February 8, 2017 at 3:28 p.m.
Patrick Durkin and the Ostrich Effect
Patrick Durkin and the Ostrich Effect
Yep, it's true, at least apparently in the eyes of Patrick Durkin, an outdoors writer whose columns are published widely in Wisconsin. After many in the Northwoods questioned the DNR's reported numbers for the 2016 gun deer season - specifically a 30-percent increase in the northern buck harvest - Durkin proclaimed the skeptics to be members of the Tinfoil Hat Club.
"Sigh," he wrote. "Here we go again. Get out your tinfoil hats and tune in IKOD (I'm Kooky Over Deer) Radio for Wisconsin's latest rumors, buck pellets and deer-hunting conspiracy theories."
The term 'tinfoil hat' comes from an urban legend that wearing a hat made from tinfoil would shield a person from aliens or from government surveillance. Over the years it became a mocking term for those who believed in so-called conspiracy theories about a faked lunar landing in 1969, aliens at Area 51, and other imaginative plots and schemes.
More recently, the label has been attached to anybody who questions government authority. In other words, the tinfoil hat is worn not merely by those who believe in a specific and quite unbelievable government plot, such as a fake moon landing, but by those who cast doubt on any government narrative.
What the government says must be true by definition, and those who don't believe it are tinfoil hatters.
For example, if you are skeptical about the significance of manmade climate change, you are a tinfoil hatter. If the federal government says there are weapons of mass destruction, then there are weapons of mass destruction, and those who doubt it, no matter how many there might be, are wearing tinfoil hats.
This is where Durkin comes in. For him, if the DNR says the buck harvest is up in northern Wisconsin, then the buck harvest is up, no matter how many people contest it or what evidence casts that narrative in doubt.
The numbers that Durkin tells us are unequivocally true show a 30-percent increase in the northern forest zone's buck harvest, even though in other regions, the antlered harvest numbers were stable, within about 3 percent or less of last year's total.
Overall, the deer kill was down from 2015; the buck harvest was up, but only because of the skyrocketing numbers reported by the DNR for the North
All of that seems odd and incongruous, given that the northern Wisconsin numbers don't match reports from the field - reports that are bolstered by actual correlative evidence - and because the DNR's reported herd population trends were no better in northern Wisconsin than elsewhere, where the harvest was weak.
Logic tells you that those areas with more deer will have more deer harvested per square mile come hunting season, and that's exactly what DNR numbers show. Following the same logic, the trend line for those population numbers should tell the tale on increasing or decreasing deer harvests.
They do, mostly, elsewhere in the state, but they don't in northern Wisconsin, and that's a red flag.
According to the agency's White-tailed Deer Population Status report for 2015, after a severe winter and a late spring, northern Wisconsin deer numbers were down by 2014 to 250,000, but "may have stabilized" through limited harvests in 2014 and 2015 and a milder winter in 2014-15.
Still, the agency, the agency cautioned going into this year, "hunters can expect that this rebuilding will still take some time, especially in northern tier counties."
So at the very least, the starting numbers would logically project a stable harvest, maybe a slight increase, but not one increasing by an astounding 30 percent. The DNR population numbers, its projections based on those numbers, and the reported kill just aren't compatible; something is wrong, either on the front end or the back end.
None of this bothered Durkin, even though numbers everywhere else more closely conformed to expectations. In other regions, most were stable or differed only slightly from the forecasts and from 2015.
More important, when there were significant departures from the forecast, rather than exceeding the projection, the harvest did not meet DNR expectations, and there was nothing remotely close to a double-digit differential between the harvest this past season and the kill in 2015, except in one instance, and that was 10 percent, not 30 percent.
In the Central Forest Zone, for instance, the population has "largely stabilized since 2009," the DNR reported in 2015. That and two mild winters led the agency to expect "deer hunting opportunities throughout the Central Forest Zone to be improved from 2015."
Based on those sets of facts, one would expect the harvest to be either stable or increased; in reality, it was stable but down by 2 percent.
In the Southern Farmland Zone, population growth has been strong, the DNR reported in 2015, and the agency produced a rosy hunting outlook for 2016: "Many counties within either the Central or Southern Farmland Zones have deer populations that are above goal, and hunters will find great opportunities to put some meat in the freezer this fall if they are able to spend some time scouting and find areas that deer frequent."
It didn't happen.
In the Central Farmland zone, the buck harvest inched upward by 1.4 percent, but antlerless fell by 8.2 percent. In the Southern Farmland zone, the buck harvest declined by 3.4 percent.
Put it all together, the northern forest zone number - compared to the forecast, compared to the 2015 harvest, compared to results in the rest of the state - was an outlier whose legitimacy and accuracy should at least be questioned.
Not Durkin, though. So afraid he was of tinfoil hats and any bad information that might discredit the official line that he stuck his head in the sand and refused to see any flags, red or otherwise. It's known as the Ostrich Effect, and Ostrich Durkin had his face deep in the dirt, facetiming, no doubt, with Mother Earth.
Add to everything else horrible weather conditions for the season opening in the North- high winds and bitter cold - and you have to question why the North would be the only region to show an increase, and a startling increase at that. Why would the projected harvest be so far outside any reasonable margin of error, and be the only one outside that margin?
