Draw it · Fill it
Checking status…

Transparency

Our data, plainly

Where every number comes from, how the opportunity score is computed, and the caveats that matter. Published rates — never guarantees.

Transparency & methodology

How Odds & Antlers' numbers are made, where the data comes from, and the caveats that matter. Short version: every draw and harvest figure is a published rate from official reports, and the opportunity score is computed from two of them — not a guarantee, and not a substitute for the official rules.

The opportunity score

The opportunity score ranks hunts that you can realistically draw and fill. For each hunt, residency class, and year it combines:

  • Draw probability — successful applicants ÷ total applicants in your residency pool for that hunt and year, from the agency's published drawing-odds report (counted across every application choice together, not just your first pick).

    Which figure leads. Where an agency prints first-choice awards, or its own first-choice figure, the Explorer row and Hunt Detail lead with the first-choice rate: the share of applicants who listed the hunt first and drew it (New Mexico's report prints the awards; Colorado's prints them for adults and youth apart, and the lead counts adults only, labelled so; Arizona's report prints the figure itself, with no award count). The overall draw rate sits beside it, labelled by what it counts ("All choices", "First choices", "Permits per first and second choices"). The score and every sort stay on the overall rate, and the page says so. A hunt no one listed first, or in Colorado no adult, has no first-choice rate, and the row says so.

  • Harvest success rate — harvest ÷ hunters for the hunt, from the agency's published harvest reports, summed across the hunt's reporting units. Colorado's 2025 report counts by license rather than by unit hunted, so its figures are not comparable with 2021–2024: charts break at the change with a divider and a note. For Colorado 2021–2024, a hunt's figure adds up CPW's figures for that kind of hunt in the units it was valid in, so it can include hunters on other licenses; a hunt page notes this beside each of those years' figures.

The default model multiplies the two (score = draw probability × harvest success rate), which rewards hunts that are both drawable and productive. A weighted-sum alternative is also available. Scores are computed separately for each hunt, residency class, and year — resident, non-resident, and outfitter applicants each see the published rates for their own pool.

When an input is missing, the score is left blank rather than guessed. The app never invents a number.

Where the data comes from

Everything comes from the covered wildlife agencies' published reports — drawing-odds summaries and per-species harvest reports. We keep a record of every load: the source, the year, the row count, and when it happened. The live table on the transparency page shows, for each jurisdiction, the publishing agency, each report we use, the years covered, and when it was last refreshed — the real, current record of what's actually loaded.

Data caveats

Honest numbers require honest caveats:

  • Summary vs. complete reports. Some agencies publish both a summary and a complete drawing-odds report. The summary's drawn-count columns can drift or drop values year to year — in New Mexico's reports we measured drawn counts 0 % populated for 2023 and only ~67 % for 2024 in the summary — so wherever both exist we ingest the complete report. Where a registered source is labeled "Summary," the loaded figures may still come from the complete edition for backfill years.
  • Not every state counts applicants the same way. Some agencies' published draw-odds reports only count an applicant toward a hunt if that hunt was their first pick — someone who ranked it second, third, or fourth and didn't get their first choice isn't counted at all for that hunt, even though they applied to it. Other states count applicants at each rank they print: New Mexico and Colorado print every hunt-code rank, while Arizona's report prints only first and second choices, though its application takes five. That changes what the raw applicant number — and the draw odds computed from it — actually means, so comparing one state's draw odds directly against another's from memory can be misleading even though both numbers come straight from the agency's own report. Where this affects a hunt, the app flags it.
  • A random draw is not a lottery ticket you can count. Draw odds describe last year's pool. Applicant counts move year to year; your odds this year will differ.
  • Refresh cadence. Ingestion runs when new reports are published or an operator refreshes — the vintage badge shows exactly how fresh the served data is. Stale data is labeled, never hidden.
  • Herd-area figures are context, not a unit's figure. Colorado's harvest reports also print hunters, harvest and success per herd area (CPW's Data Analysis Unit, a group of units), with 95% limits on hunters and harvest. We show them as printed beside each unit and never sum them, add them to a unit's or a hunt's figure, rank them or score them. The 2021 table covers hunts open to the public through the draw, landowner vouchers, over-the-counter and leftover sales; from 2022 it adds field-staff counts for hunts outside the survey, and the page marks the change. Membership is read per year, because CPW regrouped deer herd areas in 2023 and 2024.
  • Small pools. A draw figure that rests on fewer than 30 applicants, or a harvest figure on fewer than 30 hunters, swings too much to read like the rest. At 1-4 applicants a pool's draw rate moved 27.0 points a year on average, against 3.6 points at 100 or more. So we label these figures, and pools under 30 applicants sort below the rest. Nothing is withheld, and the stored opportunity score is unchanged.
    • The label. The counts replace the plain phrase and carry a "few applicants" or "few hunters" tag. Each count is named for what it is: "5 of 8 drew" where the awards went to the applicants counted; "25 permits, 2 first and second choices" where the figure is a permit quota; "11 of 27 drew on their first three choices, plus 77 on a fourth" where a total carries awards whose applicants it does not count; the awards alone ("5 awards, no applicants printed") where the report prints no applicants; and each count under its own name otherwise. Harvest always reads "N harvested, M hunters", because most harvest counts are the agencies' survey estimates and a bag can lift harvest above hunters.
    • The sort. Such pools sort below the rest on the Explorer's default order, on the search pages and, inside each band, in the planner. On the Explorer's default order, scored rows of 30 or more applicants come first, then scored rows under 30, then unscored rows. On the search pages, rows under 30 applicants rank below the rest within the scored rows and within the unscored rows, and unscored rows stay below scored ones. Their "up to" lines read only rows that are not few, and a hunt page's description says "a few applicants" or "a few hunters" where a figure is few.
    • The confidence grade. A small pool's draw figure, and a season's success figure on fewer than 30 hunters, grade at most "Moderate" (issues 0714, 0731).
    • Ladders. Point-tier rows keep their counts and take no tag (ADR 0024).
  • Published rates, not guarantees. We read imperfect public reports. We do not claim to be more accurate than the agency's own numbers — trust theirs over ours, and always confirm season dates, rules, and deadlines against the official publications before applying.

