable and conscientious behaviour came with better mood and less fatigue at the time, but more fatigue three hours later; the social effect ran through the number of people met.
• Experience sampling, small. Weak-to-moderate (one study).
• Relevance: the same item gives a gain now and a cost later. A social gain must carry its own cost rating, and the cost should be attributed with a delay (hours), as capacity criterion 20 does for delayed crashes.

1.12 Scheduling activities alone is the active part. Jacobson NS, Dobson KS, Truax PA, Addis ME, Koerner K, Gollan JK, Gortner E, Prince SE. 1996. A component analysis of cognitive-behavioral treatment for depression. J Consult Clin Psychol 64(2):295–304. doi:10.1037/0022-006X.64.2.295 — verified (Crossref).

• 150 outpatients with major depression randomised to the behavioural activation component alone, BA plus work on automatic thoughts, or full cognitive therapy: the full treatment did no better than either component, at the end of treatment or 6 months later.
• RCT. Moderate-to-strong (the trial that launched modern BA; replicated in spirit by §6.2).
• Relevance: planning and doing chosen activities is enough to carry the effect; no cognitive layer is needed in the app for the gain floor to be worth having.

2. What composes it?¶

Already covered: mastery and pleasure as Beck's two ratings (capacity report §6.4), recovery experiences detachment / relaxation / mastery / control (§4.8, §6.6), nature and attention restoration (§6.10), autonomy (§6.9).

2.1 Hedonic and eudaimonic gains differ in time course. Ryan RM, Deci EL. 2001. On happiness and human potentials: a review of research on hedonic and eudaimonic well-being. Annu Rev Psychol 52:141–166. doi:10.1146/annurev.psych.52.1.141 — verified (Crossref). Huta V, Ryan RM. 2010. Pursuing pleasure or virtue: the differential and overlapping well-being benefits of hedonic and eudaimonic motives. J Happiness Stud 11(6):735–762. doi:10.1007/s10902-009-9171-4 — verified (Crossref; full text read).

• Two traditions: pleasure and absence of pain (hedonic) vs meaning, growth, functioning (eudaimonic). Huta & Ryan (four studies: two correlational, one experience-sampling, one intervention): in the intervention, pursuing pleasure gave more well-being benefits at short-term follow-up, pursuing meaning/excellence gave more at a 3-month follow-up (full text read: https://selfdeterminationtheory.org/SDT/documents/2010_HutaRyan_JOHS.pdf). Eudaimonia, not hedonia, raised meaning right after the intervention and elevating experiences and vitality at 3 months, and lowered negative affect, but did not raise life satisfaction; people high in both hedonia and eudaimonia had the greatest well-being.
• Review (strong as a map of the field); multi-study paper (moderate).
• Relevance: a single "gain" number mixes two things that pay off on different time scales. Rest/pleasure pays the same evening; meaning/mastery pays over weeks. A daily floor fits the hedonic/restorative part better; the meaning part fits a weekly view.

2.2 What everyday activities feel like: DRM and experience sampling. Kahneman D, Krueger AB, Schkade DA, Schwarz N, Stone AA. 2004. A survey method for characterizing daily life experience: the day reconstruction method. Science 306(5702):1776–1780. doi:10.1126/science.1103572 — verified (Crossref; full text read). White MP, Dolan P. 2009. Accounting for the richness of daily activities. Psychol Sci 20(8):1000–1008. doi:10.1111/j.1467-9280.2009.02392.x — verified (Crossref). Bryson A, MacKerron G. 2017. Are you happy while you work? Econ J 127(599):106–125. doi:10.1111/ecoj.12269 — verified (Crossref). Killingsworth MA, Gilbert DT. 2010. A wandering mind is an unhappy mind. Science 330(6006):932. doi:10.1126/science.1192439 — verified (Crossref).

• DRM, 909 employed women reconstructing the previous day (full text read, Table 1): mean positive affect ranked intimate relations 5.10, socialising 4.59, relaxing 4.42, praying/meditating 4.35, eating 4.34, exercising 4.31, watching TV 4.19 … taking care of children 3.86, computer/e-mail 3.81, housework 3.73, working 3.62, commuting 3.45. Feeling competent ran the other way for work (working 4.45, computer 4.57, clients 4.65 vs watching TV 3.95). Tiredness was highest during napping (4.30), watching TV (3.54) and relaxing (3.44): restful activities are chosen when tired. Friends were the best interaction partners (4.36), the boss the worst (3.52). Poor sleep and time pressure at work lowered enjoyment a lot (press release, https://www.sciencedaily.com/releases/2004/12/041203082806.htm). White & Dolan added a "reward/purpose" rating to the DRM: work and time with children are low in pleasure but rated rewarding, so they contribute to well-being when meaning is counted. Mappiness (UK smartphone sampling): paid work ranked below every other of 40 activities except being sick in bed. Killingsworth & Gilbert (iPhone sampling of 2250 adults; details confirmed via the Harvard Gazette report, https://news.harvard.edu/gazette/story/2010/11/wandering-mind-not-a-happy-mind/): minds wandered 46.9% of the time, and at least 30% of the time during every activity but one; people were less happy while mind-wandering, and mind-wandering predicted happiness better than the activity did.
• Large DRM/ESM studies. Strong for broad rankings at population level; weak for any single person.
• Relevance: (a) category defaults for gain can come from these rankings (social, rest, meals, movement, outdoors high; commuting, admin, chores low); the competence column is Beck's mastery/pleasure split seen in population data; and because restful items happen when people are tired, a naive correlation of gain items with fatigue would make rest look harmful — learn from next-day outcomes, not same-moment ones; (b) pleasure and purpose diverge by activity (White & Dolan), so either keep two ratings or accept that one number hides this; (c) engagement in the activity matters as much as the activity itself — a gain block spent ruminating gives little (ties to detachment, capacity report §4.9).

2.3 Leisure: what it gives, and through which mechanisms. Kuykendall L, Tay L, Ng V. 2015. Leisure engagement and subjective well-being: a meta-analysis. Psychol Bull 141(2):364–403. doi:10.1037/a0038508 — verified (Crossref). Newman DB, Tay L, Diener E. 2014. Leisure and subjective well-being: a model of psychological mechanisms as mediating factors. J Happiness Stud 15(3):555–578. doi:10.1007/s10902-013-9435-x — verified (Crossref; model built from a quantitative summary of 363 articles). Kujanpää M, Syrek C, Lehr D, Kinnunen U, Reins JA, de Bloom J. 2021. Need satisfaction and optimal functioning at leisure and work: a longitudinal validation study of the DRAMMA model. J Happiness Stud 22(2):681–707. doi:10.1007/s10902-020-00247-3 — verified (Crossref).