Those are questions any worthy journalist should ask before even assessing reports from the field. Not Durkin. Ostrich Durkin had his head in the sand.
Oddly, for a so-called journalist, Durkin admitted his was a knee-jerk answer, not based on anything more than what he has seen in the past and his own rigid ideological regurgitation.
"Why discount this rumor by reflex?" he asked. "Call me jaded, but I've been covering our deer seasons since 1982. I've lost count of all the childish, ridiculous tales and accusations swirling around the state's No. 1 outdoor activity."
Durkin didn't indicate whether he had ever investigated any of those past accusations before concluding they were ridiculous and before asserting that the current concerns were false simply because past concerns might also have been false.
By Durkin's logic, a man cleared of one murder must be declared not guilty of any other murders for which he might be later charged. Why even have a trial?
Durkin's reasoning isn't the mark of a jaded man; it's the mark of a charlatan who attempts to hoodwink people by conflating the past and the present, by attempting to delegitimize present concerns by linking them to past concerns that he - acting as judge and jury - declares outrageous on their face.
Nowhere is there any evidence for his case; Durkin is simply predisposed to making the facts fit his conclusions. Such is the recipe of fake news.
To be sure, Durkin pretends to perform an analysis of the numbers, hard as that might be with so much sand scraping his corneas. The problem is, when you already know your outcome, you're prone to run numbers that lead to that outcome, which often are the wrong or irrelevant numbers to run. And that's what Durkin led with.
First, he looked at Waupaca County and found that the "county's 2016 kills with e-registration differed little from its 2014 kills with in-person registration, specifically, 1 percent higher for bucks, 1.8 percent higher for antlerless deer, and 1.2 percent higher overall."
With such scant evidence of registration rigging or deer-harvest fraud, Durkin wrote, a recount was hardly needed.
Of course, Waupaca was not where the alleged 30-percent increase occurred, and hardly anyone was questioning the numbers in Waupaca County - Durkin wrote his column apparently because he received one email from a reader there - and the final numbers were just about what the DNR expected: some increase and good opportunities.
On the other hand, many people were questioning the numbers in the northern forest zone, but, hey, why let geographic accuracy get in the way of setting up the straw man?
Sure, after setting up the red herring, Durkin got around to the northern Wisconsin numbers, but this is where he really jumped the deer, ur, shark. By now his head - and his analysis - were about six feet under.
First, he argued, the deer kills in 2016 season were well within a five-year average, with the 2016 gun-buck kill only 7 percent higher than the five-year average.
"Such differences are insignificant when spread across a zone that covers Wisconsin's northern third, roughly Highway 64 to the Upper Peninsula and Lake Superior," he wrote.
Well, that's utter nonsense, an analysis that would embarrass a serious columnist or statistician.
First, 7 percent is statistically quite high, as statistical models go.
Second, to diminish statistical significance as the land area increases is bunk because percentage expressions are proportional expressions, not absolute ones. An absolute increase of, say, 15 deer spread out over 18 counties might be insignificant compared to an absolute increase of 15 deer in one town, to use an example; but 7 percent is just as significant spread out over 18 counties as it is in one.
Here Durkin is standing the truth on its head, for in determining statistical significance, it matters not what the population size is in measuring a particular characteristic within that population; it matters what the sample size is - the larger the sample, the more accurate it is.
To wit, an increase of 7 percent in car fatalities in the U.S. wouldn't be insignificant just because the U.S. is a big place; but trying to use the number of Wisconsin fatalities to determine the national rate would indeed be a useless exercise.
In this case, the entire population in the region is being measured; the sample is 100 percent. A 7-percent differential could conceivably be a one-year fluke, but it is anything but insignificant.
More fatal to Durkin's analysis, what stands out as more statistically significant is that the deer population in the northern zone has declined by about 100,000 since 2010. That's important because when Durkin points to a 7-percent harvest increase over the five-year average without considering the population change, he is assuming a static population across the years; he is not factoring in the declining herd from which those harvest numbers emerged.
The key measurement should be the five-year average harvest as a percent of the five-year average population. Doing the math, the five-year average buck harvest has totaled about 8 percent of the total population.
To have any comparative meaning, the 2016 buck harvest should be first compared not to the five-year harvest average but to the total 2016 population from which it was harvested. When that calculation is made, it totals about 11 percent of the herd.
That's a huge 37.5-percent increase in the proportion of the total herd taken by the 2016 buck harvest compared to the previous five-year average of 8 percent, if you believe the DNR's numbers. It's not only an outlier compared to the previous five years in the northern forest zone but it is an outlier elsewhere, where the buck kills compared to total populations are also usually below 10 percent - for instance, this year's gun-buck kill was 8 percent of the estimated herd in the central forest zone, according to DNR numbers.
So while it is possible that a one-year fluke occurred and that a monstrous percentage of bucks was taken from a smaller-than-five-year average population, it's historically high, it's an outlier, it's hard to believe, and it needs to be questioned.
Next: Reports from the field
Comments:
You must login to comment.