How we rate
THMPRK · Editorial policy · Version 1.0 · October 24, 2026
The THMPRK Score is our researched assessment of a park, out of 10. It helps readers compare the experience and the likely cost of a day out. It is an editorial judgment based on documented sources. It does not imply that a THMPRK editor has visited. Visitor reviews and public star ratings are retired; source-based corrections remain a separate service.
This policy supersedes visitor-rating and review-submission requirements in the original business plan. Existing visitor data is retained privately rather than deleted. Public resort and attraction pages do not inherit an average of their children’s scores. This first rubric covers individual theme parks and independently operated water parks.
The seven categories
| Category | Weight | What we assess |
|---|---|---|
| Value for a day out | 30% | The worthwhile included experience relative to a realistic total cost: admission, tax, unavoidable fees, parking where relevant, and realistic food options. Optional queue upgrades and other extras are disclosed separately. Cheap alone does not mean good value. |
| Attractions and experiences | 20% | Quality, variety, repeat appeal and a coherent day’s worth of experiences for the park’s intended audience. Count operating attractions, not announced projects. A small family park is judged against appropriate peers, not against the coaster count of a major destination park. |
| Operations and time well spent | 15% | Reported reliability, queue handling, communication, staffing and realistic opportunities to enjoy what admission includes. Distinguish peak-day experiences from typical days, weather closures from persistent issues, and premium access from normal admission. |
| Reported cleanliness and upkeep | 10% | Recent, repeated visitor evidence about restrooms, litter, seating, dining areas and visible upkeep. This is reported visitor experience, not a hygiene inspection, water-quality measurement or safety certification. |
| Accessibility and inclusion | 10% | Practical access: clear advance information, physical and sensory access, assistance processes, family needs and evidence of how those arrangements work. Clear documentation alone cannot prove a barrier-free visit. Do not penalize a visitor’s needs or infer legal compliance. |
| Comfort and amenities | 10% | Shade or weather protection appropriate to the setting, seating, toilets, drinking water, food choice, dietary information, lockers and a manageable day. Avoid scoring cleanliness twice. |
| Character and distinctiveness | 5% | What makes this park worth choosing: coherent atmosphere, memorable setting, original experiences and local character. Licensed intellectual property and marketing budget are not proxies for quality. |
The calculation
Score = 0.30V + 0.20A + 0.15O + 0.10C + 0.10I + 0.10M + 0.05D
Editors assign each category from 0 to 10 in half-point increments. Store full calculation precision; display the composite to one decimal place. All seven categories must be supported and approved. Missing evidence never becomes zero, a neutral five, or a redistributed weight. An incomplete assessment is Not yet assessed, with no numerical score or ranking position.
Illustration only, not a real park rating: category scores 8, 9, 7, 8, 7, 8 and 9 produce 8.00, displayed as 8.0/10.
Shared anchors
0 means the lowest substantiated performance; 2 means serious recurring weaknesses; 4 means material shortcomings; 6 means a satisfactory experience with meaningful trade-offs; 8 means strong performance with limited drawbacks; 10 means exceptional performance supported across the evidence. Intermediate scores need a written reason. An absent, untested or contradictory fact has no score.
Make value comparable
Record two price baskets: two adults, and two adults plus children aged 8 and 12, applying the park’s actual height or age rules. Exclude accommodation and travel to the destination. Record the visit date, booking lead time, currency, taxes, mandatory charges, assumed parking and food basket, inclusions and exclusions. Compare a normal weekday and weekend during the published operating season, with a common advance-booking lead time where possible. Do not use an extreme promotional minimum as a typical price.
Assess value against at least five suitable regional peers of the same park type and audience, with comparable dated baskets checked within 90 days. When the peer set cannot be supported, keep value unscored. Consider included experience and the likely usable day, not cost per ride or ticket price alone. Keep both adult and family trade-offs visible; do not quietly change the target audience to improve a score. Published scores remain editorial judgments, not statistically estimated value measures.