• Kuykendall: 37 effect sizes, N = 11,834: leisure engagement and well-being correlate r = .26, through satisfaction with leisure (not its quantity); longitudinal effects run both ways (β = .15 and .16); experimental leisure interventions raise well-being (d = 1.02, few studies). Newman et al.: the DRAMMA model — leisure raises well-being through Detachment-recovery, Autonomy, Mastery, Meaning and Affiliation. Kujanpää et al. split it into six needs (detachment, relaxation, autonomy, mastery, meaning, affiliation) and followed 279 employees over five measurements around holidays: the six-factor structure held, and within a person, relaxation, detachment, autonomy and mastery tracked well-being most consistently; meaning and affiliation much less.
• Meta-analysis (strong for the association; the d = 1.02 rests on few small trials); conceptual review (moderate); longitudinal validation (moderate).
• Relevance: DRAMMA is a ready vocabulary for what a gain is made of, and the within-person data say which parts move from one period to the next: rest (relaxation, detachment), choice (autonomy) and mastery. Those suit a daily floor; meaning and connection act on a longer scale (consistent with 2.1). It overlaps with Sonnentag & Fritz (detachment, relaxation, mastery, control) and with SDT (autonomy, competence, relatedness). Satisfaction with the leisure, not hours of it, carries the effect: the person's own after-rating is the right measure.

2.4 Basic psychological needs at the day level. Reis HT, Sheldon KM, Gable SL, Roscoe J, Ryan RM. 2000. Daily well-being: the role of autonomy, competence, and relatedness. Pers Soc Psychol Bull 26(4):419–435. doi:10.1177/0146167200266002 — verified (Crossref). Ryan RM, Bernstein JH, Brown KW. 2010. Weekends, work, and well-being: psychological need satisfactions and day of the week effects on mood, vitality, and physical symptoms. J Soc Clin Psychol 29(1):95–122. doi:10.1521/jscp.2010.29.1.95 — verified (Crossref).

• Two weeks of daily reports (students): days with more autonomy, competence and relatedness were better days, controlling for the person's average; meaningful talk and feeling understood and appreciated were the best social predictors. 74 workers: the weekend and non-work lift in mood and vitality was explained partly or fully by more autonomy and relatedness.
• Daily diary studies. Moderate.
• Relevance: the gain of an item depends on how it is done (chosen? with whom? meaningful talk?), not only on its category. Autonomy is itself a gain: the person choosing what fills a gain slot is part of the gain (also capacity report §6.9).

2.5 PERMA vs a single well-being factor. Butler J, Kern ML. 2016. The PERMA-Profiler: a brief multidimensional measure of flourishing. Int J Wellbeing 6(3):1–48. doi:10.5502/ijw.v6i3.526 — verified (Crossref). Goodman FR, Disabato DJ, Kashdan TB, Kauffman SB. 2018. Measuring well-being: a comparison of subjective well-being and PERMA. J Posit Psychol 13(4):321–332. doi:10.1080/17439760.2017.1388434 — verified (Crossref).

• PERMA-Profiler: 15 items, 3 per domain (positive emotion, engagement, relationships, meaning, accomplishment), developed on N = 7188 and tested on N = 31,966. Goodman et al., 517 adults: PERMA and classic subjective well-being correlated r = .98 at the latent level; the domains converge on one well-being factor.
• Psychometric studies. Moderate-to-strong.
• Relevance: at the level of a person's overall state, many sub-scales mostly measure one thing; extra sub-scales buy little for the day's total. At the level of single activities, though, pleasure and purpose do diverge (2.2 White & Dolan). Supports: one gain number for the daily floor, plus optional kind tags, not five ratings per task.

2.6 Flow and absorbing work. Csikszentmihalyi M, LeFevre J. 1989. Optimal experience in work and leisure. J Pers Soc Psychol 56(5):815–822. doi:10.1037/0022-3514.56.5.815 — verified (PubMed 2724069).

• 78 workers, one week of experience sampling: quality of experience depended more on being in flow (high challenge, high skill) than on work vs leisure, except for relaxation and motivation; most flow episodes happened at work, not in leisure. The abstract goes on (record truncated in the copy read) to the finding usually cited as the "paradox of work": people still wished to be doing something else when working.
• ESM study. Weak-to-moderate.
• Relevance: a demanding task can carry a real gain (mastery, absorption) while still costing; Sioul must allow cost and gain on the same item. Passive leisure is restful but rarely absorbing — both are gains of different kinds.

2.7 Social contact: gain now, often cost later, and under-predicted. Sun J, Harris K, Vazire S. 2020. Is well-being associated with the quantity and quality of social interactions? J Pers Soc Psychol 119(6):1478–1496. doi:10.1037/pspp0000272 — verified (Crossref). Epley N, Schroeder J. 2014. Mistakenly seeking solitude. J Exp Psychol Gen 143(5):1980–1999. doi:10.1037/a0037323 — verified (Crossref). Leikas & Ilmarinen 2017 (1.11).

• N = 256, 3206 observations, audio recorder + 4 prompts/day for a week: more social interaction went with higher momentary well-being, observed or self-reported; deeper conversation and liked partners added to it. Commuters predicted solitude would be nicer; talking to a stranger was actually more pleasant. But sociable behaviour predicted more fatigue three hours later (1.11), and in autistic burnout social contact and masking are primary costs (capacity report §2.17).
• ESM with observer data (moderate-to-strong); field experiments (moderate); one ESM study (weak-to-moderate).
• Relevance: social items deserve a gain and a delayed cost, rated by the person. The app must never impose or push social activities (capacity report §6.9: imposed social activity at lunch raised fatigue).

2.8 Movement. Liao Y, Shonkoff ET, Dunton GF. 2015. The acute relationships between affect, physical feeling states, and physical activity in daily life: a review of current evidence. Front Psychol 6:1975. doi:10.3389/fpsyg.2015.01975 — verified (Crossref).

