How You Decide: The Science of Human Decision Making

1. 1Name the hard judgment or choice you actually face (e.g., 'What is the likely return on this rental property?'), then ask what easier question your mind is quietly substituting for it—availability substitutes 'how easily can I recall examples?' for 'how probable is this?'
2. 2Locate the anchor before you estimate any number. Write down the first figure that entered your head and label its source. If it is unrelated to the task—like a spun wheel landing on 65—flag it as noise, not evidence.
3. 3Widen your adjustment deliberately. Because System 1 fixates on a salient number and System 2 typically stops adjusting too soon, push your estimate further from the anchor than feels comfortable, then check both a much higher and much lower value.
4. 4For likelihood judgments, ask why an example came to mind easily. Recent news, vividness, or personal exposure inflate availability without changing actual base rates, so discount ease-of-recall that has a non-probabilistic cause.
5. 5Classify the interaction mode. Decide whether System 1 is influencing (biasing a number you don't notice) or informing (feeding you the feeling of difficulty as data), so you know which correction to apply.
6. 6Treat habits as trained System 1 responses awaiting System 2 approval or neglect. Before acting on a habitual choice, insert one deliberate check: does System 2 endorse this, or is it merely failing to resist?
7. 7Accept that no heuristic guarantees correctness; judge it by whether it reaches an acceptable outcome most of the time, and reserve full System 2 effort for high-stakes decisions where an acceptable-most-of-the-time rule is not good enough.

In the classic anchoring study, participants were shown a spun wheel and asked whether the percentage of African countries in the UN was higher or lower than the number it landed on—10 for one group, 65 for another. Both groups then produced their own estimate. The group anchored at 10 gave a median estimate of 25%; the group anchored at 65 gave a median of 45%—a full 20 points higher. The wheel is random and unrelated to UN membership, yet System 1 fixated on the salient number and biased the System 2 answer without participants' awareness. This is the influenced mode: System 2 disagrees with the raw anchor, starts adjusting, but stops short. The operational lesson: when you catch a salient number sitting in your head before you estimate, name it as an anchor and force your adjustment wider than feels natural, because your untrained instinct will under-adjust.

• If the first number in your head is unrelated to the quantity you're estimating, then treat it as an anchor to discount, not a data point to adjust from.
• If an example comes to mind easily because it was recent, vivid, or personal, then lower your probability estimate—ease of recall here reflects salience, not frequency.
• If a choice feels automatic and cue-triggered, then it is a habit (trained System 1); pause and require explicit System 2 endorsement before acting.
• If the decision is low-stakes, then let the heuristic run—acceptable-most-of-the-time is sufficient; if high-stakes, then engage System 2 in solo operation.
• If you can identify which system is doing what (influencing vs. informing), then apply the matching correction: strip the biasing anchor, or reinterpret the feeling of difficulty as unreliable evidence.

• Under-adjusting from an anchor. Tell: your estimates cluster suspiciously near whatever number was mentioned first, even irrelevant ones.
• Mistaking recall ease for probability. Tell: you rate risks higher right after seeing a vivid news story, and can't cite a base rate.
• Letting a habit pass as a decision. Tell: you 'chose' something you can't remember evaluating, triggered by a cue rather than a reason.
• Assuming a heuristic that usually works cannot fail here. Tell: you skip System 2 on a consequential, novel decision because the rule of thumb feels familiar.
• Substituting the easy question without noticing. Tell: you answer confidently but the question you answered isn't the hard one you actually needed solved.

• ▸Hard question I actually face: state the true judgment or choice.
• ▸Easier question my mind substituted: the shortcut question being answered instead.
• ▸First number/anchor in my head + its source: is it related to the task?
• ▸Adjustment stress-test: a much-higher and a much-lower estimate to bracket the anchor.
• ▸Why this example came to mind: recency, vividness, exposure—or genuine frequency?
• ▸Interaction mode: is System 1 influencing (hidden bias) or informing (feeling as data)?
• ▸Habit check: is this cue-triggered? Does System 2 explicitly endorse it?
• ▸Stakes rating: low (let heuristic run) or high (engage deliberate System 2)?
• ▸Corrected judgment: my estimate/choice after stripping the bias.

