data visualization

on geography, numeracy and truthiness...

In two recent pieces, the New York Times wrote articles on the death of a woman in Montana after being mauled by a grizzly bear. In both stories, to give readers a perspective on how rare these types of events are, the Times curiously quoted a statistic from Yellowstone National Park: since the opening of the park (in 1872), there have only been eight (8) recorded deaths due to grizzly bear attacks.

However, this most recent death from a suspected grizzly mauling occurred in Ovando MT, roughly 170 miles from the border with Yellowstone National Park. Grizzly bear ranges in the lower 48 states extend well beyond Yellowstone park boundaries.

A closer look at the broader Northern Rockies region reveals at least 30 deaths suspected from grizzly bear maulings since 1910. These deaths occurred in the Greater Glacier National Park region; the greater Bob Marshall Wilderness Complex; the Greater Absaroka-Beartooth Wilderness region, as well as the Great Yellowstone region which includes lands outside the park boundaries. I did not include deaths from Idaho, Colorado or California which occurred in the early part of the 20th century. I did not include deaths due to grizzly attacks across the northern border in Canada nor Alaska.

Most deaths have occurred outside the park boundaries. About half have occurred in Montana, in areas outside the park. This mortality statistic is a factor of 3.75 more than reported in both New York Times stories.


If spaced evenly over time, the eight Yellowstone fatalities come out to about one death per every 18 years. Obviously, that is not how these random events happen. But the low number stated immediately raised curiosity.

The NYT itself set up the geographic and time-scale boundary conditions for their story: in a correction where they restated a total count of grizzlies from an earlier version of their story to include only those bears in the Lower 48 states. In addition, this most recent event occurred far from Yellowstone’s boundaries so that fact extended the geography beyond Yellowstone National Park. Likewise, by using the fatality statistic from Yellowstone National Park, the NYT also set up the 149 year timeline boundary: from 1872 - present.

Fatalities due to being attacked by a grizzly are rare, whether we are talking 8 or 30 from a more restricted timeline from 1910 to present. Nevertheless, by setting up the greater geographic and time period boundaries in the story, then constricting the fatality counts to a relatively small subregion of the same territory, the story gives a misleading impression that mortality rates due to grizzly attacks are lower than they really are. If anything, there appears to be a cluster of deaths in the 1980s, before receding for two decades, and before more frequent fatalities began occurring again in the 2010s. Maybe there’s a story to be told there.

The word “truthiness” is a loaded term. It implies an authority is intentionally being misleading or is misdirecting without actually lying by giving only a selected or parsed version of the truth. I don’t believe for a second that the New York Times or any state or federal wildlife agency is intentionally trying to minimize grizzly deaths to tell this story. But they should be more careful with their numbers to avoid accusations of biased data or careless reporting.

Note: The data source on grizzly fatalities I used for my own post comes from Wikipedia. Normally, I dislike using Wikipedia as a data source; personal preferences. I did this today for convenience and expediency. But this listing looked to be well-researched, and if needed, easy to double-check using a search engine and government agency web sites.