• 14 naturalistic studies, non-clinical: physical activity is consistently followed by higher positive affect for a few hours, with some evidence of more energy; positive affect also predicts later activity.
• Systematic review of observational ESM studies. Moderate.
• Relevance: a gain for most people within hours; but a crash trigger in ME/CFS and long COVID (capacity report §2.6, §2.8, §6.13). Default gain for movement only if the person has not marked an energy-limiting condition.

2.9 Nature: an example of a dose with a plateau, counted weekly. White MP, Alcock I, Grellier J, et al. 2019. Spending at least 120 minutes a week in nature is associated with good health and wellbeing. Sci Rep 9:7730. doi:10.1038/s41598-019-44097-3 — verified (Crossref). MacKerron G, Mourato S. 2013. Happiness is greater in natural environments. Glob Environ Change 23(5):992–1000. doi:10.1016/j.gloenvcha.2013.03.010 — verified (Crossref; no abstract indexed).

• N = 19,806 (England, nationally weighted): good health and high well-being became more likely from ≥120 min/week of recreational nature contact, peaked at 200–300 min/week with no further gain; one long visit or several short ones made no difference. Mappiness: people were happier outdoors in natural settings than in urban ones.
• Large cross-sectional survey (moderate; no causal proof); large ESM (moderate).
• Relevance: the clearest published example of a gain "floor" with a plateau, and it is weekly, not daily, and indifferent to how it is split. Suggests a weekly floor for some gain kinds, with daily placement left flexible.

2.10 Sleep. Konjarski M, Murray G, Lee VV, Jackson ML. 2018. Reciprocal relationships between daily sleep and mood: a systematic review of naturalistic prospective studies. Sleep Med Rev 42:47–58. doi:10.1016/j.smrv.2018.05.005 — verified (PubMed 30404728).

• Review of naturalistic diary/ESM studies (PubMed abstract read): subjective sleep (quality, duration, latency) and daytime affect influence each other over the short term, in both directions.
• Systematic review. Moderate.
• Relevance: sleep is the largest restorative item but Sioul already plans it as a need; it should not be counted inside the gain floor (otherwise sleep alone fills it). Same for meals and naps: needs, not gains, unless the person rates them.

2.11 Occupational balance (occupational therapy). Wagman P, Håkansson C, Björklund A. 2012. Occupational balance as used in occupational therapy: a concept analysis. Scand J Occup Ther 19(4):322–327. doi:10.3109/11038128.2011.596219 — verified (Crossref). Wagman P, Håkansson C. 2014. Introducing the Occupational Balance Questionnaire (OBQ). Scand J Occup Ther 21(3):227–231. doi:10.3109/11038128.2014.900571 — verified (Crossref). Håkansson C, Wagman P, Hagell P. 2020. Construct validity of a revised version of the Occupational Balance Questionnaire. Scand J Occup Ther 27(6):441–449. doi:10.1080/11038128.2019.1660801 — verified (Crossref). Matuska KM, Christiansen CH. 2008. A proposed model of lifestyle balance. J Occup Sci 15(1):9–19. doi:10.1080/14427591.2008.9686602 — verified (Crossref). Wagman P, Håkansson C, Jonsson H. 2015. Occupational balance: a scoping review of current research and identified knowledge gaps. J Occup Sci 22(2):160–169. doi:10.1080/14427591.2014.986512 — verified (Crossref). Eklund M, Orban K, Argentzell E, et al. 2017. The linkage between patterns of daily occupations and occupational balance. Scand J Occup Ther 24(1):41–56. doi:10.1080/11038128.2016.1224271 — verified (Crossref).

• Occupational balance is defined (Wagman et al. 2012, definition as commonly quoted; paper not re-read today) as the person's own perception of having the right amount of occupations and the right variation between them (work, home, leisure, rest, self-care, social); it is subjective, not a fixed ratio of hours. Matuska & Christiansen (abstract read): balance is the perceived match between desired and actual patterns of occupation across five need-based dimensions (as listed in the paper, not re-read today: health and basic needs, relationships, challenge/competence, meaning/identity, organising time and energy). The OBQ11 (11 items, Rasch reliability 0.92) measures it as a general state. Eklund et al. separate the observable pattern of days (hours per kind) from the felt balance; the scoping review (22 studies, 2009–2014) found the link with health plausible but the evidence base young.
• Concept analyses, questionnaire validation, scoping review. Moderate for measurement; weak for causal links to health.
• Relevance: the profession that studies "a balanced day" measures it by asking the person whether the mix felt right, not by summing ratings. Sioul's end-of-day question could be exactly that ("how was the mix today: too much duty / about right / too empty"). Variety across kinds is the field's ideal, but 2.15 shows it is not good for everyone, so Sioul should offer it, not enforce it.

2.12 Nourishing vs depleting (MBCT). Segal ZV, Williams JMG, Teasdale JD. 2013. Mindfulness-Based Cognitive Therapy for Depression, 2nd ed. New York: Guilford — book, not peer-reviewed as such (the programme itself has RCT support; not reviewed here).

• Exercise in session 7 (from the book, not re-read today): list the activities of a typical day, mark each as nourishing or depleting, then plan to add nourishing ones and change how depleting ones are approached.
• Clinical practice. Opinion.
• Relevance: prior art for marking every activity on both sides (gives / takes), and for the person, not a list, deciding which side an activity falls on.

2.13 One item or several: reliability of single ratings. Wanous JP, Reichers AE, Hudy MJ. 1997. Overall job satisfaction: how good are single-item measures? J Appl Psychol 82(2):247–252. doi:10.1037/0021-9010.82.2.247 — verified (Crossref). Cheung F, Lucas RE. 2014. Assessing the validity of single-item life satisfaction measures: results from three large samples. Qual Life Res 23(10):2809–2818. doi:10.1007/s11136-014-0726-4 — verified (PubMed 24890827). Dejonckheere E, Demeyer F, Geusens B, et al. 2022. Assessing the reliability of single-item momentary affective measurements in experience sampling. Psychol Assess 34(12):1138–1154. doi:10.1037/pas0001178 — verified (Crossref).

• Single items correlate about .63 with full scales (17 studies, N = 7682); their minimum reliability is estimated at .45–.69. Single-item life satisfaction correlated .62–.64 with the 5-item scale (.78–.80 corrected) in three samples (N = 13,064, 2277, 1312) and gave virtually identical answers to substantive questions. In experience sampling, the error of a single momentary emotion rating is about a tenth of the scale and makes up about 27% of the observed variability (two empirical studies, PubMed 36074609).
• Meta-analysis (moderate); large-sample validation (strong for its purpose); methods paper with empirical data (moderate).
• Relevance: one 0–10 gain item per activity is acceptable if Sioul averages over many items and days (noise cancels), and treats a 1–2 point difference on one item as noise (capacity criterion 8). It is not reliable enough to judge a single activity from one rating: on 0–10, ±1 is pure noise (Dejonckheere: error ≈ 1/10 of the scale).