1. 1Identify which of Cialdini's three levers fits your situation: reciprocation (people feel obligated after receiving), social proof (people copy what similar others do), or authority (people defer to legitimate expertise).
2. 2To trigger reciprocation, give something first and give it unconditionally — a small, unexpected gift precedes the ask, so the recipient carries a felt obligation into the decision.
3. 3Keep the reciprocal gift small and cheap relative to what you ask for; the obligation is disproportionate to the gift's monetary value, as the address-label and restaurant-mint cases show.
4. 4Sequence the gift before the request, not after; the obligation must exist in the mind at the moment of choice.
5. 5Read the incidental social cues in an environment, not just the intentional pitches — the passages define social influence as choices shaped by both 'intentional and incidental behaviors of others.'
6. 6When you are the target, name the lever being pulled ('this free gift is a reciprocation prime') to convert an automatic obligation into a conscious, optional decision.
7. 7Test your influence attempt against boundary conditions: what works in one setting may not generalize, so treat any single field result as internally valid but not automatically transferable.

A company selling insurance to small businesses had low response rates to its market-research surveys. Because reciprocation obligates people to respond in kind after receiving something, researchers tested including a small gift or payment along with the survey request rather than asking cold. The parallel from the passages is the restaurant experiment: simply including mints with the bill produced larger tips, even though the mints cost the server almost nothing. In both cases the mechanism is identical — an unexpected, unconditional gift arrives first, and the recipient feels obligated to give back, which raises compliance with the request that follows. The gift's monetary value is trivial next to the behavior it produces, which is the point: reciprocation is disproportionate. The lever works precisely because the obligation feels like a social norm rather than a transaction.

• If you need compliance and can give something small first, then lead with an unconditional gift before making the ask.
• If the target can plausibly see similar others choosing an option, then surface that social proof rather than arguing the option's merits.
• If your credibility rests on legitimate expertise, then make the authority visible before the recommendation, not after.
• If a documented effect comes from one field setting, then treat it as internally valid but verify boundary conditions before assuming it generalizes.
• If you notice a felt obligation, warm feeling, or deference arising in yourself, then pause and check whether an incidental cue, not the option's actual value, is driving it.

• Making the gift conditional or transactional ('do this and I'll give you that') — the tell is that the target treats it as a trade, not an obligation, and the reciprocity pull disappears.
• Delivering the gift after the request instead of before — the tell is no lift in compliance because no obligation existed at the moment of choice.
• Overspending on the gift assuming bigger equals stronger pull — the tell is diminishing returns, since a cheap mint or address label already produces the effect.
• Assuming a result from one population or setting transfers everywhere — the tell is a failed replication in a new context that you didn't screen for boundary conditions.
• Treating social influence as only intentional persuasion and ignoring incidental cues — the tell is being surprised when environment or others' behavior shifts a choice you thought was purely rational.

• ▸Decision at stake: state the specific choice you want to shape or protect.
• ▸Lever in play: name it — reciprocation, social proof, or authority.
• ▸Gift/cue given first: what precedes the ask, and is it unconditional?
• ▸Timing check: does the cue exist in the target's mind before the choice?
• ▸Cost-to-effect ratio: is the gift small relative to the requested behavior?
• ▸Intentional vs. incidental: which cues are deliberate and which are environmental?
• ▸Boundary condition: what setting or population might make this fail?
• ▸Self-defense name: if you're the target, what obligation/warmth/deference are you feeling and why?

