Stratatube allows you to expose the layers beneath the standard YouTube algorithm. This tool helps you explore content more deliberately by using pattern-based searching to find hidden videos that exist beyond traditional rankings and recommendations.
Instead of an algorithm deciding what you see, Stratatube uses patterns, reusable search templates you can customize before running a search. Patterns help surface videos the recommendation system tends to ignore, including uploads with minimal descriptions, default filenames, or little engagement history. You can reuse patterns that work, share them with others, or create your own. Not every video will be interesting, some are quiet, some are odd, but you're seeing things most people never will.
Stratatube turns pattern-based searching into something playful by using a slot machine style interface. Instead of typing exact keywords, you spin combinations of eras, patterns, and parameters drawn from real upload behaviors like time periods, naming habits, or metadata quirks. It is not randomness for its own sake. Each spin builds a valid, inspectable search that you can tweak before running. The result feels more like exploration than querying, lowering the barrier to discovery while still giving you full control over what you are looking for.
"Sifting" is another of Stratatube’s exploration methods and is a great way of letting nothing go to waste. When someone runs a search, only a few results are usually clicked, but the rest still hold potential. Sifting turns those unused results into a continuous, scrollable feed where users can watch, evaluate, and save videos over time. It borrows a familiar scrolling mechanic, but the goal is slower and more intentional viewing. Instead of judging videos by thumbnails or titles, you discover their value by actually watching them.
After exploring or sifting, picking is where discovery turns into curation. When you find a video worth keeping, you can pick it and vote on it alongside others. Picks surface videos that were overlooked at upload and ignored by "the algorithm", often sitting unseen for years. We're creating a human-curated layer on top of YouTube that rescues quiet, early, and accidental uploads, guided by collective interest rather than performance metrics.