2.14 Savoring: the gain of a good moment depends on attending to it. Jose PE, Lim BT, Bryant FB. 2012. Does savoring increase happiness? A daily diary study. J Posit Psychol 7(3):176–187. doi:10.1080/17439760.2012.671345 — verified (Crossref).

• 101 people, 30 days of experience sampling: momentary savoring both carried and amplified the effect of daily positive events on happy mood.
• ESM study. Moderate.
• Relevance: same as mind-wandering (2.2): the gain of a block depends on being present in it. A gain block squeezed between anxious tasks, or spent checking messages, gives less. Supports buffer time around gain blocks and no notifications during them.

2.15 Variety of activities: not good for everyone. Lee S, Koffer RE, Sprague BN, Charles ST, Ram N, Almeida DM. 2018 (online 2016). Activity diversity and its associations with psychological well-being across adulthood. J Gerontol B Psychol Sci Soc Sci. doi:10.1093/geronb/gbw118 — verified (Crossref; volume/pages not checked).

• N = 793, two 8-day diary bursts ten years apart, seven activity kinds (paid work, children, chores, leisure, physical activity, volunteering, informal help): more diverse days went with higher psychological well-being in adults 60–74, not in 35–59, and the opposite in 24–34.
• Large diary panel. Moderate.
• Relevance: "balance as variety across kinds" (occupational balance, 2.11) is not a universal good; a younger person with few demanding roles may do better with focus. Variety should be the person's choice, not a planner objective.

3. Mixing gains with costs for a balanced day¶

Already covered: breaks >10 min after demanding tasks (capacity report §4.11 Albulescu 2022), earlier breaks restore more (§4.11 Hunter & Wu 2016), evening detachment (§4.9, §6.7), the undoing effect (§6.8), gains as a floor vs offset (§6.12), allostatic load (§6.11).

3.1 Good and bad events act on different axes. David JP, Green PJ, Martin R, Suls J. 1997. Differential roles of neuroticism, extraversion, and event desirability for mood in daily life. J Pers Soc Psychol 73(1):149–159. doi:10.1037/0022-3514.73.1.149 — verified (Crossref). Cacioppo JT, Berntson GG. 1994. Relationship between attitudes and evaluative space: a critical review, with emphasis on the separability of positive and negative substrates. Psychol Bull 115(3):401–423. doi:10.1037/0033-2909.115.3.401 — verified (Crossref). Baumeister RF, Bratslavsky E, Finkenauer C, Vohs KD. 2001. Bad is stronger than good. Rev Gen Psychol 5(4):323–370. doi:10.1037/1089-2680.5.4.323 — verified (Crossref).

• 96 men, 8 nightly diaries: undesirable events raised negative mood and also lowered positive mood; desirable events raised positive mood but did not lower negative mood. Positive and negative evaluation are separable processes, not two ends of one scale. Across many domains bad events weigh more than good ones.
• Diary study (moderate), reviews (strong as reviews of many studies, but broad).
• Relevance: a gain does not subtract from a cost. Netting "cost − gain" into one score is not supported; keep costs (ceilings) and gains (floor) as separate quantities and check both.

3.2 Ong AD, Bergeman CS, Bisconti TL, Wallace KA. 2006. Psychological resilience, positive emotions, and successful adaptation to stress in later life. J Pers Soc Psychol 91(4):730–749. doi:10.1037/0022-3514.91.4.730 — verified (Crossref).

• Three daily-diary studies (older adults, recently widowed women): on days with positive emotions, stress raised negative emotions less, and recovery from stress was faster, especially in resilient people.
• Multilevel diary studies. Moderate.
• Relevance: gains buffer the reaction to costs on the same day (an interaction) and speed recovery the next days; consistent with §6.8. Not a licence to schedule more cost.

3.3 The U-index: time spent in an unpleasant state. Kahneman D, Krueger AB. 2006. Developments in the measurement of subjective well-being. J Econ Perspect 20(1):3–24. doi:10.1257/089533006776526030 — verified (Crossref).

• Proposes the U-index: the share of time in which the strongest feeling is negative; it needs no cardinal (summable) scale of feelings. Feelings reported close to the experience are more accurate.
• Methods paper by the DRM authors. Moderate (widely used).
• Relevance: a "balanced day" can be judged by how much of it is spent in blocks whose cost dominates their gain, rather than by summing points. A planner check could be: share of the day's hours where max(cost ratings) > gain rating, kept under a learned level. This avoids adding up 0–10 numbers that are not on a ratio scale.

3.4 Order within a day: people prefer improving sequences and spread good things. Loewenstein G, Prelec D. 1993. Preferences for sequences of outcomes. Psychol Rev 100(1):91–108. doi:10.1037/0033-295X.100.1.91 — verified (Crossref).

• When choices are framed as sequences, people prefer outcomes that improve over time and want good outcomes spread evenly, violating additive models.
• Experiments on stated preference. Moderate.
• Relevance: do not lump all gains at the end of the day (or all in the morning); spread them, and let the day end on something light or pleasant (also peak-end, capacity report §1.8).

3.5 Pairing a gain with a cost (temptation bundling, immediate rewards). Milkman KL, Minson JA, Volpp KGM. 2014. Holding the Hunger Games hostage at the gym: an evaluation of temptation bundling. Manage Sci 60(2):283–299. doi:10.1287/mnsc.2013.1784 — verified (Crossref). Kirgios EL, Mandel GH, Park Y, Milkman KL, Gromet DM, Kay JS, Duckworth AL. 2020. Teaching temptation bundling to boost exercise: a field experiment. Organ Behav Hum Decis Process 161:20–35. doi:10.1016/j.obhdp.2020.09.003 — verified (Crossref). Woolley K, Fishbach A. 2016. For the fun of it: harnessing immediate rewards to increase persistence in long-term goals. J Consum Res 42(6):952–966. doi:10.1093/jcr/ucv098 — verified (Crossref). Premack D. 1959. Toward empirical behavior laws: I. Positive reinforcement. Psychol Rev 66(4):219–233. doi:10.1037/h0040891 — verified (Crossref; classic, not re-read).