Random draw vs. points

Some states allocate tags by preference or bonus points accumulated over years; others run a pure random draw. Covered jurisdictions now include both kinds: New Mexico's draw, for example, is pure random — no points, no waiting line, everyone in a pool has the same chance each year, and the published pool sizes are what the odds mean. Colorado, for example, allocates by preference points instead — the draw odds we serve for every hunt are still the overall applicant-pool rate (summed across every point tier), not personalized to your accumulated points. That rate also counts applicants who listed the hunt as a later choice, so it can read well under 100% on a hunt where every first choice drew. It also overstates your chance on premium, point-creep tags if you hold few points.

Where an agency publishes per-point-level results — Colorado's draw recap and Utah's draw reports do — each hunt's page now also carries a free "Odds by points" ladder: the printed applicants, awarded permits, and empirical rate at every published point level (first-choice-scoped, since that is what the agencies publish). Colorado prints adult and youth columns separately, and a Colorado ladder serves the adult columns: the applicants, permits and rate at each level are the adults'. Its point cutoff is the adult floor CPW prints, and where CPW prints none, the lowest level at which an adult drew. If your profile's birth year makes you a youth for the coming draw, the page also shows CPW's printed youth counts beside each level; they never feed a ranking. On some Colorado hunts, part of that year's licenses, up to a share CPW prints, could also go by random draw among applicants with enough points, and the page says so. An award from that draw counts in its level's rate like any other, so a level below the adult floor can show an adult award. On a Utah bonus-point ladder from the 2026 draw, some of a hunt's permits can go to the applicants with the most bonus points, and at least part of every quota, all of it where a residency class has one permit, goes by random draw, where every applicant holds one entry plus one for each point. So the ladder says that a level where no one drew in one year still had a chance (Utah Admin. Code R657-62-8(4), (5)). We never interpolate, extrapolate, or simulate between published levels. The ladder feeds one ranking only: the Strategy Planner's 1st choice, where the agency's ladder is rankable and you have saved your points (see "Season plans" below). Everywhere else it is display context: where your points are known (in Your residency pool, a Pro feature), the figure at your points leads, with the rate across all point levels beside it, and it never changes a ranking. Arizona publishes no per-point breakdown, so Arizona hunts carry only the overall-rate caveat. Each hunt's page labels all of this via a per-hunt caveat, so you always know which kind of odds you're looking at.

Season plans (Strategy Planner)

The Strategy Planner builds a choice stack for each species you pick, one choice per slot on the agency's application. Here is how its numbers are made:

