How we score
Review Methodology
How 76 sites are scored, what gets checked, and how the overall score is calculated. Updated September 30, 2026.
The weighting
- Content Quality35%
- Value For Money25%
- Site Design20%
- Update Frequency20%
Same four weights for every site. Commission rate carries zero weight.
Current review profiles are research-based. They use public tours, publisher information, join-page prices where checked and available user reports. No current review is marked in our review data as tested from a paid subscription. That means member-only playback, search, downloads and cancellation experience should be treated as unverified unless a review is later backed by documented direct access.
The pillars and their weights are fixed across all reviews — every site is scored against the same rubric, and that rubric doesn't shift based on affiliate relationships, niche, or how new the site is.
The "Researched" badge describes the current catalog. A "Tested from a paid subscription" badge is reserved for a review with recorded access and specific findings. Both use the same four-part framework, but a research-based number is an editorial estimate with clear limits, not a hands-on verdict.
The four pillars
Each one scored on its own, then weighted into the overall score.
Content Quality
How good is the actual content — and is there enough of it?
- Publicly visible scene count, labeled as publisher-reported when not independently counted
- Percentage exclusive to this site vs. licensed/recycled
- Default streaming resolution (4K / 1080p / 720p / SD)
- Performer roster size and turnover
- Variety of acts and scenarios covered
Value For Money
Are you getting fair content per dollar — and can you leave easily?
- Effective monthly cost on the cheapest legitimate plan
- Cost per video (library size ÷ annual price)
- Trial availability and trial-to-paid pricing transparency
- Published cancellation terms, without assuming the actual flow is easy
- Disclosed checkout upsells, when visible before payment
Site Design
How does the actual product feel to use?
- Public tour navigation and visible filters
- Streaming reliability only when member access is documented
- Public mobile tour; member playback only when directly tested
- Visual design and information density
- Published account features, with member-only functions marked unverified
Update Frequency
Is the site still actively producing — or coasting on archives?
- Dated releases visible on the public tour, where available
- Consistency of upload schedule
- Newest scene date vs. site age
- Recycled vs. genuinely new content
- Network update spillover (for network memberships)
The maths
How the overall score is calculated
Each pillar is scored 0–10, then multiplied by its weight (Content Quality 35%, Value for Money 25%, Site Design 20%, Update Frequency 20%). The four weighted scores are summed to produce a 0–10 overall score, which is converted to a 5-star display by multiplying by 0.5. Pillar scores shown on review pages are rounded to the nearest 5 for visual clarity — the underlying calculations use the precise scores.
A site needs to score consistently across all four to land above 4.5/5. A 95 in content can't carry a 60 in updates — the weighting is deliberately balanced so a single strong pillar doesn't whitewash weakness elsewhere.
Zero weight
What's not in the score
- Affiliate commission rate. Has zero weight. A site I don't earn a cent from can outrank a site paying 70%, and several do.
- Marketing claims. A publisher's "4K HDR" or "weekly updates" claim is labeled as such until it can be checked. Public release dates are evidence of visible activity, not proof of member playback quality.
- My personal taste. A niche site I'm not into the content of can still score well if it executes the niche cleanly.
- Tour-page promises. A public tour does not prove what a paid member can actually access. Member-only claims remain qualified until directly tested.
The reviewer
Who writes these reviews
TwinkVault is operated by Isaac in the US. Current review profiles are research-based and use the evidence described above. Reach me at hello@twinkvault.com if you spot something I got wrong, want to suggest a site, or work in the industry and want to share context I should know.
Special cases
Edge cases
A few things that come up often:
- Network sites (Say Uncle, Buddy Profits, ASGmax, and similar) are scored at the network level when one subscription gives access to all sites in the network. Individual profiles describe each site's content focus and access differences; the shared score reflects the network value.
- Pricing changes are added with a check date when an official join page can be verified. Older figures remain historical snapshots until rechecked; a changed price alone does not imply a membership was retested.
- Future paid or press access will be disclosed on the specific review only after it is documented. Current review records do not claim such access.
- A site that relaunches with substantially new branding or ownership is treated as a new review. The old review is archived with a redirect to the new one.
Change log
Changes to this methodology
The rubric is stable but not frozen. Material changes are noted here with the date.
- September 2026: Corrected unsupported claims of paid testing; current review records are research-based and member-only observations are unverified.
- May 2026: Added pillar-score display rounding (nearest 5) for visual clarity.
Questions
FAQ
The questions people ask about how the scores work.
Why these four pillars?
Content, price, usability and update pace are the practical factors behind a membership decision. Public research can estimate them, but member-only details stay unverified until direct access is documented.
Why is content quality weighted highest?
Because if the content's bad, nothing else saves the subscription. A beautifully designed site with a thin, recycled library still isn't worth paying for. Conversely, a great content library forgives a lot of UX sins.
Why is design only 20%?
Most paying users adapt to a clunky UI within a week. Bad search is annoying, but it doesn't make the content worse. I weight what affects the actual value of the subscription more heavily than what affects the first-impression.
Do scores ever change after publishing?
Yes. New, checkable pricing, catalog or ownership information can change an assessment. A page's last-updated date records an edit; it does not mean that a membership was retested.
Can I see the actual numbers behind a site's score?
Each individual review page shows the four sub-scores out of 100 alongside the overall score. The score breakdown card on every review is exactly the data this methodology produces — nothing hidden.
Scored on this rubric
See it applied
Every site below was scored with the four pillars above. Tap one for its full breakdown.
