Analysis of Nevada Department of Wildlife 2020 & 2021 Hunt Data

Introduction

In the state of Nevada, the management of wildlife resources requires placement of quotas on the hunting of big game species.  These quotas are created based on wildlife surveys in hunt management units covering the entire state.  An analysis of historical data shows that some units provide hunters with better opportunities than others, and that there are differences in opportunity across different species of big game.

This study is designed to assist hunters in the selection of hunt units in to order to maximize the utility of the hunt experience.  The data analyzed is representative of the previous two completed hunt cycles (2020 and 2021) across the state of Nevada and covering all “big game” hunts.  The reports are generated using Microsoft Power BI Desktop with additional work completed using Synoptic Designer for Power BI.  Also, some exploratory and visual analysis was performed using Python (specifically the Numpy, Pandas, Matplotlib and Seaborn libraries).  Original datasets are publicly available from the Nevada Department of Wildlife, and hunt unit maps are customized based on work originally distributed by HuntNV.

The use of two years of data has benefits and drawbacks.  In the case of lower-population species, it is beneficial to use as much data as possible.  However, the changes in hunt unit combinations and date ranges from year to year creates some minor inconsistencies in reporting.  The various species’ populations are migratory, so this also factors into the year-to-year differences.  These inconsistencies are minor, however, primarily resulting in inclusion of adjacent hunt units in one year but not the other.  Some hunts have been removed from the analysis due to incomplete survey data returned by hunters and/or insufficient data, but this is an insignificant amount and does not affect overall results.

Analysis:

• Tag applications far outweigh opportunities

The Nevada Department of Wildlife conducts an annual allocation of hunting tags (known as a “tag draw”), with each tag essentially representative of a single animal a hunter is legally allowed to “take”.  A maximum of one tag of each species can be allocated to each hunter as part of the process, and it is common for hunters to request tags for anywhere from one to all available species.  Once the allocation process has been completed, hunters are notified of which tags they were awarded for the upcoming season.  Hunters generally cannot request more tags until the following year.  There is a significant imbalance between supply and demand, as evidenced by figure 1 (below).  It is clearly visible that for each of the eight managed species there is a far greater count of applications for tags than there are tags to distribute (or “tag quota”). 

From figure 3, we see that the highest-volume species (mule deer, elk, and antelope) appear to have much lower success rates than do the lower-volume species (big-horn sheep, mountain goats, and bear).  One causal factor of this is that the lower-volume species tags tend to go to more experienced hunters (due to a “points” system that rewards longevity of activity in the tag draw process; hunters can accrue points over many years to increase chances of drawing scarce tags), and as such these hunters tend to devote more time and preparation to these hunts (see figure 2).  One apparent outlier is the mountain goat, which has one of the lower average “days of effort” (days of pre-hunt scouting combined with actual hunting days) at 7.2 days (only antelope had fewer days at 5.6).

• Odds of Success and Correlations

There are some clear trends in tag draw success that can be interpreted from the data.  Looking again at figure 3, the average odds of drawing a tag are far lower than the average odds of actually filling that tag. 

Interestingly, the species with the highest odds of drawing a tag (mule deer) carries the second lowest odds of filling that tag.  Still, on average a hunter historically has a higher chance of overall success when applying for a mule deer tag than with any other application.  This implies that if a hunter has limited resources, he or she can optimize overall success likelihood by focusing tag application efforts on mule deer in lieu of all others.

Additionally, figure 3 shows that there is a tendency for the least available species tags (black bear, big horn sheep, and mountain goat) to have an extremely high tag fill rate.  To better understand this difference, let us look at figure 4 and figure 5.  Figure 4 shows the relationship in average effort days across the various species, which (outside of mountain goat) shows a cluster of high average effort days to fill tags for low-quota species.  Figure 5 shows a scatter plot of each species ranking average effort days vs. average tag fill success odds (with bubble sizes reflecting counts of tags awarded for each species).  Note that Rocky BHS has a significantly higher average effort days as compared to all other species yet has an approximately average rate of tag fill success.  However, mountain goats and the other species of BHS have very high rates of tag fill success while having average effort days more closely aligned with the other species.  This indicates that the Rocky BHS hunts may be more challenging than the mountain goat and other BHS hunts.

Note that the feature ‘Total Fill Odds’ is calculated as ‘Tag Draw Success’ x ‘Tag Fill Success’ as a series of dependent events.  Although not directly related “in the field”, these are considered related events for this analysis because success or failure in the first event (the tag draw) directly impacts a hunter’s ability to participate in the second event (the tag fill).  Hence, if a tag has a 0.2 (20%) chance of being awarded, and that same tag has a 0.7 (70%) chance of being filled, then the total odds of success (both acquiring and filling the tag) can be calculated as 0.2 x 0.7 = 0.14 (14%). 