  • Which hunts it considers. Each species is planned from that species' latest published hunt list — the newest year the agency has published draw results for — and only from hunts you are eligible for. A species whose report comes out later than the others keeps its own latest list. The plan targets the next application cycle, using each species' latest published hunt list as its stand-in for that cycle's list.
  • The odds for each hunt. The planner uses the latest year's published draw odds, and does not carry them forward to the next cycle. Each pick says which year it used.
  • How hunts are ranked. Inside each odds range, a hunt whose pool had fewer than 30 applicants in the basis year is listed after the rest; it still fills the slot when no larger pool is in the range. Within each group, hunts rank by a score that combines those odds with the hunt's most recent harvest success rate published for the latest list's year or earlier. Harvest reports usually lag the draw by a year, so this is often the year before. A hunt with no harvest figure for that year or earlier ranks after every hunt in its group that has one.
  • A point-aware 1st choice, where the ladder is rankable. Where the agency's point ladder is registered as rankable — today only Colorado preference points, through the 2027 draw — and you have saved your points, the 1st choice ranks on the agency's published first-choice rate at your point level, which in Colorado is the adult rate (a random-draw award stays in the rate at its level): the rate at your level; the nearest published level below yours when yours is not printed; the top level's rate when you hold more points than any level printed; and 0 when you hold fewer than the lowest level printed, which in Colorado is the lowest level an adult applied at (the plan then says where the ladder starts). The rate is the published one, never projected forward or estimated between levels, and the plan reports how many applicants and permits it rests on. A hunt with no current-year ladder at your residency class is not offered as a 1st choice, so the 1st choice is compared on one kind of odds only; it can still fill a later choice. If no laddered hunt fits the 1st choice's odds range, that choice stays empty rather than falling back to overall odds. Many published rates are exactly 100%, so hunts with the same rate are ordered by harvest success rate, then hunt code. The ranking does not weigh how many points a pick spends: a hunter who draws a Colorado 1st choice forfeits that species' preference points (Colorado Parks and Wildlife), so the plan reports the point cutoff (in Colorado, CPW's printed adult floor, or the lowest level where an adult drew; elsewhere, the lowest level that drew) and says what drawing it spends. Without saved points, and in every other point system, the 1st choice uses the hunt's first-choice rate in New Mexico, AZGFD's printed first-choice figure in Arizona, the share of adult applicants who listed the hunt first and drew it in Colorado, and its overall odds elsewhere, as described below.
  • The plan says what a draw spends, and leaves its picks alone. Where a state's rule is on file and you hold at least one point, a pick says so beside its point figures, worded by that rule. In Colorado a 1st choice reads "Drawing this as your 1st choice spends all {points} of your points. A later choice spends none.", because CPW's brochure says "If you are issued a first-choice license, all accumulated preference points for that species become void and return to zero." and that later choices "don't use or generate points". In Utah a draw on any choice spends them, because the rule turns on obtaining the permit (R657-62-8 and R657-62-9). One saved balance can cover more than one of the totals Utah keeps: preference points are kept by kind of hunt, and bull moose and antlerless moose bonus points are kept apart. So the preference line says "for this kind of hunt", the bonus line says "for this species and kind of hunt", and each says "up to" your points rather than "all" of them. In Arizona a computer-draw tag spends your bonus points on any choice, except a permanent education point and a loyalty point (Arizona Game and Fish Department, 2026-27 Arizona Hunting Regulations, R12-4-107(F)). Arizona publishes no point ladder, so its plan reads the balance you saved and Hunt Detail does not carry the line. Each rule is a sourced row per state and point system. Colorado's row ends with the 2027 draw until CPW's 2028 rules have been read, and the line stops there. Ordering Colorado's 1st choices by the points they reach was tested and not adopted: within a unit it saves 0 to 2.5 points at the median and gives up 1 to 9 points of success.
  • In New Mexico every choice shows the hunt's first-choice rate. New Mexico gives each application one random number, so how hard a hunt is to draw is best read off its first choices. Every choice, the 1st included, shows the share of applicants who listed the hunt first and drew it, with those counts, and is sorted into its odds range on that rate. It is not your chance at that choice. Checked against the next year's reports, hunts this rate put in the likely range drew about two in three of their first choices. A hunt nobody listed first has no such rate, so the plan leaves it out and says how many it left out and how many of those still gave tags to later choices. Where NMDOW's report prints no first-choice award count, the hunt keeps its overall odds.
  • In Colorado every choice not ranked on your points shows the adult rate at that choice. CPW's draw recap prints, for each choice, how many adult and how many youth applicants listed the hunt and how many of each drew. Every choice not ranked on your points, the 1st included when you have saved no points, shows the share of adult applicants who listed the hunt at that choice and drew it, with those counts, and is sorted into its odds range on that rate. It is the share of applicants who listed the hunt at that choice and drew it, never your chance at that choice. At a later choice its awards go only to applicants who missed their earlier choices, but it counts everyone who listed the hunt there, including applicants who drew an earlier choice elsewhere. Youth applicants and youth awards are left out, as they are from a ranked 1st choice's rate, so a hunt where only youth drew at a choice shows 0 there. A hunt no adult listed at that choice has no rate there, so the plan leaves it out of that choice and says how many it left out. Where a report prints no adult and youth split, a choice not ranked on your points keeps the hunt's overall odds.
