Future rewards tend to lose motivational value faster in ADHD than in people without it.
Naming the phenomenon
Why can I desperately want to do something, be fully capable of doing it, and still be unable to begin?
What looks from the outside like doing nothing is, from the inside, a wrestle you are losing in real time. This is not laziness, not distraction, not lack of motivation. It has a shape, and the shape has a name: intention collision. Two things try to happen at once — the task you intend, and everything competing with it — and they collide before either can start.
Not a lack of will. A collision. And collisions can be resolved.
Meta-analytic imaging work has found reduced ventral striatal response during reward anticipation.
Failure to start may change the emotional value of the task, making the next attempt harder until another force shifts the balance.
The central mystery
Knowing is not the same thing as doing.
A person can understand a task, value the outcome, care deeply about the consequences, and have the skills required to complete it. And still nothing starts.
“I feel stuck at the edge of something I know exactly how to do.”
That gap is what this project names.
The Intention Collision Model
A dynamic competition, not a character flaw.
The model treats task initiation as an evolving state. Two broad forces compete: approach pressure, which pushes toward action, and avoidance pressure, which pushes away from it. When they collide near-equally and neither wins cleanly, the outcome is intention collision — the intention is real, but it can't get through. Executive control helps resolve the conflict.
Approach pressure
- Immediate reward
- Interest and novelty
- Urgency
- Social accountability
- Clear first action
- Visible progress
- Personal meaning
Conflict
NO START
Action begins when approach pressure reliably exceeds avoidance pressure and the system can sustain the transition.
Avoidance pressure
- Delayed reward
- Ambiguity
- Boredom
- Emotional discomfort
- Uncertain duration
- Risk of failure
- Task switching cost
1
Task appears
The person intends to act.
2
Conflict is evaluated
Reward, delay, effort, uncertainty, emotion, and current arousal all matter.
3
Action does not start
Nothing starts. You avoid, switch tasks, or sit with it open and untouched.
4
The task changes emotionally
Guilt arrives. The task gets heavier. You can see it happening. You cannot stop it.
5
The next attempt starts from a worse state
The collision compounds.
6
Something shifts the balance
Urgency, novelty, accountability, medication, or a tiny first step can push the system across threshold.
Interactive model
Move the variables and watch the state change.
This is a conceptual simulator, not a diagnostic tool. Its purpose is to make the theory intuitive.
Try «invoice» then «medication support» — the task did not change; the executive support did.
Approach pressure0
Avoidance pressure0
Illustrative state
INTENTION COLLISION / STALL
Single-moment conceptual calculator, illustrative only — not a validated clinical model, and not a simulation of the recursive dynamic described elsewhere on this page. Solid green line: approach pressure. Dashed pink line: avoidance pressure. An orange marker appears only when approach genuinely crosses avoidance during the shown window.
Evidence map
What is known, what is plausible, and what is still ours to test.
Intention collision is a proposed synthesis, not a validated theory. The components below are established or emerging in the literature; the recursive dynamic combining them is the claim this project makes and is willing to have tested.
Strong evidence
ADHD and delay discounting
A meta-analysis of 21 investigations and 25 case-control comparisons found significantly steeper monetary delay discounting in ADHD, with a medium pooled effect.
Strong evidence
Altered reward anticipation
Meta-analytic fMRI work has found medium-sized ventral striatal hyporesponsiveness during reward anticipation in ADHD.
Moderate evidence
Procrastination and ADHD
Systematic reviews have found a consistent positive association between ADHD symptoms and procrastination, with inattention the most robust correlate. Executive, emotional, and motivational mechanisms are implicated.
Emerging evidence
Avoidance of mental effort
Research suggests avoidance motivation and response inhibition may interact in ADHD, but the literature on subjective effort remains sparse and mixed.
Strong evidence
Catecholamines and executive control
Dopamine and noradrenaline modulate prefrontal cortical function. ADHD medications act in part by changing catecholamine signalling in networks important for attention and executive control.
What this builds on
Prior art the model rests on
Approach-avoidance conflict is a long-established construct (Lewin; Miller, 1940s), with gradient models and decades of animal and human evidence. Temporal Motivation Theory (Steel and König, 2006) formalises how delay and impulsiveness erode motivational value. Sirois and Pychyl (2013) model procrastination as short-term mood repair: avoidance relieves present discomfort at the cost of future discomfort, making the next approach harder.
What intention collision adds is not the conflict, and not the emotional loop. It is the claim that the two are the same system, that the collision state is a distinct and recognisable phenomenology rather than a failure of will, and that it is specifically state-transitional: the same person, same task, same intelligence, crosses the threshold when a term changes.
Our synthesis
Recursive intention collision
The specific claim that non-initiation recursively increases aversive task value until urgency or another factor causes a state transition is built from established components. Its narrower value is not that the loop exists — Sirois and Pychyl have already shown that — but that the loop and the approach-avoidance conflict describe one system, and that the resolvable moment when a state term changes is the useful clinical and personal target.