• Milkman (university gym members, field experiment): audiobooks available only at the gym raised gym visits 51% at first (29% with mere encouragement), fading over time; 61% then paid for the restriction. Kirgios: N = 6792 gym members, audiobook + encouragement raised weekly workout likelihood 10–14%, lasting up to 17 weeks after; the encouragement added little beyond the free audiobook. Woolley & Fishbach: choosing the more enjoyable version of a goal activity, or adding music, increased persistence more than focusing on delayed benefits. Premack: a more-preferred activity can reinforce a less-preferred one that precedes it.
• Field RCTs (moderate-to-strong for small effects on exercise), lab/field experiments (moderate), classic theory.
• Relevance: two legitimate ways to mix: (a) put a gain inside a cost (music, company, pleasant place: raises the item's own gain rating), (b) put a chosen gain right after a cost. Both raise follow-through. Neither shows a well-being benefit, only adherence; and making leisure a "reward to earn" risks Tonietto's effect (1.9) and the "leisure must be earned" frame. Sioul may offer pairing; it should not lock gains behind tasks.

3.6 Who decides when the gain happens: scheduled vs prompted. Howe E, Suh J, Bin Morshed M, McDuff D, Rowan K, et al. 2022. Design of digital workplace stress-reduction intervention systems: effects of intervention type and timing. Proc CHI 2022. doi:10.1145/3491102.3502027 — verified (Crossref).

• 86 workers, 4 weeks: system-prompted ("just-in-time", at detected stress) interventions were done more often than self-scheduled ones, but stress reduction did not differ. People rated low-effort interventions highest, yet high-effort ones reduced stress most. Participants liked the accountability of prompts but wanted partial control over timing.
• Field experiment, conference paper. Weak-to-moderate.
• Relevance: let the planner propose the slot, let the person move or decline it. Offer both low-effort gains (likely to happen) and higher-effort ones (more effect), without pushing.

3.7 Lunch-break gains have small, short effects. de Bloom J, Sianoja M, Korpela K, Tuomisto M, Lilja A, Geurts S, Kinnunen U. 2017. Effects of park walks and relaxation exercises during lunch breaks on recovery from job stress: two randomized controlled trials. J Environ Psychol 51:14–30. doi:10.1016/j.jenvp.2017.03.006 — verified (Crossref).

• 153 knowledge workers, 15 min/day for 10 working days: in autumn, restoration rose and fatigue fell after lunch and in the afternoon (d = 0.22–0.58), most consistently with park walks; in spring, hardly any effect.
• Two small RCTs. Moderate for "small, short-lived, context-dependent".
• Relevance: one 15-minute mid-day gain block has a real but modest same-day effect; gains are local in time, so placement matters more than the daily total. Supports placing gains between demanding blocks rather than banking them for the evening.

3.8 Protected time on the calendar. Saha K, Iqbal ST. 2023. Focus time for wellbeing and work engagement of information workers. CHI EA 2023. doi:10.1145/3544549.3585688 — verified (Crossref).

• 15 treatment vs 10 control, 6 weeks: automatically blocking focus time in Outlook went with higher well-being (more relaxation and satisfaction, less tiredness and stress).
• Small experiment, extended abstract, authors from the vendor. Weak.
• Relevance: reserving time on the calendar can itself be a gain (time affluence, 1.8; autonomy, 2.4).

3.9 Which gain shows up in which next-morning outcome. Sonnentag S, Binnewies C, Mojza EJ. 2008. "Did you have a nice evening?" A day-level study on recovery experiences, sleep, and affect. J Appl Psychol 93(3):674–684. doi:10.1037/0021-9010.93.3.674 — verified (Crossref).

• 166 employees, one work week of daily surveys: low detachment in the evening predicted next-morning negative activation and fatigue; evening mastery predicted next-morning positive activation; relaxation predicted serenity; sleep quality related to all of them.
• Diary study. Moderate.
• Relevance: the outcome to learn the gain floor from is next-morning state ("rested / not", "looking forward / not"), and it responds to evening detachment most. Evening gains should be low-arousal (detach, relax), with mastery-type gains earlier.

3.10 An evening look at what went well is itself a gain. Bono JE, Glomb TM, Shen W, Kim E, Koch AJ. 2013. Building positive resources: effects of positive events and positive reflection on work stress and health. Acad Manage J 56(6):1601–1627. doi:10.5465/amj.2011.0272 — verified (Crossref).

• Three-week field study with an intervention: positive events and negative events each contributed independently to stress, blood pressure and symptoms; a brief end-of-workday reflection on what went well lowered stress and improved health that evening.
• Field experiment, healthy employees. Moderate.
• Relevance: (a) more evidence that good and bad add up on separate axes (3.1); (b) the evening check-in can be framed as noticing what was good (a gain) rather than grading the day (a cost). This partly answers Etkin's warning (§7.6): the question itself can be restorative if it is about noticing, not scoring.

3.11 Operations research / recommender work. Searched OpenAlex 2026-10-05 ("multi-objective daily schedule well-being", "wellbeing-aware scheduling calendar recommender", "activity recommendation daily schedule wellbeing", "personal scheduling stress recovery breaks optimization"). No peer-reviewed model was found that plans an individual's day with well-being terms and validates it on outcomes; hits were industrial scheduling, crew fatigue rules, or design essays (e.g., "Perfect days. A benevolent calendar to take back your time", Zenodo 2019, doi:10.5281/zenodo.2628970 — grey literature, not read). Commercial auto-schedulers (Reclaim, Motion, Viva Insights focus time) protect habit or focus blocks by priority rules; no published outcome data found. Conclusion: the constraint formulation in §5 is engineering inference from the psychology above, not an established method.

4. Prior art in apps and their evidence¶

Already covered: Structured's Energy Monitor (capacity report §5.1), Tiimo and other planners (§5.2–5.10, no outcome studies found), Visible (§2.19), Bearable (§2.20).

4.1 Structured "Energy Monitor" — the currency model in practice. Vendor documentation (capacity report §5.1). Tasks are marked draining (three intensities), recharging or neutral; recharging points add back to the day's energy, but only mid-day and never above the daily limit. One scale, a hand-set limit, no learning, no evidence of outcome. This is the "gain offsets cost" model, with two caps (only mid-day, never above the daily limit; in the vendor's words "you cannot entirely recharge your energy or more than your daily available energy limit"). Sioul's floor model needs no such caps because gains never enter the cost budget.

