State wildlife departments release a huge amount of deer data gathered honestly and with care. Yet those figures mislead readers again and again because context disappears as soon as a number lands in a headline, caption or chart. The confusion comes from that missing context rather than deception. Readers take a partial count to cover everything. They treat a management target as a measured fact. A season gets labeled a year off. This practical field guide covers the five traps we run into most often, with a real example readers can reproduce and one question that clears up each problem. It aims to help readers understand the numbers while keeping the context that gives them meaning.
KEY FINDINGS
- A headline can cover just part of the harvest. The often-quoted North Dakota count covers deer killed during gun season alone: 20,877 in 2025. Once bow and muzzleloader kills are included, official figures from North Dakota Game and Fish Department put the statewide tally across seasons near 32,000, a substantially larger total. The gun-season headline leaves roughly a third of statewide harvest uncounted.
- Targets sound exactly like measured results. At Camp Atterbury in Indiana, '25 deer per square mile' describes a deer-management objective. The Flint Hills refuge in Kansas has a measured density of '23 per square mile,' an almost identical figure. Matching units do not give these figures the same meaning.
- A crash tally alone does not tell you deer density. Clayton, Iowa's leading county for deer harvest (42,354 deer taken), recorded 1,401 animal-vehicle collisions. Metropolitan Polk (10,415 deer taken) recorded 2,405. Traffic exposure is absent from this comparison.
- County harvest tallies are absent from some sources. Among the twelve states we track, five release no county harvest figures. North Dakota, South Dakota, Kansas and Nebraska derive their harvest estimates from hunter surveys. Minnesota uses deer registration but releases no county breakdown. Any county harvest number assigned to those states is made up rather than obtained from a source.
ITRAP ONE · THE SUBSET HEADLINE
The count is accurate, but it covers only a portion
Deer statistics most often mislead when an accurate count for part of a total is presented as the whole. North Dakota makes the problem easy to see. The widely repeated harvest number counts deer taken during gun season: our record shows 20,877 for the 2025 season. Hunters also kill deer with bows and muzzleloaders there. The North Dakota Game and Fish Department's total across all seasons is nearer 32,000. No one is lying, and the gun-season count is exact. But reading that number as 'North Dakota's deer harvest' leaves roughly a third of the harvest out. Headlines often give a partial count a broader label: 'buck harvest' becomes 'harvest,' a 'firearm' count becomes a season total, or a 'reported' count becomes an estimate.
What to check: ask what is being totaled and which weapons and seasons are included before accepting the figure. Without that scope, a harvest statistic tells only half the story.
IITRAP TWO · THE MISLABELED SEASON
A caption's date does not always match the harvest year
A deer season spans calendar years, beginning in fall and ending in winter. Agencies therefore organize harvest records around the season year. Reports bearing a single year's date regularly cover kills made across two calendar years. A file marked '2025' may contain a season spanning 2024 to 25. Using the publication year in a chart caption instead of the season year quietly moves every data point to a different year. Filenames have the same problem, as do datasets replaced at the same location. Minnesota overwrites its individual-deer harvest interface annually, according to our acquisition notes. Unless those records are downloaded after a season ends, that season's per-deer information is lost. The date in a caption identifies the season a figure is supposed to represent. Those dates are wrong often enough that they need checking.
What to check: establish whether the date refers to the hunting season or the report's release. Do that before lining up any series from two agencies. Incorrect year labels can make separate seasons appear to match.
IIITRAP THREE · THE GOAL IN A MEASUREMENT'S CLOTHES
A planned deer density looks just like an observed figure in print
Even careful readers fall for this because the figures are valid and share the same presentation. A management plan states a goal for how many deer each square mile should hold, the level managers aim to reach. A field study uses those units for a measurement, the density researchers actually observed. Without the context supplied by a single word, either type could carry the label '25 deer per square mile.' Our property layer includes both kinds. Camp Atterbury's management objective is 25 deer for each square mile. At the Flint Hills refuge, the measured density across the entire refuge is 23 deer for each square mile. Close as they are, these figures represent a target and an observed density. Taking their average, or putting them on the same map, mixes what managers want with what was observed.
Swipe table to see every column
| Property | Figure | What it actually is |
|---|---|---|
| Camp Atterbury (IN) | 25 deer / sq mi | Management OBJECTIVE, a target |
| Flint Hills NWR (KS) | 23 deer / sq mi | MEASURED density, an observation |
| Glendale (IN) | 17.9 → 10.3 / sq mi | MEASURED pre/post-harvest, with CI |
An observed result can equal a management target, with both expressed in identical units. The target gives the desired density; the study gives the reported density. Do not combine them into an average.
