Shark Incident Classifications: What “Unprovoked,” “Provoked” and Other Labels Mean

Shark Incident Classifications: What the Labels Really Mean

Shark incident statistics can look simple until two organizations publish different totals for the same year. Often the disagreement is not an error. It comes from differences in definitions, evidence and the way unusual cases are classified.

That is why Shark Incidents does not place every news report into a single “attack” bucket. A bite involving a swimmer in open water is not the same type of event as a person being bitten while handling a hooked shark. A shark striking a boat is different again. And some widely reported cases remain uncertain because the available evidence does not establish that a shark caused the injury.

Understanding those distinctions is essential before comparing countries, years or species.

Unprovoked incidents

An unprovoked incident generally describes a shark biting a living human in the shark’s natural habitat without the person first attempting to handle, capture, feed or otherwise provoke the animal.

The label does not mean the shark was malicious, hunting a person, or acting without any environmental trigger. It is a classification of the human–shark interaction, not a judgment about the animal’s intention.

This category is commonly used in annual statistics because it provides a more consistent basis for examining encounters that arise during ordinary water use such as swimming, wading and board sports.

Provoked incidents

A provoked incident occurs when a person initiates an interaction that materially changes the circumstances. Examples can include attempting to catch or unhook a shark, spearfishing interactions, feeding a shark, handling a captured animal or deliberately touching or harassing one.

This distinction matters. Combining provoked and unprovoked events into one headline number can make a location appear more hazardous to ordinary beach users than the evidence supports.

It can also hide important information for fishers, divers and wildlife professionals, who may face a different type of exposure than swimmers.

Watercraft incidents

Sharks sometimes bite, bump or ram boats, kayaks, paddlecraft or other vessels. These events may be scientifically interesting, but they are not automatically human bite incidents.

Shark Incidents stores watercraft events as a separate classification when the evidence supports it. If a person is also bitten during the event, the record should make that distinction clear rather than forcing a complicated event into an inaccurate single label.

Air/sea-disaster and scavenging cases

Historical databases may include events associated with shipwrecks, aircraft accidents or other disasters at sea. These can be difficult to interpret because the sequence of events is often uncertain.

Likewise, there are cases in which shark scavenging occurs after a person has already died from another cause. Those records must not be presented as though a shark caused the death unless the evidence supports that conclusion.

Questionable or disputed cases

Some cases begin with a dramatic headline and become less certain as investigators review the evidence. Wounds may have another cause. A body may not be recovered. Witnesses may disagree. The species may be guessed from location rather than identified.

A responsible database needs a place for uncertainty.

On Shark Incidents, a questionable or disputed record should clearly state what is known, what is uncertain and which sources support each point. An incident should never be upgraded to “verified” merely because many websites repeat the same original report.

Classification is separate from verification

A useful way to think about the database is that every incident has two different questions:

1. What kind of incident was it? 2. How certain are we that the reported details are correct?

For example, a new media report may appear to describe an unprovoked bite, but the event can remain in “Reported” verification status until more evidence becomes available. Later reporting might confirm the event but change the species from “reported tiger shark” to “species unknown.”

The classification and the confidence level should therefore be stored separately.

Why totals differ between databases

Different projects may use different inclusion rules, investigate cases at different speeds, or revise older records when new evidence appears. One project may count only confirmed unprovoked bites in its headline statistics; another may show all reported shark-human incidents on a map.

Neither number is meaningful without its methodology.

Whenever Shark Incidents displays a total, the page must state the scope: the included classifications, verification statuses and date through which the database was reviewed.

The practical rule for readers

Do not judge shark risk from a headline count alone. First ask what the number includes.

A well-labeled map is more useful than a larger but ambiguous total. That is why every public record on this site displays both its incident classification and its verification status, with source links so readers can examine the evidence themselves.

For a complete explanation of how records enter and change in this database, see the [Methodology](/methodology.html) page.

---