4.2 Moodivate (behavioural activation app). Dahne J, Wahlquist AE, Carpenter MJ, et al. 2025. A digital depression treatment program for adults treated in primary care: a randomized clinical trial. JAMA Intern Med 185(6):692–701. doi:10.1001/jamainternmed.2025.0494 — verified (PubMed 40227715). Dahne J, Lejuez CW, Diaz VA, et al. 2019. Pilot randomized trial of a self-help behavioral activation mobile app for utilization in primary care. Behav Ther 50(4):817–827. doi:10.1016/j.beth.2018.12.003 — verified (PubMed 31208690; app description from the PMC full text).

• App: the person names values, plans value-driven activities, marks them done, and monitors mood daily; badges reward use and completed activities. 2025 RCT, N = 649 with at least moderate depression, 22 primary-care clinics, 12 weeks: BDI-II fell by 10.3 (Moodivate) and 9.9 (with EHR link) vs 5.9 (usual care); odds of clinically significant improvement 2.5–3.0× and of remission 2.3–2.6× usual care. Weekly retention 68–100% in month 1; 33% still using at 12 weeks. Two developers co-own the company selling it (declared). Pilot (N = 52): both a BA app and a CBT app beat usual care; ~47 sessions of ~3.5 min over 8 weeks; 70% still using at 1 month, 50% at 8 weeks.
• Large pragmatic RCT. Strong for "a self-guided app that schedules valued activities reduces depression"; it cannot separate the activity scheduling from the other parts (mood tracking, badges, values work).
• Relevance: the best direct evidence that a planner which schedules chosen, valued activities helps. It schedules by values, not by a gain score, and it does not set a daily gain quota. Its badges conflict with Sioul's no-gamification rule; nothing in the trial shows the badges were needed.

4.3 Ly KH, Trüschel A, Jarl L, et al. 2014. Behavioural activation versus mindfulness-based guided self-help treatment administered through a smartphone application: a randomised controlled trial. BMJ Open 4(1):e003440. doi:10.1136/bmjopen-2013-003440 — verified (PubMed 24413342).

• N = 81 with major depression, 8 weeks: no overall difference between the BA app and the mindfulness app; BA did better in more severe depression, mindfulness in milder depression (6-month follow-up).
• Small RCT, no inactive control. Weak-to-moderate.
• Relevance: activity scheduling did better than mindfulness when depression was more severe — when anticipated pleasure is lowest too (1.10). In milder states, the way activities are experienced (attention, detachment) may matter more than adding them.

4.4 Huguet A, Rao S, McGrath PJ, et al. 2016. A systematic review of cognitive behavioral therapy and behavioral activation apps for depression. PLoS ONE 11(5):e0154248. doi:10.1371/journal.pone.0154248 — verified (Crossref).

• 117 depression self-help apps; 12 (10%) were consistent with CBT/BA principles; none had efficacy studies; usability variable; privacy and safety policies rare.
• Systematic review of the market. Moderate (dated).
• Relevance: most "mood + activity" apps implement the core of BA loosely; the app market is not evidence.

4.5 Daylio (mood and activity journal). Chaudhry BM. 2016. Daylio: mood-quantification for a less stressful you. mHealth 2:34. doi:10.21037/mhealth.2016.08.04 — verified (Crossref; an app review, not a study).

• The person logs a mood (five faces) and ticks the activities of the day; the app shows which activities co-occur with good or bad moods. No outcome trial found.
• App review / vendor. Opinion.
• Relevance: an alternative to rating each activity's gain: infer which kinds of activity go with good days from one daily mood entry plus the list of what was done. Fewer ratings (Etkin, capacity report §7.6), but confounded (good days allow pleasant activities, not only the reverse) and slow to learn.

4.6 Learning each person's own sources of pleasure from their data. van Roekel E, Vrijen C, Heininga VE, Masselink M, Bos EH, Oldehinkel AJ. 2017. An exploratory randomized controlled trial of personalized lifestyle advice and tandem skydives as a means to reduce anhedonia. Behav Ther 48(1):76–96. doi:10.1016/j.beth.2016.09.009 — verified (Crossref; abstract read from the authors' preprint, doi:10.31234/osf.io/b94sy).

• 69 young adults with persistent anhedonia rated pleasure and lifestyle (activity, social contact, sleep…) 3 times a day for 2 months; after month 1, two groups got advice based on their own observed links between lifestyle and pleasure (one also a tandem skydive). Monthly questionnaires showed no group difference, but momentary positive affect and pleasure were higher in the following month in both advice groups than in controls; the skydive added nothing.
• Small exploratory RCT. Weak-to-moderate.
• Relevance: the closest test of what Sioul would do — learn from a person's own ratings which activities give them pleasure, then plan more of those. It moved daily pleasure a little, not global measures, within a month. Supports G19 (suggest from the person's own well-rated items), and modest expectations.

4.7 Finch (self-care with a virtual pet). No peer-reviewed outcome study found (searched 2026-10-05: web, OpenAlex); only press and app reviews (e.g., https://yourstory.com/2022/06/app-review-self-care-pet-finch-gamifies-mental-wellbeing). Grey literature.

• Self-care goals (from brushing teeth to yoga) feed a virtual pet that grows; mood check-ins; reflections. Reported high retention (market-analyst claims, unverified).
• Relevance: the "gain floor" turned into a game of daily goals. Retention is not well-being; the gamified quota is exactly what Sioul rules out (capacity criteria 44–47; Etkin; Silverman & Barasch).

5. Design criteria for Sioul¶

These extend capacity report criteria 38–44 (gains) and keep all its other criteria. "Inference" marks a rule built from several sources; numbers marked "default" are starting values to tune, not findings.