1. 1Separate the decision process from the decision outcome: name the specific behavior you want to predict or produce (which option chosen, amount spent, risk taken, or whether any choice happens at all).
2. 2Trace the outcome back to the System 1/System 2 interaction that generated it — was the choice an approval (System 2 rubber-stamped a gut response), an override (System 2 blocked it), or informed (System 2 read System 1's ease-of-recall as evidence)?
3. 3Check whether a heuristic is silently driving the outcome: anchoring on a starting number, availability from vivid/recent examples, or representativeness from a stereotype match.
4. 4Predict the direction of bias: familiar, recent, and vivid inputs inflate perceived likelihood, so outcomes tied to those events (buying insurance, avoiding a risk) will be overweighted.
5. 5To move an outcome, plant a seed before the choice — ask the person to simulate the behavior (mere measurement) rather than argue for it directly.
6. 6Measure the outcome as a rate against a baseline, not as a single anecdote, so you can see whether your intervention actually shifted behavior.
7. 7Attach the emotional appraisal: identify the specific emotion present (not just positive/negative) using certainty, control, and pleasantness, because distinct emotions push risk and spending in different directions.

In the car-purchase study, the outcome measured is a single behavior: buying a new car. The baseline group of roughly 5,000 people who answered the survey once showed a purchase rate of about 2.4%. The treatment group of about 3,500 was first asked when they planned to buy a car, then whether they had recently bought one — and their purchase rate rose to 3.3%. That is roughly a 50% relative increase in the outcome, produced by nothing more than one question on a mail-in survey. The mechanism is mere measurement: asking about the behavior triggers a mental simulation of life with and without a new car, and that simulation seeds System 1, which later flows into a System 2 purchase decision. The lesson for shaping outcomes is that prompting simulation of a behavior can raise the rate of that behavior even for large, deliberate purchases.

• If the outcome must be predicted and the event is vivid, recent, or media-heavy (shark attack, flood), then assume the person overestimated its likelihood and expect over-cautious choices.
• If a numerical outcome (price paid, quantity, estimate) is at stake, then look for the anchor that was present first — the outcome will be pulled toward it.
• If you want to raise the rate of a behavior, then have people simulate or answer a question about that behavior beforehand rather than persuade them.
• If System 2 merely approved a System 1 impulse, then treat the outcome as unscrutinized and easily reversed by adding a moment of deliberation.
• If you judge decision quality, then include the specific emotion at the time of choice, since certainty and control alter risk preference beyond mere pleasantness.

• Judging an outcome by its process quality alone — the tell is calling a lucky System 1 approval a 'good decision' because it worked out, ignoring that System 2 never scrutinized it.
• Believing a small run of results is representative (law of small numbers) — the tell is betting the next coin flip is 'due' after four tails, or trusting a behavior rate from a tiny sample.
• Reading vividness as probability — the tell is spending on travel insurance or avoiding an activity because dramatic examples come easily to mind.
• Anchoring the estimate and forgetting it — the tell is an outcome that clusters suspiciously near the first number mentioned, with only small adjustments away from it.
• Treating all negative or all positive moods as equivalent — the tell is predicting the same choice under fear and anger when their certainty and control appraisals differ.

• ▸Target behavior: state the exact outcome (option, amount, risk, or no-choice)
• ▸Baseline rate: the outcome frequency with no intervention
• ▸System interaction: mark approval / override / informed
• ▸Active heuristic: anchoring, availability, or representativeness (name the input)
• ▸Bias direction: is likelihood over- or under-estimated, and why (familiar/recent/vivid)
• ▸Emotion present: name the specific emotion and its certainty + control reading
• ▸Seed used: what simulation or question was planted before the choice
• ▸Measured rate: the outcome frequency after intervention vs. baseline
• ▸Downstream satisfaction: how the chooser now rates the decision