Figure 6 (below) shows boxplots of tag draw success broken out by individual species.  It provides value by showing not only averages, but by breaking the data into quartiles in order to identify the distributions of tag quotas.  For example, although the Desert BHS tag draw success shown above (in figure 2) was generally low, the boxplot shows that there were a few Desert BHS tags awarded at a much higher rate, with one at nearly 0.2.  Although from a relatively small sample, this shows that not all of the scarce tags are equally unlikely to be awarded.  It is possible that a hunter can use this information to identify specific hunts that may not get as much attention from other hunters, and which can provide an opportunity to focus resources on opportunities with better odds of success.

• 2021 was more challenging than 2020

NDOW data over the previous two years provides the opportunity to compare hunts year-over-year for a basic trend analysis.  Although only two years of data are available it is possible to make some useful inferences.  Figure 7 shows a comparison (by species) between 2020 and 2021.  It can be seen from the distributions that the average tag fill success dropped slightly in 2021 when compared to 2020 numbers.  In addition to this, in most cases the IQR (inner-quartile range, or the middle 50% of the data observations) also showed a general shift towards lower tag fill success rates in 2021. 

Backing up a step in the hunting process, figure 8 shows tag draw success by species.  Although visually similar to figure 7, the boxplots show a different story.  2021 shows a general tendency for greater tag draw success than that experienced in 2020.  However, although hunters experienced better overall success in drawing tags, they had (as figure 7 shows) lower success in filling those tags.

Of note, the average hunt quota (by species) decreased in 2021 (vs 2020).  The average hunt quotas (with error bars indicating 95% CI) are shown in figure 9.  Average hunt quotas for more scarce species remained virtually identical, but quotas for mule deer, antelope, and elk were notably reduced in 2021.  A logical expectation would be that this adjustment would result in improved tag fill rates, but as shown in figure 7 this was not the case.  Possible reasons could include too small of an adjustment to estimated population changes, changes in herd migration patterns, increased predation from mountain lions, disease, wildfire, or other factors affecting either herd populations or unit accessibility (both for hunters and for prey).

• Planning to optimize opportunities

Understanding the distribution of hunters by hunt type (figure 10) provides hunters with a way to plan hunt activity.  For example, figure 10 shows that more than half of all hunters in the field are hunting some variety of mule deer.  This knowledge helps individual hunters determine strategy, as the number of competing hunters puts pressure on animals and affects their behaviors.

This data can also be used year-over-year by hunters to compare (for example 2020 vs 2021) how changes in hunter counts may have contributed to differences in experiences.  Additionally (when filtered by units or individual hunts), it can be used to help plan to reduce the total number of hunters competing in the same unit.

• Weapon Success Rates

While species and unit difficulty are major factors in hunt selection, weapon selection is also important.  Part of the management of hunts includes designation of specific weapons for specific hunts.  This technique helps to “level the playing field” for both hunter-animal and hunter-hunter interactions.  Not all hunts are available for all weapons, and different weapons present different difficulties and challenges.  To increase the likelihood of success, it is important for hunters to assess not only the “what” (species) and “where” (unit), but also the “how” (weapon). 

Figure 11 (below) shows quotas by species and weapon, as well as the fill odds by species and weapon.  This is an important breakdown, because not only does each weapon present unique challenges, but these challenges differ by species.  Awareness of the differences can make a dramatic difference in hunt experience. 

Table 1 provides a breakdown for each of the eight big-game species.  The “Quota by Weapon” values in the left-hand column all add up to 100%, with each species’ pie chart showing their share of each weapon category:

• ALW: Any Legal Weapon
• AR: Archery
• SWR: Seasonal Weapon Restriction
• M: Muzzleloader

Additionally, each of the table 1’s “Quota by Weapon” pie slices added up according to its color will correspond to figure 11’s “Quota by Weapon” for the same color.  Figure 11’s and table 1’s “Tag Fill Success” pie charts correspond similarly, with the exception that in this case figure 11 uses an average in lieu of a total.

Using the first species (mule deer) as an example, table 1 shows that ALW hunts are far more plentiful than all other weapon categories (as per the quota percentages).  However, SWR hunts historically show a slightly  vs higher success rate (0.567 vs 0.520).   It  appears that an SWR hunt provides a slight advantage (0.047) to the hunter.  Contrasting this information with the overall data in figure 11, ALW hunts for mule deer are 0.083 less effective than overall population results, while SWR hunts for mule deer are 0.019 less effective than overall population results.  Ignoring other factors, mule deer SWR shows better results than mule deer ALW, which is the opposite of the overall big game population.  This would indicate that the SWR mule deer hunts are indeed more likely to provide a successful outcome for the hunter than an ALW mule deer hunt.