What to check: look for the verb attached to a population or density figure. 'Maintain,' 'goal,' 'target,' and 'objective' signal intended management levels. 'Measured,' 'surveyed,' 'observed,' and 'estimated' signal estimates or observations. Estimating does not produce an exact count. When the document fails to distinguish these, consider the figure unusable.
IVTRAP FOUR · THE CRASH MAP
Deer density cannot be read from crash totals alone
People turn county collision totals into maps of deer density, even though those totals map traffic. A crash requires both a deer and a vehicle. Our analysis of Iowa found only a weak correlation (r = 0.31) between county collision counts and county deer harvest. Traffic volume is a more likely explanation, although traffic was not measured for this page. Clayton County had Iowa's highest deer harvest, with 42,354 taken during 2015 to 2024 yet recorded only 1,401 animal-vehicle collisions. Metropolitan Polk County took about a quarter as many deer (10,415) and recorded 2,405 collisions. Sorting Iowa counties by crashes produces a list of cities. Sorting by harvest points to deer country. Our companion paper about crash maps examines this problem separately.
What to check: animal-vehicle collision totals tell you where cars hit animals. Before drawing any conclusion about deer from a crash map, require a rate adjusted for exposure, with collisions expressed per vehicle-mile. Also establish whether the tally covers deer specifically or all animals hit on roads.
VTRAP FIVE · THE COUNTY THAT DOESN'T EXIST
County harvest figures in survey states are made up rather than measured
Within the twelve states we track, seven report harvest by county for registered deer. County figures are absent in five. Michigan, Ohio, Indiana, Wisconsin, Missouri, Iowa and Illinois make deer reporting mandatory for hunters, so the harvest county is recorded. North Dakota, South Dakota, Nebraska and Kansas instead estimate the take with hunter surveys conducted after the season, releasing results for management units or the whole state. Minnesota registers deer without publishing a county tally. When a map colors individual counties in those five states, its county pattern reflects an assumption about splitting a unit-level estimate into county shares rather than a measurement. Our companion paper about the two-tier map explains this fully.
What to check: find out whether a state uses deer registration or hunter surveys before accepting its county harvest tally. In a survey system, no county measurement was taken.
VITHE FIELD CHECKLIST
Before accepting a deer figure, work through five questions
Check these in sequence
- Scope: what is being totaled, and which seasons and weapons does it cover? The most common mistake is a headline that makes a partial count sound like the whole.
- Season year or release year: does the date identify when the deer were taken or when the report came out? Check that before lining up any two data series.
- Target or observed result: check the verb used. Words describing intent ('target,' 'objective') differ from those describing observations ('measured,' 'estimated').
- Count or rate: has the denominator been left out? If traffic levels vary, use collisions per vehicle-mile for comparison instead of total crashes.
- Reporting geography: is the area actually reported by the source, or has someone created that breakdown? Establish whether harvest is available at the county level or just for management units.
Good data and honest reporting can still lead to these problems when a figure moves a single step beyond the context supplied with it. The five questions bring back the background readers need to assess a deer statistic. Even sound data calls for careful interpretation.
NOTES AND SOURCES
- The statewide harvest record provides North Dakota's gun-season count (20,877, 2025). North Dakota Game and Fish Department combines gun, archery and muzzleloader kills in its published ~32,150 all-season total.
- Public-hunting management documents from federal and state agencies supply an objective of 25 deer/sq mi for Camp Atterbury and a measured density of 23 deer/sq mi for Flint Hills National Wildlife Refuge. The distinction between target and measurement matches the published documents.
- Iowa DNR records and records from Iowa DOT supply the figures comparing Clayton with Polk for harvest and animal-vehicle crashes (2015 to 2024). The comparison is covered by 'Why Deer-Crash Maps Lie.' Details of the tier split appear in 'The Two-Tier Map.' These are companion papers.
- Limits: these examples show common ways figures can mislead. Each agency's numbers come with their own cautions about sampling and reporting. All figures are snapshots for the seasons labeled 2024 to 2025 and the 2026 source collection.
Fully Drawn Research is an independent data analysis desk, not affiliated with any state wildlife agency, transportation agency, mapping provider or outfitter. Fully Drawn Research is an independent data analysis desk, not affiliated with any state or federal wildlife agency. Figures are reproduced from public agency records except the North Dakota all-season total, cited from NDGF. Crash counts are treated as traffic-exposure indices and survey-state harvest is never apportioned to counties. Figures current to the 2024–2025 seasons.
Keep reading
FDR Scout puts research like this on the map of the ground you hunt. Open Scout →