Overview

Imagine waking at 3 a.m. with a vivid sense that your brother is standing in the doorway. The next morning you learn he died in an accident at that exact hour. This is the core pattern of a crisis coincidence report. The experience is spontaneous: nobody planned it or set up an experiment. It arrives without warning, often with strong emotional force. And its timing appears to match a real crisis happening to someone the person knows.

What makes these reports scientifically interesting is not any single case. Any one story could be a coincidence, a false memory, or a dream misremembered after the fact. What drew researchers’ attention was the sheer number of such reports. The key question is whether their timing can be explained by chance alone. Deaths are common, and people have vivid experiences every night. Some overlap is inevitable. The question is whether the overlap happens more often than probability would predict.

The First Large-Scale Investigations

The systematic study of crisis coincidences began with the Society for Psychical Research (SPR) in London, founded in 1882. The SPR’s early researchers knew that individual stories prove nothing. They needed a large, structured collection of cases. They also needed independent checks on both the experience and the distant event. Two projects defined this era.

The first was Phantasms of the Living, published in 1886 by Edmund Gurney, Frederic Myers, and Frank Podmore. The book gathered hundreds of cases that people reported on their own, outside any experimental setting. The authors did not simply accept stories at face value. They wrote to witnesses, sought testimony from others who had been told of the experience before the news of the crisis arrived, and tried to find independent records of the distant event. Reviewers noted the extraordinary care the authors took in questioning witnesses.¹¹

The second project was the 1894 Census of Hallucinations. It surveyed roughly 17,000 people across several countries. Participants were asked whether they had ever, while awake and in good health, had a vivid impression of seeing, hearing, or being touched by a person who was not physically present. Among those who said yes, researchers looked for cases where the impression matched a death or crisis involving the person perceived. The committee concluded that the number of death-coincident cases was not due to chance.⁵¹

What the Experiences Look Like

Crisis coincidences are not a single type of experience. They arrive in several forms. The most dramatic are visual: a person sees a lifelike figure of the distant individual, sometimes in ordinary clothing, sometimes looking distressed. But hearing sounds is also common. So is a sudden overwhelming sense of the person’s presence, or a vivid dream that feels very different from ordinary dreaming.

Later analysis of large case collections found that visual hallucinations make up only about a quarter of all hallucinatory psi experiences. Auditory, tactile, and other forms account for the rest.¹⁸ Early researchers, influenced by the dramatic visual cases in Phantasms, may have overestimated how often crisis experiences take visual form. The broader picture is more varied.

One recurring question is who, if anyone, is doing the communicating. The traditional picture imagines the dying person somehow reaching out to the other person at the moment of crisis. But a careful analysis of 208 visual hallucination cases pointed to a different possibility. The person having the experience, not the distant person in crisis, may be the one who starts it. In case after case, the apparition appeared in the experiencer’s own familiar environment. It showed up in places where they might naturally expect to see that person. It did not appear in settings that matched where the distant person actually was at the time of crisis.⁵

The Statistical Argument

The central scientific claim about crisis coincidences is not that any single case is convincing. It is that the timing of these experiences, across a large collection, departs from what chance would produce. To evaluate this claim, you need two things. First, a reliable count of how often the experiences occur at the moment of a distant death. Second, a reliable estimate of how often that overlap would happen by accident.

The second number is harder to calculate than it sounds. Deaths are not rare events. In a large population, many people die every day. Anyone who has a vivid experience involving a friend or relative has some chance of that experience accidentally overlapping with a death. This is called the base rate problem. You have to know how common the coincidence would be by chance before you can say whether the observed rate is surprising.

The Census committee calculated that only about one apparition in 19,000 should coincide with a death by chance. The observed rate in the Census data was approximately one in 48, roughly 400 times higher than the chance expectation.²³⁸

The SPR researchers were aware of this problem and tried to address it directly. They estimated the annual death rate, calculated how many waking hallucinations the average person might have in a year, and worked out what fraction of those hallucinations should fall within a twelve-hour window of any given person’s death purely by accident. The resulting baseline was very low. That made the observed clustering look highly significant.¹ Whether that baseline was correctly calculated remains a point of dispute.

