If-Then Plans and the Designed Environment

Implementation intentions: the best-evidenced technique in this subject

An implementation intention is a plan written in the form “if situation X arises, then I will perform behaviour Y”, and it is not a reminder. The mechanism is that specifying the cue makes it perceptually distinctive, so it is noticed earlier, and specifying the response in advance connects the two, so the behaviour is initiated by the situation rather than by a decision. A meta-analysis of more than ninety independent studies found a medium-to-large effect on goal attainment, which is larger than most self-regulation interventions and larger than anything a lecture on discipline produces. The technique appears to work best where the goal is genuinely wanted and the obstacle is a failure to act on it, which is exactly the shape of every bias in this subject. The form has two constraints, and both are where poorly written plans fail. The cue has to be detectable without judgment — a clock time, a loss count, a dollar figure, an order in the ticket, a specific setup appearing on the list. “If I feel off” is not a cue, because recognising it requires the judgment the state has already compromised; “if the session reaches three losses” is a cue that anybody could apply on your behalf. The response has to be concrete and available: close the platform, halve the size, open the journal, place the order at the level. A response that itself needs planning is a second decision carried into the same moment. The evidence for the wider family of self-regulation techniques is uneven, and the differences are worth knowing because the popular version is the wrong one. Fantasising about the successful outcome — the account at a number, the trade working — reliably reduces effort, because the mind treats the imagined achievement as partial progress. What does work is mental contrasting: name the wish, name the best outcome, then identify the internal obstacle and plan against it, which is the sequence behind the WOOP protocol. Paired with an implementation intention, the obstacle in the plan is what the cue is built from, and the two techniques are doing one job — the daydream is not. The remaining lever is the environment, and it is the most reliable of the three because it does not depend on the trader at all. Friction is a design choice: an app that must be reopened, a broker that requires a call, a P&L window that is not on the second screen, a size that is entered from a plan rather than chosen at the ticket. A trader who removes the decision has not resisted the state, they have arranged for it not to matter, which is the same move as the pre-commitment in the previous lesson applied to the machinery rather than to the rule. Where the state still arrives, self-distancing is the tool for the moment: describing the situation from the outside — one’s own name, a third-person frame, what a stranger would see — reduces emotional reactivity and improves the reasoning available, without requiring the trader to feel differently. • Write plans as “if [detectable cue], then [concrete response]”. • Choose cues a stranger could apply for you: a count, a time, a dollar figure, a setup on the list. • Make the response immediate and available; a response that needs planning is another decision. • Use mental contrasting — wish, outcome, obstacle, plan — rather than positive visualisation, which reduces effort. • Put friction on the behaviour and remove the decision; use self-distancing at the moment the state arrives. An if-then plan can also automate something harmful — “if I am down, then I double up” is an implementation intention and it will work. The technique amplifies whatever response is attached to the cue, which is why the response has to be checked against the arithmetic of the previous lessons before it is written.

The sheet, written on a Sunday

A plan that works is short enough to be read in front of the ticket and specific enough that somebody else could apply it while you are in the state it addresses. The natural mistake is to write one line for every failure that can be imagined, which produces a document nobody opens; the useful rule is one line per failure that has actually been experienced. A trader with five real failures gets five lines, and the fifth is usually about the morning rather than the market, because the inputs to a decision are also decisions the plan can hold. Two properties run through the lines below and they are the test of any new one. The cue arrives inside a session and can be recognised without judgment — a count, a clock, a dollar figure, a window on a screen — so it does not depend on the trader assessing themselves. And the response is executable at that moment with nothing further to work out, which means it can be the smallest action that changes the situation: closing something, halving something, leaving for twenty minutes, writing a line. Whenever a proposed line fails either test it is a preference with better wording, and the way to find out is to hand the sheet to somebody else and watch where they have to ask a question. • One line per failure actually experienced, not per failure imagined. • Every line keeps the cue detectable without judgment and the response executable now. • Include the inputs to the decisions — sleep, schedule, screen — not only the trades. • Rehearse the lines in a calm moment; an if-then plan that has never been read will not be recognised in a hot one. • Keep it to something readable in a minute, because it has to be read in the minute it is needed. One trader’s sheet — Sleep under six hours before the open: Small size or no size that day — a check on the person, not the market · Before the session: mark the levels and the setups: The list is fixed before the open; nothing enters it later ← · If the session reaches three losses: Halve the unit until two planned trades are followed ← · If the profit-and-loss window is open: Close it; the written exit governs the position · If an order would exceed the day’s unit: The order is not sent — the cue is a number rather than a feeling · At the close: log the day, the tag and the state: One line each, so the audit has a denominator The sheet is the artefact and the if-then form is the grammar, and the two together are what the audit later prices: the plan says what the trader does, and the record shows whether they did it. A sheet that never changes between reviews is working; a sheet rewritten every week is usually one that is not being followed.

When the plan meets the day

Implementation intentions work because they attach a response to a cue, and they fail for exactly the same reason: an if-then is only as good as the cue’s specificity and the response’s availability. Four failure modes account for most of the plans that get written and not kept. The first is a **vague cue**. “If the market falls sharply” is not a cue; it is a category. A cue has to be a thing you will observe without judgement — a price crossing a level written down in advance, a stop being triggered, a specific number appearing on a specific screen at a specific time. The test is whether two people reading the cue would agree, at the same moment, that it had occurred. The second is a **response that is not possible in the moment**. A rule that says “if I feel the urge to add to a loser, I will call my accountability partner” fails when the partner is asleep. The response should be something you can do from where you are, with what is in front of you, in under a minute — closing the platform, writing the impulse in the log, moving to a different room. A response that requires a different situation is a wish. The third is a **competing habit**. An if-then competes with whatever you already do when the cue arrives, and the existing habit is stronger because it has years of repetitions behind it. The evidence on habit formation is that the automatic version is built by repetition of the same cue-response pair in the same context, not by the initial intention — so the first weeks are the expensive part, and the plan should assume the automatic version does not exist yet. The fourth is the absence of a consequence. Most if-thens specify what to do and not what happens afterwards if the rule is broken, which is where the sequence in this lesson’s lab belongs: a breach is logged, a specific consequence attached, and the rule re-committed rather than quietly abandoned. Without that, one failure ends the plan, and the failure is nearly certain to occur. Two things are worth saying about the wider claim. The technique is well evidenced for closing the gap between intention and action, and it is not well evidenced for producing enduring behaviour without repetition and without an environment that makes the cue reliable. The plan closes the intention-action gap; it does not remove the need for repetition. • A cue has to be observable without judgement, and two people would agree it happened. • The response must be possible in the moment, in under a minute, from where you are. • The competing habit is stronger than the intention until repetition has built the new one. • Write the consequence for a breach, or the first failure ends the plan. A short audit of any written plan: read each if-then and ask whether the cue is specific, the response is immediately available, and the breach has a stated consequence. The rules that fail the audit are the ones that were never really rules.

Sources

• Implementation intentions: strong effects of simple plansGollwitzer (1999), American Psychologist
• Implementation intentions and goal achievement: a meta-analysisGollwitzer & Sheeran (2006), Advances in Experimental Social Psychology
• Mental contrasting and the self-regulation of fantasyOettingen (2012), “Future Thought and Behaviour Change”
• Self-distancing and adaptive self-reflectionKross & Ayduk (2011), Current Directions in Psychological Science