The short version
Every public golf course gets a score out of 100.
It’s built from what the experts, the golfers and the course itself tell us, and every input is cited on the course’s page.
A great architect and a real story earn bonus points. A missing one never costs a course anything.
The score is national. An 85 in Kansas means the same thing as an 85 in North Carolina.
How the score reads
Out of 10, with one word that says what it means.
| Score | Label | What it means |
|---|---|---|
| 8.5+ | Legendary | The bucket list |
| 7.5–8.4 | Destination | Worth the trip |
| 6.5–7.4 | Memorable | You’ll talk about it after |
| 5.5–6.4 | Playable | A good round, any day |
| under 5.5 | Unsolved | Nobody’s cracked this one yet. Go solve it. |
| no score shown | Uncharted | Not enough to go on: no expert list and fewer than 5 golfer reviews. Go see for yourself. |
The cut-offs follow where real courses land: in North Carolina, about 1 in 30 public courses is Legendary. An Uncharted course isn’t a bad course, just an unknown one, and it gets a score as soon as golfers start reviewing it.
Next to every score we show the Online Golfer Rating: the average stars golfers gave the course online, with how many reviews and where they came from, one tap away. Under the hood the math below runs on a 100-point scale; the score you see is that number divided by 10.
The formula
Total Score = min(100, Core + Story) Core = (E + G + C + A + T + D) × 100 / 85 when we know the architect Core = (E + G + C + A + T) × 100 / 75 when we don't
| Symbol | Part | Points |
|---|---|---|
| E | Expert lists | 0–40 |
| G | Golfer stars | 0–20 |
| C | Challenge | 0–5 |
| A | Access | 0–5 |
| T | Threads and blogs | 0–5 |
| D | Architect | 3–10 (bonus) |
| Story | Story and heritage | 0–15 (bonus) |
If we don’t know who designed a course, we don’t guess and we don’t punish it. We leave D out and score the course on everything else, stretched to the same 100 scale.
The parts
E — Expert lists (0–40)
E = max(20, 40 × (0.7 × P + 0.3 × L / 4)) P = 1 − (rank − 1) / max(list size, 25% of the state's courses) ← best showing on any list L = number of lists the course is on (out of 4)
The lists:
- Golf Digest Best in State
- Golfweek Best Courses You Can Play
- GOLF.com Best in State
- Top 100 Golf Courses
How it reads:
- #1 on a list of 25 gives P = 1.00.
- #13 of 45 gives P = 0.73.
- Small lists in big states: a list can’t count as shorter than a quarter of the state’s courses. Being #10 on a 20-course list means more in a state with 70 courses than in one with 700.
- Not on any list gets half credit (20). Experts haven’t ranked every course, and a course they haven’t seen hasn’t failed. A listed course always scores at least 20, and almost always more.
G — Golfer stars (0–20)
avg = Σ (w × reviews × average) / Σ (w × reviews) ← for each age group n = Σ (w × reviews) w = 1 (played in the last 2 years), ½ (3–5 years ago), ¼ (older) S = (avg × n + 3.5 × k) / (n + k) G = 20 × clamp((S − 3) / 2, 0, 1)
Source: GolfPass golfer reviews.
We count the stars ourselves. GolfPass shows a headline average at the top of each course page, and it often doesn’t match the page’s own count of 5-, 4-, 3-, 2- and 1-star reviews. One course showed 2.4 while its 593 reviews averaged 4.44. We ignore the headline and work from the counts.
Recent reviews count more. A course changes: new owners, a renovation, a bad summer. So a review from the last two years counts fully, one from 3–5 years ago counts half, and anything older a quarter. A course coasting on old praise slides, and a course that just got better rises.
What S does: every course starts as if it had k extra 3.5-star reviews. A 5.0 from two reviews can’t beat a 4.6 from two hundred. The more reviews a course has, the more its real rating shows through. Old reviews count for less here too, so a course rated mostly years ago is held closer to 3.5.
k depends on the state. Golfers in busy states write far more reviews. k runs from 2 in the state with the fewest reviews per course to 10 in the state with the most, so a quiet state’s courses aren’t held back for being quiet.
| Adjusted rating (S) | Points |
|---|---|
| 3.0 or lower | 0 |
| 4.0 | 10 |
| 5.0 | 20 |
C — Challenge (0–5)
C = 5 × clamp((slope − 113) / 32, 0, 1)
Source: the USGA slope from the longest tee.
| Slope | Points |
|---|---|
| 113 (average) | 0 |
| 145 or harder | 5 |
| No rating on file | 2 |
A — Access (0–5)
| Who can play | Points |
|---|---|
| Public, municipal, county, state-run | 5 |
| Resort, university | 4 |
| Semi-private | 3 |
Private clubs aren’t ranked at all. If you can’t book a tee time, it’s not on the list.