1. 1Map the raw information first: list every option, exactly how each is described, and the context it will be evaluated in—because objectively equivalent information can produce different choices depending on wording.
2. 2Identify the reference point. Ask whether the frame anchors people to a status quo, a gain, or a loss; the default and the wording set the baseline against which everything else is judged.
3. 3Choose your frame deliberately. To exploit loss aversion, describe outcomes as losses to push risk-seeking; to secure sure choices, describe outcomes as gains to push risk-aversion.
4. 4Set the default to the outcome you want as the path of least resistance, since default/status-quo information gets preferential treatment by decision-making machinery.
5. 5Partition the decision if you want to shift weighting—e.g., ask a screening question first ('How much do I want to spend?') before the choice question to reframe tradeoffs.
6. 6Arrange the assortment: add a decoy (asymmetrically dominated option) or a middle option to steer choice via the decoy or compromise effect.
7. 7Control evaluability—decide whether options are seen jointly or separately, because context changes how attributes are weighted.
8. 8Test boundary conditions: check whether the effect survives when the frame is pushed to extremes (e.g., an aggressively low restraint) rather than assuming it always holds.

Kahneman and Tversky's Asian disease problem shows framing at work. A group told that Program A saves 200 of 600 for sure, versus Program B with a one-third chance to save all and two-thirds chance to save none, chose the sure gain—72% picked Program A. A separate group read the identical scenario framed as deaths: Program C means 400 die for sure, Program D gives a one-third chance nobody dies and two-thirds chance all 600 die; here 78% chose the gamble, Program D. The two pairs are mathematically identical—200 saved equals 400 dead. But the gain frame ('saved') made people risk-averse toward the sure outcome, while the loss frame ('die') made them risk-seeking. The only variable changed was the description, yet it flipped the majority preference. This is the raw-material effect: same information, different weight, different manufactured decision.

• If you want people to accept a certain outcome, frame it as a sure gain; if you want them to take a risk, frame the sure option as a loss.
• If you want higher participation on any option, make it the default—opt-out organ donation systems ran 85.9–99.98% versus 4–27.5% for opt-in.
• If tradeoffs between price and quality are blocking a sale, partition the decision so price is screened first, making any in-budget option feel 'already acceptable.'
• If you're adding a third option, use an asymmetrically dominated decoy to boost a target option, or a middle option to trigger the compromise effect.
• If someone claims a framing effect is universal, test its boundary conditions before relying on it—an aggressively low restraint or hypothetical stimuli may break it.

• Assuming equivalent descriptions produce equivalent choices. Tell: you present '90% survive' and '10% die' interchangeably and are surprised by inconsistent decisions.
• Treating the default as neutral. Tell: you set opt-in as the default expecting people to actively choose, then see participation collapse to single digits.
• Ignoring irrelevant surrounding information. Tell: you assume people filter to only what matters, but once context is collected it gets used and shifts weighting.
• Overgeneralizing an effect past its limits. Tell: you apply a lab finding (e.g., a price restraint) without checking whether it holds when set aggressively low or in real, non-hypothetical stakes.
• Framing gains and losses backward for your goal. Tell: you emphasize what's saved when you actually needed people to take a risk, and they default to the safe choice.

• ▸Decision to influence: state the single choice you are shaping
• ▸Options list: write each option and its exact wording as presented
• ▸Reference point: name the status quo/baseline each option is judged against
• ▸Frame chosen: mark gain-frame or loss-frame and the risk behavior it should trigger
• ▸Default set to: name the option positioned as path of least resistance
• ▸Partition question: write the screening question asked before the main choice, if any
• ▸Assortment move: note any decoy or compromise option added and its target
• ▸Evaluability mode: state whether options are seen jointly or separately
• ▸Boundary check: list the extreme condition you will test before relying on the effect

