Your personal relapse prevention map
G. Alan Marlatt spent decades studying something that treatment programmes had largely ignored: what actually happens in the period leading up to a lapse. Not the dramatic moment of giving in. The hours and days before it , the chain of circumstances, emotional states, and small decisions that, in retrospect, was leading somewhere specific all along.
His relapse prevention model, developed in the 1980s and still one of the most influential frameworks in substance use treatment, identified something important: lapses are not random. They happen in high-risk situations , specific, predictable combinations of internal state and external context that represent elevated vulnerability for each individual person.
The situations are different for different people. But they're consistent within a person. And they're entirely readable in the data , if you've been collecting it.
What Marlatt's model identified
Marlatt's research found that the majority of lapses , across substances and populations , occurred in one of a small number of situational categories:
Negative emotional states , frustration, anxiety, depression, loneliness, boredom. The single most common category. The person isn't celebrating; they're managing something difficult, and the substance is the management tool that's available.
Interpersonal conflict , arguments, unresolved tension, difficult conversations. The substance provides relief from the emotional activation of conflict, or creates distance from it.
Social pressure , direct pressure from others to use, or the implicit pressure of being in a social context where use is the norm and abstaining or moderating requires active resistance.
Positive emotional states , sometimes overlooked, but significant: celebration, excitement, the feeling that a special occasion warrants special indulgence.
Testing personal control , the thought "I can handle just one" after a period of successful moderation. The confidence that the pattern has been broken is itself a risk factor.
Each of these is a recognisable category. What Marlatt found is that individual people have their own hierarchy , their particular combination of high-risk categories, their specific contexts, their characteristic emotional precursors. And those specific patterns can be mapped.
Apparently irrelevant decisions
One of the most useful concepts in Marlatt's model is what he called "apparently irrelevant decisions" , the small, seemingly unrelated choices that set up a high-risk situation without the person recognising them as part of a chain.
The person who "just happens" to take the route home that passes the bottle shop. The person who agrees to meet friends at a venue they know will be difficult when they're trying to moderate. The person who skips dinner before a work event where alcohol will be served. None of these feel like decisions about drinking. Each is a step in a chain that reliably ends in a place they didn't consciously choose.
Marlatt called these decisions "apparently irrelevant" because they appear irrelevant to the substance use outcome , and feel irrelevant to the person making them. But in retrospect, and in the data, they're the scaffolding that the lapse was built on.
How your data maps the pattern
The ayodee daily log, accumulated over several weeks, is a relapse prevention map. The antecedent data , mood scores, stress ratings, context notes , shows the emotional and situational conditions that preceded heavier use. The pattern of when urges peaked, what was happening when consumption was highest, which days of the week or month are reliably more difficult , all of this is in the data.
Reading it through the Marlatt lens involves specific questions:
What are your three highest-consumption days in the last month? Look at the mood ratings that day. Look at the time the first drink was logged. Look at any notes. What was the emotional state? What was the context? Are there common elements?
When do your urges peak? The urge log, reviewed across several weeks, typically shows clustering. 6pm to 8pm on weeknights. Friday afternoons. Days following difficult interpersonal events. The clustering is the high-risk window , the time and context where the probability of use is elevated for you specifically, not for people in general.
What are the days before heavy days like? Sleep quality the night before. Mood during the day. Stress levels. The apparently irrelevant data from the day before the heaviest day is often where the chain starts.
What's different about the low-consumption days? The solution-focused flip of the Marlatt analysis: the days where the risk was present but the outcome was different , what distinguished them? This is your personal resilience data, and it's as useful as the risk data.
The abstinence violation effect
Marlatt identified something specific about what happens after a lapse that determines whether it stays a lapse or becomes an extended episode of heavy use: the abstinence violation effect.
When someone who has set a goal , a drink limit, a sober day, a period of reduced use , breaks that goal, two things happen. They feel guilty and self-critical. And they make an attribution: "I failed because of something about me. I'm the kind of person who can't manage this."
The self-critical attribution converts a behavioural event (exceeded the limit on Tuesday) into an identity event (I'm someone who can't control their drinking). Once it's an identity event, the limit no longer applies , because the limit was a person who could moderate, and that person has now been revealed as a fiction. So the Tuesday overshoot becomes a Wednesday and Thursday of abandonment.
Marlatt's intervention is reattribution: the lapse is a behavioural event, not an identity event. It happened in a specific high-risk situation that can be identified and understood. It's information about that situation , and about the chain of apparently irrelevant decisions that led there , not evidence about who you are.
The data supports this reattribution directly. The overshoot on Tuesday happened in specific conditions that are in the log: the mood rating that morning, the urge intensity in the afternoon, the context in the evening. It's an explicable event with identifiable antecedents. It's not a revelation about your character.
Using the map
The relapse prevention map your data draws is not a prediction of doom. It's a weather map , it shows you where the pressure systems are, where the risk is elevated, what the conditions look like when things get harder.
You can't change the weather. You can know it's coming, and make different choices about what to do when it arrives.
ayodee tracks mood, urges, substance use, and context , the data that draws your personal risk map. Anonymous, no account needed.
References Marlatt, G.A., & Gordon, J.R. (Eds.) (1985). Relapse Prevention: Maintenance Strategies in the Treatment of Addictive Behaviors. Guilford Press.
Witkiewitz, K., & Marlatt, G.A. (2004). Relapse prevention for alcohol and drug problems. American Psychologist, 59(4), 224.