Built an AI-ranking engine for sweeps casinos — methodology + sheet link
March 15th, 2026 at 6:25:52 PM
permalink
alright, for the nerds (said with love) 🎰
i wanted a way to objectively rank sweepstakes casinos so i built a scoring engine that handles everything autonomously.
here's how it all works:
8 scoring dimensions, each weighted:
each dimension is normalized and fed into a weighted composite. the weights aren't equal — trustpilot and payout speed carry more because imo those are the dimensions where sites diverge the most and where players actually get burned
output:
the autonomous piece: the engine scrapes fresh trustpilot data every 5 days, recalculates all composites, and re-ranks the entire list — zero manual intervention. this isn't a snapshot. it's a living ranking that reflects current sentiment. check it today vs. 5 days from now and you might see movement
55 sites scored so far. the spread is significant — top tier composites are nearly 2x the bottom tier, and you can see exactly where each site gains or loses points
View the Full Rankings Spreadsheet:
open to feedback on the weighting. if someone has a compelling argument for why a dimension should carry more or less, i'm all ears. the whole point was to remove subjectivity but the weights themselves were a judgment call 💎
don't hesitate to reach out if you want to dig into how it works further ✅
i wanted a way to objectively rank sweepstakes casinos so i built a scoring engine that handles everything autonomously.
here's how it all works:
8 scoring dimensions, each weighted:
- • trustpilot score — scraped programmatically via their public pages, refreshed on an autonomous 5-day cycle. not self-reported, not cherry-picked
- • payout speed — categorized by reported redemption windows
- • welcome offer value
- • daily bonus structure
- • game count
- • referral program quality
- • minimum purchase threshold (lower = better for the player)
- • crypto support (binary + which coins)
each dimension is normalized and fed into a weighted composite. the weights aren't equal — trustpilot and payout speed carry more because imo those are the dimensions where sites diverge the most and where players actually get burned
output:
- • composite score per site
- • 4-tier classification (🟢🔵🟡🔴) with natural breakpoints in the distribution
- • 🏆 badge winners across all 8 categories
the autonomous piece: the engine scrapes fresh trustpilot data every 5 days, recalculates all composites, and re-ranks the entire list — zero manual intervention. this isn't a snapshot. it's a living ranking that reflects current sentiment. check it today vs. 5 days from now and you might see movement
55 sites scored so far. the spread is significant — top tier composites are nearly 2x the bottom tier, and you can see exactly where each site gains or loses points
View the Full Rankings Spreadsheet:
open to feedback on the weighting. if someone has a compelling argument for why a dimension should carry more or less, i'm all ears. the whole point was to remove subjectivity but the weights themselves were a judgment call 💎
don't hesitate to reach out if you want to dig into how it works further ✅
March 15th, 2026 at 6:34:48 PM
permalink
For some reason the link isn't showing up in the post:
It's in Google Docs @:
/spreadsheets/d/17fE_BVBn1-rJo2cD66c4LCE2t1F5OyKUjrAhX2sA-Sg/edit?usp=sharing
It's in Google Docs @:
/spreadsheets/d/17fE_BVBn1-rJo2cD66c4LCE2t1F5OyKUjrAhX2sA-Sg/edit?usp=sharing
March 15th, 2026 at 8:02:24 PM
permalink
Quote: GARRETTG85For some reason the link isn't showing up in the post:
link to original post
Indeed. New members aren't allowed to post links during the probationary interval.
This greatly reduces spam.
May the cards fall in your favor.