1. 1Map your day's demands against a single fuel tank: list every task that requires attention control, emotion regulation, impulse override, or behavioral modification, and treat them as drawing from one shared executive resource, not separate reserves.
2. 2Schedule high-self-control decisions (tradeoffs, precise calculations, rule-following—all System 2 work) for periods before you have exercised much self-regulation, not after a chain of resisted impulses or prior choices.
3. 3Chunk similar tasks so you stop switching mindsets: group abstract work together, concrete work together, auditing separate from forecasting, because each mental-gear switch consumes executive resources.
4. 4Convert repeated System 2 decisions into habits so System 1 handles them for free; deliberately train System 1 on repetitive tasks to free System 2 for decisions that genuinely require tradeoffs.
5. 5Recognize decision fatigue as a source of depletion: after making many consumer-style choices, expect weakened self-regulation, so avoid stacking a self-control demand right after heavy decision-making.
6. 6Account for individual training: the same stimulus deletes people unequally—resisting a cigarette costs a long-time smoker executive resources but costs a never-smoker nothing—so calibrate demands to whom the System 1 impulse is strong for.
7. 7When you notice resources are low (more emotional slips, diet cheating, skipped gym), defer any decision that requires a tradeoff or rule-following and let the resource replenish with time.

In the mindset-switching study, people made a series of choices; some were told to keep switching their decision strategy on every choice, others kept a single strategy throughout. Afterward, the dependent measure came from a different self-regulation domain entirely: participants watched several minutes of a funny stand-up routine while their faces were filmed and were instructed to show no emotion. On average, people were about twice as bad at maintaining a neutral expression when they had switched back and forth between decision strategies than when they had used a single strategy. Because there is one unitary executive resource powering one unitary executive function, the strategy-switching drained the same tank that emotion suppression needed. The effect held for switching of all kinds—abstract to concrete, approach to avoidance, even switching languages for bilinguals. The practical lesson: an accountant who both audits and forecasts should chunk those tasks rather than alternate all day, because each switch taxes the shared reserve and leaves them prone to self-control failures later.

• If a decision requires tradeoffs, precise calculation, or overriding an impulse, then make it while resources are fresh; if resources are depleted, then defer it or expect a System 1 default.
• If a role forces multiple mindsets (auditing vs. forecasting, abstract vs. concrete), then batch like tasks together instead of alternating, because each switch spends executive resources.
• If you have just made many decisions in a row, then treat your self-regulation as compromised and avoid scheduling a diet, temper, or discipline test immediately after.
• If a task is repetitive and stable, then train System 1 to automate it so System 2 is freed for decisions that need it.
• If an impulse is strong for a given person (a trained System 1 response like a smoker's craving), then budget more executive resource for their resistance than for someone with no such trained response.

• Assuming each self-control domain has its own tank—the tell: someone resists dessert all afternoon then blames unrelated stress for snapping at a colleague, missing that both drew on one unitary resource.
• Treating decision-making as free—the tell: a day of choosing t-shirts, candles, and shampoos leaves you unable to stick to a plan that evening, and you attribute it to weak willpower rather than decision-induced depletion.
• Multitasking self-control—the tell: alternating between mindsets all day produces more emotional outbursts and skipped commitments, because switching itself burns resources like the neutral-expression failures showed.
• Applying a uniform depletion estimate to everyone—the tell: expecting a former smoker and a never-smoker to be equally tired after a smoke-free event, ignoring that System 1 training makes the resistance costly for only one of them.
• Expecting instant recovery—the tell: pushing a second demanding decision right after a draining one and getting a System 1 default, when resources only replenish with time.

• ▸Decision to make: name the choice and note whether it needs tradeoffs or rule-following (System 2) or can run on habit (System 1).
• ▸Self-control demands today: list attention, emotion, impulse, and behavioral tasks drawing on the single tank.
• ▸Prior depletion: what self-regulation or decisions have you already made before this point?
• ▸Mindset switches: count the abstract/concrete, approach/avoidance, or role switches you're forcing.
• ▸Chunking plan: which similar tasks can you batch to stop switching?
• ▸Person-specific cost: whose trained System 1 impulse makes this resistance expensive, and how much?
• ▸Timing: is this scheduled while resources are fresh or after depletion—defer if depleted.
• ▸Habit candidate: which repeated decision can you train System 1 to automate to free System 2?
• ▸Recovery window: when can resources replenish before the next high-control demand?