What to rate G1. One gain rating per item, 0–10, "what it gave back", recorded after the item. Anchors: 0 = it gave nothing back (it only took); 5 = some rest, pleasure or sense of having done well; 10 = left me clearly restored, glad or proud. One item is enough for a daily total: multi-domain well-being collapses onto one factor at the day/person level, single items track full scales at r ≈ .6, and their noise (≈1 point on 10) averages out over many items — 2.5 Goodman 2018, 2.13 Wanous 1997, Cheung & Lucas 2014, Dejonckheere 2022, §6.4 Beck. G2. Add one optional "kind" tag per item instead of more ratings: rest (relax / switch off), pleasure, mastery (doing something well), meaning (values, care for others), connection, movement/outdoors. Defaults come from the item's category; the person can change them. Used for placement (G15) and for optional weekly kinds (G11), not for the floor — 2.3 Newman 2014 / Kujanpää 2021 (DRAMMA), §4.8 Sonnentag & Fritz, 2.1 Huta & Ryan, 2.2 White & Dolan. Offer Beck's two ratings (mastery, pleasure) as an advanced option for people who want them (§6.4). G3. Every item can carry costs and a gain at once. Never derive a gain from the category alone for movement or social items, and let social items carry a cost charged a few hours later — 1.11 Leikas 2017, 2.6 Csikszentmihalyi 1989, 2.7, §6.13, §2.17. G4. Needs that Sioul already plans (sleep, meals, naps, medication) do not count toward the gain floor, so that sleeping long cannot "fill" it; a meal counts only if the person rates it (a shared dinner, a meal out) — inference from 2.10 and the need/gain split.

When to rate G5. Ask for a gain rating after the item, rarely: on the first few uses of an item or kind, then about one time in five (default), and whenever the person opens the item; never during it. Between ratings, use the learned value — §7.6 Etkin 2016 (measuring enjoyment lowers it), 2.13 (single ratings are noisy anyway), §7.6 Ancker 2015. G6. Plan with the learned after-rating of the item or kind (median of its last 5 after-ratings, default), not with the person's forecast; the forecast is used only until after-ratings exist. Do not "correct" low gain forecasts upward: in daily life people predicted pleasure about as they then felt it, depressed people included (both low), while displeasure was over-predicted — 1.10 Wu 2017; this revises the gain half of capacity criterion 7 (the anxiety half stands). G7. One optional end-of-day tap: "Today's mix: too much duty / about right / too empty", with an optional "something good today?" note. Together with the capacity report's next-morning "rested / not" (criterion 14), this is the outcome the gain floor learns from — 2.11 occupational balance (perceived mix, OBQ), 1.7 Sharif 2021 (too little and too much free time), 3.9 Sonnentag 2008 (evening gains show up next morning), 3.10 Bono 2013 (noticing what went well lowers evening stress).

Learning the floor G8. Daily gain = Σ over items of (gain/10) × hours, each item's hours capped at 2 h (default), so the unit is "good hours". The cap reflects plateaus (free time beyond ~2 h/day adds nothing; nature beyond 200–300 min/week adds nothing) and keeps one long film night from filling a week — 1.7, 2.9, 2.2 (DRM weights by duration), inference. G9. Learn from days with a known outcome only (G7 or morning "rested"), and keep rest days in: unlike cost budgets, gains are informative on rest days. Floor = 25th percentile (default) of daily gain on good days, weighted with the capacity report's 14-day exponential kernel over at most 28 days (criteria 16–17); bounded between 0.5 and 4 good hours (defaults); change the shown floor only when the new value differs by >30% (criterion 48). Use next-day or end-of-day outcomes, never same-moment co-occurrence: restful items are chosen when tired, so gain and fatigue go together in the moment (2.2 DRM tiredness column) — inference from §6.12, 1.4, 1.7, 3.9. G10. Do not learn the floor downward from a run of bad days (depression, flare, crisis): a low-gain stretch is when gains are needed most. If fewer than 5 good days exist in 28, keep the default — 1.4 Sonnentag 2018 (recovery paradox), §6.2–6.3 (BA works when mood is low), 4.3 Ly 2014. G11. Default before data: two gain blocks a day of at least 15–30 min each (one mid-day, one evening), about 1 good hour; plus optional weekly kinds the person turns on: time outdoors ≥120 min/week in any split, something for meaning or connection once or twice a week — 3.7 de Bloom 2017 (15-min lunch blocks), §4.11 Albulescu 2022 (>10-min breaks), 2.9 White 2019, 2.1, 2.3 Kujanpää 2021 (meaning and affiliation act on longer scales).

Planning: mixing gains with costs G12. Plan with constraints, never a net score: each cost scale ≤ its budget (capacity criteria 27–28) and daily gain ≥ floor. A gain never raises a cost budget and a cost never "uses up" a gain — 3.1 David 1997 / Cacioppo 1994 / Baumeister 2001, 1.5 COR, 1.6 JD-R, 3.10 Bono 2013, §6.12. G13. When both cannot be met, move optional costs first, then shrink gain blocks to their low-effort minimum, and the planner never removes the last gain block of the day (the person may); if the day is still impossible (unavoidable events), say so plainly and lighten tomorrow (capacity criterion 34) — 1.4 recovery paradox, 3.2 Ong 2006 (positive emotions buffer stress reactivity). G14. Optional check, U-index style: the share of planned waking hours in blocks where the highest cost rating ≥5 and exceeds the gain rating stays under a learned limit (same estimator) — 3.3 Kahneman & Krueger 2006; inference. G15. Placement: (a) a low-effort gain block of ≥15 min right after each heavy block (capacity criteria 29 and 42); (b) spread gain blocks through the day rather than saving them for the evening, because their effect is local in time; (c) in the last 1–2 h before sleep, only low-arousal gains (switching off, relaxing) and no item rated ≥5 on anxiety (default); (d) end the day on something light — §6.8 Fredrickson 2000, §4.11, 3.7 de Bloom 2017, 3.4 Loewenstein & Prelec 1993, 3.9 Sonnentag 2008, §1.8 peak-end. G16. On heavy days protect gain blocks more, not less, and make them low-effort (a walk, lying down, music), because high-stress days are the ones where recovery fails by itself — 1.4 Sonnentag 2018, 3.6 Howe 2022 (low-effort options get done). G17. Pairing is offered, not imposed: suggest adding something pleasant to a costly item (music, company, a nicer place), which raises that item's own gain; never lock a gain behind finishing tasks ("earn your break") — 3.5 Woolley & Fishbach 2016, Milkman 2014, Kirgios 2020, 1.9 Tonietto 2021. G18. The planner proposes the gain slot; the person fills it, moves it, or leaves it empty. An unplanned, empty block counts as gain by default (free time is a gain in itself), up to the item cap — 3.6 Howe 2022 (partial control wanted), 1.8 time affluence, 2.4 autonomy, §6.9 Trougakos 2014. G18b. Keep gain blocks free of notifications and leave a few minutes of slack around them, because the gain of a block depends on being present in it — 2.2 Killingsworth & Gilbert 2010, 2.14 Jose 2012. G19. Suggest gain activities only from the person's own past items with good after-ratings, rotate them, and when an item's after-ratings drift down over repeated use, suggest a variation — 4.6 van Roekel 2017 (personal advice from own ratings raised daily pleasure), 1.3 hedonic adaptation (HAP), 1.2 person–activity fit and variety.