• Interactive Dashboards to Prioritize Species, Unit, and Hunt Opportunities

Previous analysis has focused on use of static displays to identify and compare the characteristics of different hunt features.  However, interactive dashboards and tools provide the ability to dig deeper into data that otherwise would have to remain averaged (to make it readable).  This has the benefit of allowing a granular exploration into the data without displaying an overwhelming amount of data points.

Through use of the Species and Unit Selection Interactive Dashboard (figure 12) it is possible to identify hunt opportunities by species and individual units/unit groups.  Each selection in the top-left panel dynamically updates the remaining panels in the dashboard.  The goal of this dashboard is to provide a visual means of comparing hunts in different areas to identify opportunities that best fit a hunter’s preference by species.

Shown below is a view of average success for Elk in unit group 071 – 079.  Multiple options can be selected, with results averaged.  The remaining three panels show corresponding information both visually and numerically.  To provide balance with the percentages provided in other panels, the bottom-right panel provides count values.  Additional filters (not pictured) can be applied to select specific date ranges and to narrow the focus to a single-year’s data through a simple selection box similar to the one shown on the left.

The Hunt Success Dashboard (figure 13) provides information on each individual hunt (selected in the left-side panel).  The points on the scatter plot represent historical Tag Draw Success for each of the hunts, and the size of each bubble represents the quota for each hunt (larger bubbles equal higher quantities).  The points are arranged by species and with highest Tag Draw Success to lowest.  On the right side of the dashboard are the key metrics related to the selected hunt.

Each year, NDOW schedules over 900 individual hunts.  These hunts occur approximately August 1^st through February 20^th, with individual hunt lengths ranging from a few days to a few months.  Using a tool like the Hunt Success Dashboard it is possible to identify multiple high-value hunts that overlap date ranges.  This can be accomplished by first identifying hunt opportunities by species, then overlapping dates to optimize timeframes.  This can be done to either overlap multiple species during the same timeframe (in the same or adjacent units/unit groups), or to sequence timeframes to extend the overall hunting timeline.

The Unit Success Rate Visualization Tool (figure 14) uses tree maps to identify (from highest to lowest) the Tag Draw Success, Tag Fill Success, and Total Fill Odds by hunt unit.  Selection of an item in one section automatically selects the corresponding value in the other sections and also highlights the corresponding unit/unit group on the map.  The selection works in any direction, so if a user were to click in any section (including the map) the corresponding selection will highlight in all sections. 

Note in the example that unit 171 is selected in all four sections.  Of interest is the fact that unit 171 has the highest Tag Draw Success, but a very low Tag Fill Success.  The Total Fill Odds, though, are relatively high.  Individual hunters can use their own self-assessment of their skill and experience level to identify which unit provides the best opportunity for his/her preference.  Not shown is the filter that allows selection of individual or groups of species. 

Also note also that this tool is useful to non-hunters as well.  By selecting either the Tag Draw Success or map areas, non-hunters planning to use an area for recreation can obtain an estimate of how densely hunters will be populating the selected area.  This (when augmented with the previously described date-range selection filters, not shown) provides a way to increase safety by helping non-hunters to avoid heavily hunted areas.  This reduction in human presence would have an accompanying benefit of reducing stress on the local animal population.

Conclusion:

The analysis above identifies some key ways that hunters can use existing NDOW data to identify existing possibilities to improve hunt success odds through a focus on species and unit-specific imbalances in opportunities.  Specific insights include:

• Analysis of readily available data can help improve the odds of being awarded a hunting tag, which is the biggest constraint on overall hunt success.
• Different species have different difficulty levels; in some cases, there is an opportunity to dramatically increase likelihood of success by increasing the total effort days.
• Not all years carry the same opportunities.  2021 was more challenging for hunters than 2020 and adding upcoming hunt completion data will allow for better trending.
• The high number of mule deer hunters in the field may be related to the fact that mule deer tag fill rates compare poorly to those of other species.
• Where weapon selection is an option, the selection of weapons can have a significant impact on likelihood of success.
• The use of interactive dashboards provides the opportunity to plan for success at a much more granular level than static data visualizations allow, but both have utility.

Through thoughtful use of the information available, a savvy hunter can significantly improve the likelihood of success in both award of tags and filling of tags.

Resources:

1. NDOW Hunt Statistics: https://www.ndow.org/blog/hunt-statistics/#
2. NDOW Big Game Status Book: https://www.ndow.org/wp-content/uploads/2022/01/2020-21-BIG-GAME-STATUS-BOOK-COMPLETE.pdf
3. HuntNV: https://www.hunt.wildlifenv.com/map