1. 1Name the active goal before you decide: state the desired end state as a gap between the world now and the world you want, since that gap is the tension driving the choice.
2. 2Identify which knob on the control panel is live — a goal, a mindset, or an emotional appraisal — because each bends the same cognitive machinery toward a different end.
3. 3Check whether a mindset is spilling over from a prior task: once a set of cognitive procedures activates, it stays active and colors the next unrelated decision.
4. 4Watch for self-imposed structure that quietly reframes the goal: setting a spending limit, listing reasons, or reserving an option changes what you optimize for, not just how you decide.
5. 5Audit the emotional state along the six appraisal attributes (certainty, pleasantness, attentional activity, anticipated effort, control, responsibility) rather than lumping it as merely positive or negative.
6. 6Decide whether you want a deliberative mindset (compare all options on all attributes) or an action mindset (eliminate options one by one) — and avoid switching between them mid-session, which depletes self-regulation.
7. 7Offload stable, repeated decisions to System 1 as habits so System 2 is free to govern the decisions that actually need judgment.
8. 8Recheck the control-panel setting after the decision: if the reason you gave doesn't survive removing the trigger cue, the motivation, not the logic, drove the choice.

In Simonson's vacation study, students imagined finishing an exam they weren't sure they'd passed, with a discounted Hawaii package expiring before results came out. Given three options, 61% paid a nonrefundable $5.00 fee just to defer the buy until they knew the outcome. But when other groups were told in advance they passed or failed, about 55% bought the trip either way — the exam result didn't change the behavior. What changed was the motivational state: knowing the outcome supplied a reason (passing = a reward, failing = consolation), and without a reason the tension of an undecided goal made people pay to wait. The lesson: the motivational control panel, not the facts, decided; the same information produced a different choice depending on whether a reason-based mindset was satisfied.

• If you catch yourself paying or delaying just to obtain a 'reason' for a choice you'd make anyway, then the goal is already resolved — act now.
• If a mindset from a previous task is still active, then defer consequential decisions until it dissipates, since active procedures bias subsequent judgments.
• If a decision repeats with a consistent cue and reward, then convert it to a habit to free System 2 for higher-stakes choices.
• If you must alternate between deliberative and eliminative decision styles, then expect self-regulatory depletion and reduce the number of hard choices that follow.
• If you're about to set a pre-commitment (a spending limit, a plan), then confirm it's steering you toward your goal, not silently licensing more expense as in the pen study.

• Treating emotion as just positive/negative — the tell: two 'negative' emotions (fear vs. anger) produce opposite choices you can't explain because you ignored certainty and control.
• Mistaking a self-imposed constraint for restraint — the tell: setting a spending limit before choosing, then spending more ($2.10 vs. $1.64) because the limit reframed the goal as 'spend up to X.'
• Switching decision mindsets repeatedly and blaming poor choices on the options — the tell: fatigue and emotion-regulation lapses after alternating styles across categories.
• Believing you control the panel fully — the tell: switches flip and dials spin without your awareness, so you rationalize a motivated choice as a reasoned one.
• Manufacturing reasons to ease or defer a decision — the tell: your choice stays the same regardless of the outcome that supposedly justified it, as in the Hawaii study.

• ▸Desired end state: write the concrete gap between now and what you want.
• ▸Active goal type: is this a biological set point or a higher-order aspiration?
• ▸Live knob: goal, mindset, or emotion — which is driving this decision?
• ▸Mindset carryover: what prior task might still be active and coloring this?
• ▸Emotion appraisal: rate certainty, pleasantness, attention, effort, control, responsibility.
• ▸Decision style: deliberative (compare all) or action (eliminate) — pick one, don't switch.
• ▸Reason check: would I choose the same if the triggering reason were removed?
• ▸Habit candidate: can this repeated choice be offloaded to System 1?