T — Threads and blogs (0–5)
One point for each independent, cited mention, up to 5.
Mentions come from golfers writing about the course on their own: forums, blogs and trip reports.
D — Architect (3–10, bonus)
D = 3 + 3.5 × V + 3.5 × R V = ln(courses built) / ln(most courses built by any architect) R = max( min(1, listed / 5), min(1, (listed / built) / 0.25) )
What goes in:
- Built is the architect’s career total of original designs, not how many we happen to have on file.
- Listed is how many of their courses appear on Golf Digest’s America’s 100 Greatest, Golf Digest’s 100 Greatest Public and the state lists above.
How the three numbers work:
- The 3 floor: anyone who designed a course starts at 3.
- V (volume): uses a log scale, so going from 1 course to 10 counts as much as going from 10 to 100. The busiest architect in our data, Tom Bendelow at about 600 courses, sets the top.
- R (recognition): an architect earns full credit either way: 5 or more listed courses, or a quarter of everything they built made a list. Big careers and small, brilliant ones both get their due.
| Architect | Built | Listed | D |
|---|---|---|---|
| Donald Ross | ~400 | 32 | 9.8 |
| Pete Dye | ~200 | 16 | 9.4 |
| Tom Doak | ~40 | 11 | 8.5 |
| Mike Strantz | 9 | 3 | 7.7 |
| Arthur Hills | ~200 | 1 | 6.6 |
| A one-course architect | 1 | 0 | 3.0 |
Story — Story and heritage (0–15, bonus)
Story = min(15, History + Heritage + Milestones + Era)
| What | Points |
|---|---|
| The course’s history is researched and published on its page | 5 |
| A published Heritage story | 5 |
| Milestones | up to 5 total |
| · A renovation or restoration in the last 25 years | 2 (+1 by a recognized architect) |
| · Hosted a major event: pro tour, USGA, state or college championship | 2 |
| · A documented first: first public course in the county, first integrated course and more | 1 |
| · An award: Golf Digest or Golfweek Best New, design and “best of” honors | 2 |
| Era | course-building history |
| · Opened before 1930 | 3 |
| · Opened 1930–1959 | 2 |
| · 1960 or later, or unknown | 1 |
Era is a small nod to course-building history, so a brand-new course is never more than two points behind a classic. New courses catch up through milestones: awards, the events they host and the work done on them. Every milestone is cited, like everything else in the score.
Story is added after the 100-point scale, so it can lift a course but never sink one.
Worked example: Pinehurst No. 4 = 9.1 Legendary
Pinehurst, North Carolina
| Part | Inputs | Points |
|---|---|---|
| E | Best showing: Golfweek #3 of 20, counted as 88 (a quarter of NC’s 353 courses) → P = 0.977; on all 4 lists | 40 × (0.684 + 0.300) = 39.4 |
| G | 21 reviews (13 five-star, 6 four, 2 three) = 4.52. None from the last 2 years; 3 from 3–5 years ago (4.67, half weight), 18 older (4.50, quarter weight) → avg 4.54 from n = 6. NC k = 10 → S = 3.89 | 20 × 0.45 = 8.9 |
| C | Slope 138 | 5 × 0.78 = 3.9 |
| A | Resort | 4 |
| T | Not yet collected | 0 |
| D | Donald Ross (~400 built, 32 listed) | 9.8 |
| Core | (39.4 + 8.9 + 3.9 + 4 + 0 + 9.8) × 100 / 85 | 77.6 |
| Story | History 5 + milestones 5 (2018 Gil Hanse renovation 3, 2019 U.S. Amateur 2, 2019 GOLF award 2 = 7, capped at 5) + opened 1919 → 3 | +13 |
| Total | 77.6 + 13 = 90.6 | 91 → 9.1 |
No. 4’s golfer score is low only because so few golfers have reviewed it lately. Ten good reviews this season would add several points.
The rules we hold ourselves to
- Every number is cited. The course page shows each list, rank and star rating with a link to the source.
- No paid placement. A course can’t buy points or position.
- No curve across states. The score is national. On a state’s Top 10/50/100 page, we may also show how each course compares with that state’s #1.
- Missing data never counts against a course. That covers an unknown architect, no story yet, or no threads collected yet.
- The score moves when the evidence moves. New lists, new reviews and new stories update it automatically.
Coming next: ratings from Formula 18 Golf players themselves, and rounds played on Formula 18 Golf.
See it in action: Best Public Golf Courses in North Carolina →