Why deadlines can feel magical
The task did not get easier. The state changed.
A deadline can collapse ambiguity, make future consequences immediate, increase arousal, reduce competing options, and make the next action obvious. In the model, multiple terms change at once.
10:00
“I cannot start.”
Reward distant. Task ambiguous. Urgency low. Alternatives available.
16:15
“OH SHIT.”
Urgency immediate. Consequence salient. Decision space collapses. Action wins.
Same person.
Same intelligence.
Same task.
The collision resolved.Medication rabbit hole
“I did not feel motivated. I just did it.”
Treatment doesn't necessarily make the task appealing. It may just lower the threshold between intention and action — the point where intention collision resolves.
People often describe it not as feeling better, but as simply watching themselves begin.
Goal maintenance
The intended task stays in view instead of being crowded out by everything else.
Signal quality
Catecholamine modulation can improve the signal-to-noise characteristics of executive networks within an optimal range.
Action threshold
Our hypothesis is that better executive support may reduce the gap between “I should” and “I have begun,” without requiring an emotional feeling of motivation.
Important: This is an explanatory model, not medication advice. ADHD treatment is individual. Talk to your doctor, not a website.
A theory must make predictions
If the model is useful, we should be able to break it.
The strongest version of this project is not “this feels right.” It is a model that generates measurable predictions.
Time to first action should respond dramatically to context, not just task difficulty.
Changing urgency, clarity, accountability, or reward timing should alter time-to-first-action even when task difficulty stays constant — the intention collision resolves when a term changes, not when the person tries harder.
Task difficulty and activation cost should separate.
A short boring administrative task can take longer to initiate than a difficult but interesting task.
The collision state should be behaviourally distinguishable from both low motivation and simple avoidance.
A person in intention collision should show high approach indicators — checking the task, opening it, orienting toward it, elevated arousal — simultaneously with non-initiation. Low motivation predicts low approach indicators. Simple avoidance predicts orienting away. If people who report "I cannot start" show the same behavioural signature as people who report "I do not want to," the model is wrong.
Reducing the first action should have disproportionate effects.
Making only the first step smaller and clearer should cut starting time by far more than its tiny reduction in total work would explain.
Urgency should create non-linear transitions.
Performance may not improve smoothly. There may be a threshold where behaviour changes abruptly from avoidance to sustained action.
Scientific boundaries
What this project does not claim.
It does not claim ADHD is simply “low dopamine.” Dopamine effects depend on circuit, receptor, timing, and context.
It does not claim intention collision is literally a primitive fear freeze response. The similarity is in how it feels from inside, not necessarily in mechanism.
It does not claim one mechanism explains every person with ADHD. ADHD is heterogeneous, and intention collision is neither exclusive to ADHD nor universal within it.
It does not claim to distinguish intention collision from initiation difficulty in depression, autism, or anxiety. Anergia in depression, autistic inertia, and anxiety-driven avoidance can all present as "I want to and cannot begin". That differentiation is an open question.
It does not claim the Intention Collision Model is validated. It is an evidence-informed synthesis designed to generate better questions and testable predictions.
The series
Eight rabbit holes, one question.
Why can’t I just start?
Name intention collision. Separate wanting from doing.
Dopamine is not motivation.
Reward prediction, salience, and action selection.
The strange economics of ADHD.
Delay discounting, uncertainty, reward timing, and competing value.
Avoidance is not laziness.
Negative reinforcement, emotional cost, and executive control.
Why deadlines suddenly work.
Arousal, urgency, state regulation, and threshold effects.
Why medication can make starting ordinary.
Catecholamines, prefrontal control, and the intention-action gap.
The Intention Collision Model.
Integrate the components into a recursive dynamic system.
Can we prove it?
Experiments, falsification, competing explanations, and limitations.
Selected evidence base
Start here.
These sources support components of the model. They do not, by themselves, validate the full Intention Collision Model.
Jackson & MacKillop
ADHD and monetary delay discounting: meta-analysis of case-control studies.
Plichta & Scheres
Meta-analytic review of ventral striatal responsiveness during reward anticipation in ADHD.
Effort in ADHD scoping review
A systematic mapping of the sparse and mixed literature on subjective mental effort in ADHD.
Orhan, Corr & Krupić
ADHD, avoidance of mental effort, response inhibition, and avoidance motivation.
Suriano, 2026
Systematic review of ADHD and procrastination.
Arnsten
Catecholamine influences on prefrontal cortical function and relevance to ADHD treatment.
The working hypothesis
Intention collision is not a failure of will. It's a system waiting for a threshold that other brains clear without noticing.
The next job is to turn that idea into a sharper model, compare it against competing explanations, and design tests that could prove it wrong.