1. 1Classify emotion as cognitive machinery, not distraction: treat the affective state as an active input that transforms information into a decision, rather than something to suppress before deciding.
2. 2Separate valence from appraisal dimensions before acting: name whether the feeling is positive or negative, then rate its certainty, sense of control, and anticipated effort — because two states with the same valence (fear vs. anger) push risk preference in opposite directions.
3. 3Check for incidental affect leaking into an unrelated choice: ask whether the current mood comes from the decision itself or from something irrelevant (weather, a lost game, loneliness) that will bias the outcome.
4. 4Estimate executive-function fuel before deciding: if you have recently exercised heavy emotion regulation, impulse control, or focus, assume System 2 is depleted and more System 1 responses will slip through.
5. 5When depleted, defer high-stakes choices or add friction: postpone until resources replenish, or force a deliberate step (write down the decision rule) so System 1 does not dominate through neglect.
6. 6Correct for known state-driven biases: if lonely or socially isolated, expect more risk seeking in financial decisions and dial back accordingly; if in a negative mood, expect larger downside reactions than the situation warrants.
7. 7Re-appraise to shift the state you decide in: change certainty, control, or effort appraisals (gather one fact to raise certainty, break the task down to lower anticipated effort) before committing.

The Journal of Finance study collected daily returns from the main exchanges in 26 countries over 15 years and found a significant correlation between sunny mornings in the exchange cities and positive daily returns — the traders' incidental good mood bled into pricing decisions that had nothing to do with the weather. The Applied Economics Letters study on the English national soccer team and the FTSE 100 sharpens the point: winning produced a mild positive market effect, but losing produced a much larger negative effect. Same domain, but the negative emotional shock moved the market far more than the equivalent positive one, echoing that we are more sensitive to losses than to comparable gains. Neither the weather nor a match result carries real financial information, yet both reliably shifted committed money. The lesson for a decision maker: if your emotional state today traces to something irrelevant to the choice, its influence is a bias to correct, not a signal to trust — and the negative pull will likely be the stronger one.

• If the emotion originates outside the decision (incidental affect like weather or a sports loss), then treat it as noise and discount its pull rather than reading it as information.
• If you are socially isolated or lonely, then expect elevated risk seeking in financial choices and require an extra check before committing to risk.
• If you have just done heavy self-regulation (emotion regulation, impulse control, dieting, focus), then assume depleted executive fuel and defer or add friction to any choice needing System 2 override.
• If two negative states share valence but differ on control (fear = low control, anger = high control), then do not treat them as interchangeable — they move risk preference differently.
• If the emotion arises from the decision itself and reflects genuine stakes, then let it inform the choice rather than suppressing it.

• Treating emotion as a pure distraction to be eliminated (the Plato-era view): the tell is trying to 'decide rationally' while ignoring an affective signal that is actually carrying stake-relevant information.
• Collapsing all feeling into good/bad valence: the tell is being surprised when two positive or two negative moods produce opposite risk behavior because certainty and control were ignored.
• Deciding while depleted and mistaking a System 1 impulse for a considered choice: the tell is a pattern of 'outbursts, diet cheating, gym skipping' clustering after periods of intense self-control.
• Reading incidental affect as decision-relevant: the tell is confident commitment on a sunny day or right after a win that you cannot justify on the facts.
• Underweighting the loss side of emotional swings: the tell is being blindsided by an oversized negative reaction (as when a team loss moved the market far more than a win) after planning only for symmetric effects.

• ▸Decision at hand: state the choice in one line.
• ▸Current valence: positive / negative / neutral.
• ▸Certainty appraisal: how sure do I feel about outcomes (1-5)?
• ▸Control appraisal: how much do I feel I can influence the result (1-5)?
• ▸Anticipated effort: how effortful does acting feel right now (1-5)?
• ▸Source of the feeling: from this decision, or incidental (weather, loss, loneliness)?
• ▸Executive-fuel check: heavy self-regulation recently? yes/no — if yes, defer or add friction.
• ▸Bias correction: name the likely tilt (e.g., risk seeking if lonely, oversized negative if loss-driven) and the adjustment.
• ▸Commit or defer: proceed now, or postpone until state/fuel replenishes.