Never G20. No gain score, progress bar, ring, streak, badge, pet or "goal not reached" message; at most a line in words ("there is room for something good this afternoon") — §7.6 Etkin 2016 and Silverman & Barasch 2022, 1.9 Tonietto 2021, 4.7 Finch and 4.2 Moodivate's badges as counter-examples. G21. Never schedule or suggest social activities by default, never present any activity as "good for you", never sort activities into good and bad for the person — §6.9 Trougakos 2014, 2.7, 2.12 (the person decides what nourishes), 2.15 Lee 2018 (variety helps some ages, not others). G22. Neutral wording for gains ("time for you", "something you like"), never "recharge so you can work" or "self-care task"; the person may opt in to a framing that ties a gain to their goals, which helps people who see leisure as waste — 1.9 Tonietto 2021.

Special cases G23. Depression and anhedonia: predicted and felt gains are both low; activity scheduling still works — it matched medication in more severe depression (§6.2) and beat a mindfulness app in more severe cases (4.3). Keep gain blocks small and linked to the person's values, count mastery as well as pleasure, do not drop an item because its learned gain is low, keep the floor at its default (G10), and never display gain totals — 1.10 Wu 2017, §6.2–6.3, 4.2 Moodivate, 4.3 Ly 2014, §6.5 Sherdell 2012. G24. Energy-limiting illness (ME/CFS, long COVID) and burnout: every gain item also costs; its cost ratings count against the budgets like any other item, the default gain kinds are rest and switching off (low cost), movement and social items get no default gain, and social items carry a delayed cost — §2.6 NICE 2021, §2.8 Chu 2018, 1.11 Leikas 2017, §6.13. G25. Autistic burnout: social contact and masking default to cost, not gain, unless the person rates otherwise — §2.17 Raymaker 2020.

Open questions and conflicts between sources¶

1. Duration. DRM weights feelings by duration (2.2); BA practice counts activities, not hours; plateaus (1.7, 2.9) argue for diminishing returns. The 2-h cap per item (G8) and "good hours" unit are guesses.
2. Floor estimator. The 25th percentile of good days (G9) is a guess between "the least that went with a good day" and target-chasing; with 14–28 days of noisy data, a dose–response fit per person is out of reach (capacity open question 16).
3. Daily or weekly. Rest and detachment act within the day and next morning (3.7, 3.9); meaning, connection and nature act over weeks (2.1, 2.3 Kujanpää, 2.9); kindness worked better concentrated on one day a week (1.2). A second, weekly floor per kind may be needed; it adds complexity.
4. Forecast bias. Sherdell 2012 (lab, §6.5) found low anticipatory pleasure drives avoidance in depression; Wu 2017 (daily life, 1.10) found pleasure forecasts matched experience (both low). G6 sidesteps this by planning from after-ratings; whether showing people their past after-ratings ("last time this gave you 6") helps them start, as exposure does for anxiety, is untested.
5. Rating burden vs benefit. Etkin: measuring enjoyment lowers it; Bono: an evening look at what went well lowers stress; BA's effective trials use ratings. How rarely to ask (G5's one in five) is unresolved.
6. Gamification works for adherence. Moodivate (badges) beat usual care in a large RCT; Finch reportedly retains users. Sioul's refusal rests on other evidence (Etkin, Silverman & Barasch, Tonietto) and on its users; whether a calm, unscored app keeps people using it is untested.
7. Offset model. Structured lets recharging tasks offset energy, capped at the daily limit and only mid-day (4.1); no outcome data either way. The floor model rests on separable-axes evidence (3.1, 1.5, 1.6), not on a head-to-head test.
8. Too much gain. Sharif's decline concerns idle time, not rated gain; adaptation (1.3) lowers returns; no source shows harm from a high-gain day. Sioul needs no gain ceiling, only the per-item cap (G8).
9. Gains and tomorrow's capacity. Capacity criterion 26 treats gains as a predictor of tomorrow; buffering is partial and specific (1.6: 18 of 32 combinations). Learning cost budgets only from days that also met the floor is plausible and untested.
10. Variety. Occupational balance values variety (2.11); activity diversity helped older adults, not younger ones (2.15). Left to the person.
11. Framing leisure. Tonietto: tying leisure to goals helps those who see leisure as waste; autonomy and anti-instrumental arguments point the other way. G22 makes it opt-in.
12. Operations research. No validated personal-day scheduler with well-being terms was found (3.11); the constraint design (G12–G18) is engineering inference.

Sources checked (verification log)¶

• Method (2026-10-05, 12:06–12:40): every DOI was resolved on Crossref (title, authors, year, venue); abstracts were read from OpenAlex or PubMed; two full texts were read (see below). "Verified" in the entries means the Crossref (or PubMed) record matched. Details not in the abstract were confirmed, where stated, through the named web page (authors' institution, press release, Greater Good summary); otherwise they are marked "not re-read today" and should be checked before quoting.
• Not peer-reviewed, marked as such: Segal, Williams & Teasdale 2013 (book); Structured, Daylio, Finch, Bearable, Visible (vendor/press); "Perfect days" (Zenodo design essay, not read); Saha & Iqbal 2023 and Howe et al. 2022 are peer-reviewed conference papers (CHI), the first an extended abstract.
• Full texts read: Huta & Ryan 2010, Kahneman et al. 2004 (both in pdf-wellbeing/, 0.4 MB together); Moodivate pilot (PMC, online).
• Claims carried with a "not re-read today" flag: Lyubomirsky 2005 gratitude-frequency finding; Wagman 2012 definition wording; Matuska & Christiansen's five dimensions; MBCT session-7 exercise; the "paradox of work" sentence (Csikszentmihalyi & LeFevre 1989).
• Searches that found nothing usable: OpenAlex queries for well-being-aware personal scheduling (3.11); peer-reviewed outcome studies of Finch (web, OpenAlex).
• No reference in this file was invented; none is left unverified at the DOI level.