Skeptical Critiques and Discussion

Critique 1: Selection bias: people remember hits and forget misses

This critique is a standard concern in evaluating collections of spontaneous cases and is discussed in the context of survival research methods.⁴⁹

The argument runs as follows. People have vivid experiences involving distant friends and relatives fairly often. When nothing bad has happened, the experience is forgotten within days. When a death or crisis follows, the experience is remembered vividly and reported. The result is a collection that looks full of meaningful coincidences. In reality it is just the memorable subset of a much larger pool of unremarkable misses. The SPR researchers tried to counter this by requiring prior testimony. The person had to have told someone about the experience before learning of the crisis. This requirement does not eliminate the problem entirely. But it substantially reduces it for the cases where it was applied.¹¹

Evidence strength: strong (as a concern); partially addressed by prior-testimony requirement in the best cases

Critique 2: Base-rate neglect: deaths are common, so coincidences are inevitable

This concern is central to any statistical evaluation of spontaneous case collections and is raised in critical reviews of survival research.⁴⁹

If you know many people and have vivid experiences regularly, some of those experiences will coincide with deaths by pure chance. The SPR researchers acknowledged this and built a baseline calculation to account for it. The question is whether their baseline was accurate. Critics argue the assumptions underlying the calculation were optimistic in ways that inflated the apparent excess. Defenders respond that the committee made generous allowances for error and the excess remained large even after those adjustments.⁸²³ The debate has never been fully resolved because the original data cannot be re-examined with modern methods.

Evidence strength: contested

Critique 3: Retrospective embellishment: memories shift to fit the outcome

This is a well-documented feature of human memory and applies directly to spontaneous case reports.⁴⁹

Human memory is reconstructive, not like a video recording. When people learn that a friend died at a particular time, they may unconsciously adjust their recollection of when they had their unusual experience. The timing shifts to match the death. This process is not deliberate deception. It is a normal feature of how memory works. The prior-testimony requirement is the main safeguard against this problem. Cases where the experience was written down or reported to a witness before the news arrived are much harder to explain by embellishment. The SPR researchers were aware of this and noted it explicitly in their case evaluations.¹³

Evidence strength: strong (as a concern); mitigated in documented prior-testimony cases

Critique 4: Psychological predisposition: some people are simply more prone to vivid experiences

Research on nurses and other groups has found that people who report unusual experiences score higher on measures of psychological absorption and hallucination proneness.²⁶⁶²

A survey of 100 nurses found that those who reported paranormal or unusual experiences scored significantly higher on absorption and on hallucination proneness. Absorption means the tendency to become deeply immersed in mental experiences. Job stress did not differ between the two groups.²⁶ This finding suggests that individual psychological traits predict who reports crisis-type experiences. It does not by itself prove the experiences are not real. Even genuine experiences may depend on the same predisposing traits as non-genuine ones.⁶² But it does mean that people who report crisis coincidences are not a random sample of the general public.

Evidence strength: moderate

Where Things Stand Today

Crisis coincidence reports occupy an unusual position in parapsychology. They are among the oldest and most extensively documented phenomena in the field. Yet they resist the kind of controlled experimental testing that modern science demands. You cannot schedule a crisis. You cannot randomly assign people to have vivid experiences at set moments. The phenomenon is inherently spontaneous and can only be studied after the fact.

The strongest cases in the historical literature are those where the person recorded or reported the experience to a witness before learning of the distant death. These cases are harder to dismiss as embellishment or coincidence. But they are also a small fraction of the total collection. And even they cannot rule out all alternative explanations.

Modern parapsychology has largely shifted toward controlled laboratory experiments, where conditions can be standardized and results replicated. Crisis coincidences remain important as the historical foundation of the field. They were the phenomenon that originally motivated the SPR’s founding. They also serve as a reminder of what the laboratory experiments are ultimately trying to explain: the persistent human experience of knowing, without any normal means, that something has happened to someone far away.