1. 1Identify which system is likely dominating: if the response is fast, automatic, and feels obvious, it's System 1; if it requires slow, serial effort and attention, it's System 2.
2. 2Diagnose your own override tendency using Frederick's cognitive reflection test — the bat-and-ball, widget, and lily-pad questions — noting that intuitive answers ($0.10, 100 minutes, 24 days) are wrong and the correct ones ($0.05, 5 minutes, 47 days) require deliberate processing.
3. 3Classify the interaction mode at play: approval (System 2 lets the System 1 impulse through), override (System 2 blocks it), neglect (System 2 too loaded to supervise), influence (a blocked System 1 impulse still colors a later choice), informing, or solo operation.
4. 4Check System 2's load before any decision requiring restraint: if it's occupied, exhausted, or distracted, expect neglect and the System 1 default to win.
5. 5Reduce cognitive load in advance for decisions where you want System 2 in charge — don't decide about higher-order goals (diet, budget) while multitasking.
6. 6Use System 2 to train System 1: repeat a deliberate response until it becomes automatic, as happens when driving shifts from full attention to effortless routine within months.
7. 7Set constraints and rules ahead of time so the right response is the automatic one when System 2 may not be available at the moment of choice.

In the 'Heart and Mind in Conflict' experiments, participants chose between chocolate cake (the System 1 craving) and fruit salad (the System 2, goal-consistent choice). Before choosing, they got a cognitive-load task: the low-load group memorized a two-digit number, the high-load group a seven-digit number. Among the cognitively unburdened, 41% chose the cake. Among those whose System 2 was tied up with the seven-digit number, 63% chose the cake — roughly a 50% increase. System 1 always wants the cake; System 2 normally overrides it in service of a higher-order goal like losing weight. When System 2 is occupied, it can't restrain System 1, so the impulse goes through by neglect. The tell that neglect drove the choice is that nothing changed except the memory burden.

• If a response is fast, effortless, and feels obviously right, treat it as a System 1 output and verify before acting — obvious answers in this class are often wrong.
• If System 2 is distracted, exhausted, or otherwise loaded, then expect neglect and pre-commit to constraints rather than trusting in-the-moment restraint.
• If the decision serves a higher-order goal (weight, budget), then make it while System 2 is free, not while multitasking.
• If a System 1 impulse is overridden but the appetite persists (still thinking about ice cream), then expect influence to surface later and guard against the delayed lapse.
• If you want reliable good behavior at the point of choice, then train System 1 through repetition so the desired response becomes automatic instead of requiring effortful override each time.

• Trusting the intuitive answer on reflection-test-type problems; the tell is answering $0.10 / 100 minutes / 24 days without pausing.
• Making restraint decisions while cognitively loaded and then blaming willpower; the tell is that the lapse (cake, ice cream) tracks distraction, not desire.
• Assuming approval means the choice was scrutinized; the tell is endorsing a System 1 response System 2 never actually examined for accuracy or wisdom.
• Believing an override 'settled' a temptation; the tell is an unexpected later urge — dishing up ice cream after dinner despite passing it up earlier — from lingering System 1 influence.
• Relying on in-the-moment self-control instead of pre-set rules; the tell is getting 'carried away' at the store because no spending limit was fixed beforehand.

• ▸Decision at hand: state the choice in one sentence.
• ▸Fast/automatic pull: what does System 1 want right now?
• ▸Deliberate/goal answer: what would slow System 2 processing recommend?
• ▸Higher-order goal at stake: which long-term goal does this touch?
• ▸Current System 2 load: am I distracted, tired, or multitasking? (yes/no)
• ▸Interaction mode: approval / override / neglect / influence / informing / solo — which is operating?
• ▸Lingering-impulse risk: could a blocked System 1 pull resurface later?
• ▸Pre-commitment set: what constraint or rule did I fix in advance to protect the goal?