  • In Utah every choice uses the hunt's overall odds. In Arizona the 1st choice shows AZGFD's own printed draw odds for the hunt as a first choice, and later choices use the overall odds. Utah's portal prints only an all-choice count. AZGFD's report prints its figure beside the number of first choices and no award count, so the plan shows that count and no award count. It is not your chance at that choice. A hunt the report prints no figure for, because nobody listed it first or the cell holds no number, is left out of the 1st choice only, and the plan says how many; it can still fill a later choice. Arizona draws the 1st and 2nd choices together and prints no award for either. Overall odds can overstate a later choice: a hunt can fill on first choices and still show a high overall rate.
  • In New Mexico the plan states your chance of a tag; nowhere else is a chance combined. New Mexico gives each application one random number and walks its choices in order (NMDOW's booklet, "How the Draw Hunt System Works", and its "How Draw Works" page), so the order of your list decides which hunt you draw, not whether you draw. Your chance of a tag is the first-choice rate of the easiest hunt you list, whatever its position, and your chance at each choice is that hunt's rate minus the highest rate listed above it: a hunt below or level with an easier one has none. The card names the hunt the figure rests on with its counts, shows each choice's chance, and gives a range of two standard errors on that count plus five points, which last year's figure landed inside on about nine in ten resident elk and deer plans and nearly every pronghorn and oryx plan. The reports are consistent with the rule and cannot confirm it, because none shows how one application's choices turned out; every report year the app carries re-tests it. The figure counts your three hunt choices and reads this year's counts as next year's chance; the deer quadrant choice adds a little over one point on top. A party application needs a license for every member, so a hunt's cutoff is not quite one position for every application; the reports cannot size that. A stack with a hunt the report prints no first-choice award for shows no figure. Every other state states none: no report can check a chance combined across hunts under their rules, and the planner shows each hunt's own figures.
  • In New Mexico the plan lists its hunts hardest first. Your risk level decides which hunts go in, and the rate each choice shows decides the order, lowest first: the hunt's first-choice rate, or its overall odds where the report prints no first-choice count. With one number per application a hunt listed below an easier one gets nothing. A conservative plan therefore reads mid odds, likely, likely, and an aggressive plan lists its harder long shot first. The order is a rule of thumb read from the printed counts, how many first choices drew, not a chance the app computes, and those counts move from year to year. It has a cost: the easiest hunt drops from first in line to last resort. If you most want the easiest hunt, list it first yourself. If a swap leaves a hunt below or level with an easier one, that choice says so. Other states keep the order of your risk level.
  • A later choice's rate is likely optimistic. Where a later choice reads the overall odds, its rate is likely optimistic, because a later choice is considered only after the earlier ones miss. That holds from the first choice the agency does not draw in its first pass: the 2nd choice in most states, and the 3rd in Arizona, which draws the 1st and 2nd choices together.
  • Season warnings. Utah lets any applicant draw only one limited-entry or once-in-a-lifetime species a year, so when a plan holds more than one, it warns you, because each species' odds read as if that species could be drawn. Where residents may apply for only one species per hunt type, a resident's plan holding several species of one type warns you too. Swapping in an alternative shows that hunt's own figures on the page.
  • Season dates. The window and the overlap check apply to the hunts that carry season dates, and the plan tells you how many had none. Hunts in every served state carry their published dates, except Utah's CWMU hunts, where the operator sets the days. A Utah Dedicated Hunter hunt carries the general archery, muzzleloader and any legal weapon seasons of its unit, which the program's rule names. Where the unit also has an early any legal weapon season, that window is kept but not planned on, because no agency source says whether the permit covers it. A hunt whose windows are dated by weapon is read on the windows the hunter's chosen weapons may hunt, and meets the weapons preference through such a window; a hunt with no such window, or a hunter with no weapon chosen, is read on every window. Odds are per hunt code: where one code's windows carry their own licences (a New Mexico bighorn code), the code ranks on its pooled odds, not on the odds of the window that fits, because the draw report publishes applicants per code only.

Independence

Odds & Antlers is an independent product. It is not affiliated with, endorsed by, or speaking for any wildlife agency. Agency names appear only to identify the public sources of the data.

One more thing about states that use preference or bonus points: the odds you see here are still the overall pool rate, summed across every point tier, not adjusted for your own points. Where your points are known, a hunt’s page and the Explorer lead with the figure at your points, beside that rate. When an agency publishes results by point level, you’ll find that ladder on the hunt’s own page too, built from their printed figures.

Photo credits

The photos on this site are public-domain or openly licensed. Each is credited where it appears; the full list is below.