Evidence before a number
Every category needs a rationale, at least two independent public publishers, actual source URLs, check dates and medium or high confidence. For published facilities and rules, one publisher may be the park’s official website. For visitor-experience judgments, both must be independent third parties. Two pages from one publisher, syndicated stories and an operator-owned account on another domain do not establish independence.
For operations and cleanliness, review at least 20 relevant, nonduplicate first-hand visitor accounts per category, distributed across at least two independent platforms. The accounts must concern visits within the previous 12 months, and the sample must have been reviewed within 90 days. Prefer the latest full operating season for seasonal venues; where that falls outside the window, leave the score unpublished until fresh evidence exists. Review dates alone do not establish visit dates.
Use a repeatable sample rather than a keyword search for complaints: record the date range, platform split, selection method, number screened, relevant accounts and exclusions. Include positive, mixed and negative evidence. Repeated or syndicated accounts count once. Distinguish separate operating seasons, special events and unusual incidents. Do not infer cleanliness from a platform’s overall stars. If the platform does not offer a relevant cleanliness measure, assess relevant narratives and label them as such. The minimum sample is an editorial threshold, not a representative survey or margin of error.
Keep private research notes with brief original summaries and source references. Do not copy entire reviews, photos or reviewer identities. Use permitted access or authorized APIs; do not bypass login walls, CAPTCHAs or access restrictions. If credible sources disagree, record the conflict and lower confidence or withhold the category. Official promotional claims cannot resolve an independent experience dispute.
Approval, freshness and independence
Automation can find changes, propose evidence summaries, identify duplicates and calculate a score from approved inputs. A named editor must inspect the actual sources, confirm publisher independence, apply the rubric and approve the assessment. A fetch or an AI-generated summary does not count as an editorial check.
Value evidence expires after 90 days; other categories after 180 days. Operations and cleanliness also require their recent sampling window. The complete assessment must have an approval date within 90 days. Refresh after material pricing, operating or accessibility changes even before expiry. When any required component expires, withdraw the current composite and ranking eligibility; retain its private history. Do not roll forward dates automatically.
Publish the assessment date, named editor, research basis, criterion scores, weights, reasoning, confidence, source links, best-fit audience, strengths and caveats. Disclose paid partnerships, affiliate relationships and hosted visits. A park cannot buy a score, suppress a justified weakness or edit its assessment. Corrections receive evidence review and a dated decision trail. A change to weights creates a new methodology version; compare or rank only current assessments on the same version.
How it appears on THMPRK
Use THMPRK Score — researched assessment and a clear “8.0 / 10” treatment, never visitor stars or a fabricated review count. A restrained breakdown table is accessible without JavaScript, with text labels and no color-only signals. Show confidence separately from quality. An unrated park is not a bad park. Do not display a ranked leaderboard until a useful comparable group has current, complete assessments. Ties share a position; rankings use displayed one-decimal scores, with alphabetic ordering for ties.
HTML and Markdown must expose the same approved public assessment, methodology version, sources, dates and uncertainty. Private sampling notes, draft numbers and personal research information must stay out of REST, JSON-LD, Markdown, feeds and sitemaps. No Review or AggregateRating rich-result markup will be emitted for these desk-researched local-business assessments. Google’s current review-snippet guidance prohibits aggregating other sites’ ratings and instructs that local-business ratings come directly from users rather than editors. Google Search Central: review snippets, checked October 24, 2026.
Catalog rules accompanying this policy
Park names use official spelling and capitalization, including EPCOT. City, region and country live in separate location fields and tile subtitles. Clean park slugs omit appended location text; genuine naming collisions receive a deliberate disambiguator. Canonical paths contain the real country and, where applicable, the real resort slug. Missing country or resort does not create an “anywhere” placeholder. Preserve old paths as redirects.
A theme park and a distinct water park may form a THMPRK resort grouping when sources establish the relationship. An embedded, included water area remains part of the theme park, as with Lagoon. A grouping created for navigation is labeled as a THMPRK grouping unless the operator uses that official resort name. Preserve individual park identities and show useful resort-wide planning, all child parks and links back to the parent.