You have replayed it already. The round you did not close early enough. The hire you held on to for two quarters too long. The pivot you considered in month eight and set aside. Each one runs with the same opening: if only.
This is upward counterfactual thinking. You are comparing what happened to a better version of events that did not occur. It feels productive because it looks like a post-mortem. You are tracing causes, identifying leverage points, running the board on what you could have done differently. Research on how the mind generates and uses counterfactuals shows it is doing something quite different from analysis, and understanding what that is matters for how the next few months go.
What counterfactual thinking is and what it does
A counterfactual is a mental simulation of an alternative to reality. The mind generates them automatically after unexpected or negative outcomes. Psychologists draw a line between two types. Upward counterfactuals compare what happened to something better: if only I had moved faster, if only I had read the market signal in Q2. Downward counterfactuals compare what happened to something worse: at least the team found roles, at least I had not taken on personal debt to extend runway.
The distinction sounds minor. The measured effects are not. Upward counterfactuals consistently increase regret and negative affect, and drive rumination: the mind returning to the same material without generating new information. Downward counterfactuals reduce emotional distress and improve forward orientation. Neither type requires you to pretend the outcome was good. The difference is which reality you are comparing against.
Kahneman's work on the simulation heuristic adds a specific mechanism. The easier it is to mentally simulate an alternative outcome, the stronger the emotional response to comparing against it. After a shutdown, every near-miss and delayed decision sits in recent memory, available and easy to simulate. That ease amplifies the regret, not the actual closeness to a better outcome. The mind treats a vivid alternative as evidence that the better outcome was close. It was not necessarily close. It was just easy to imagine.
Why the loop feels like it is doing something
The if-only replay carries the texture of work. You are building a causal model of what happened. You are not indulging, you are learning, which is what you are supposed to do after a failure. The problem is: the useful information, if any exists, arrives in the first few passes through the events. What continues beyond that is not learning. It is the emotional machinery running on the same material without new inputs.
The loop has no natural stop condition. Each if-only generates a scenario to evaluate, which produces another counterfactual, which continues the cycle. The mind does not signal when the analysis is complete because there is no threshold to meet. It signals when the material feels resolved, and rumination tends to keep the material feeling unresolved by design.
There is a second effect worth naming. Repeated upward comparison across enough events gradually shapes something in how you explain things to yourself. Each if-only is a version of: if I had been different, this would not have happened. Applied broadly and repeatedly, that becomes a stable, global attribution: I am the kind of person this happens to. That is not the lesson of a single shutdown. It is what the loop turns a single shutdown into, given enough iterations.
The direction matters more than the content
Here is the part the research makes actionable. You cannot stop the replay. The mind returns to events that are unresolved, high-stakes, and recent. That is not a choice, it is how memory handles material it has not yet filed away. What you can change is whether the replay runs only upward or also downward.
For every if-only pass, generate one at-least. Not as a consolation or a reframe. As a second counterfactual of the opposite type, run through the same event. At least the product shipped and was actually used by real people. At least the market timing problem is legible and external, not a personal capability failure. At least you are out now rather than two years from now with more sunk cost. At least the people you worked with know what you can do.
The goal is not to balance a ledger or cancel the regret. You are using the same mechanism, counterfactual comparison, in a direction that does not anchor the mind in the loss. Downward counterfactuals work because they give the simulation machinery a different baseline. The loss is the same. The comparison shifts.
What this is not
This is not a call to forced optimism. Pushing positive framing over a real loss tends to suppress processing rather than complete it, and founders who have read enough self-help material will recognize the move and correctly distrust it. The distinction here is mechanical, not motivational. You are choosing which type of comparison runs alongside the upward one, not choosing to feel fine about what happened.
You may encounter the claim that the right mindset would have produced a different outcome, that belief or mental state controls external results. That is not what this research addresses. The research here is about what the mind does with events after the fact, specifically: which processing patterns tend to keep someone stuck and which ones tend to produce forward movement. The outcome is already fixed. What is still open is how the next few months go.
One thing to do with this
After the next replay, before closing the loop, note the if-only, then ask for the at-least version of the same event. Write it down if you need the friction to slow the cycle. You do not need to believe the at-least statement immediately or find it convincing. You are building the habit of running both directions, not producing a revised verdict on the shutdown.
Over time the downward comparison becomes available without forcing. The upward replay still runs. The difference is it no longer has the space entirely to itself, and the mind gradually has more material to work with than the single baseline of a better outcome that did not happen.