binary comparisons
users are presented with two options and must choose a winner. it is fast, intuitive, and drastically reduces cognitive load.
in development
rankd.
five-star rating systems are fundamentally broken. users tend to rate items as either one star or five stars, destroying any nuance. rankd is a rating engine that fixes this by forcing binary choices.
building for Androidrankd is currently in active development. we are building it as a native Android application using Flutter. the core elo engine is functional, and we are refining the user interface.
the binary advantage
when presented with a one-to-five scale, it becomes a measure of extreme emotion rather than objective quality. average items get inflated, and polarizing items get buried.
rankd forces binary choices. by asking users to simply choose which of two items is better, we eliminate rating inflation. over thousands of comparisons, a mathematically sound hierarchy emerges.
how it works
elo-based ranking
we use battle-tested algorithms to adjust scores based on the relative strength of the matched items. beating a top-tier item matters more.
custom datasets
create your own categories. rank anything from mechanical keyboards to espresso machines using the same underlying math.
fraud prevention
statistical models detect and ignore bot voting and coordinated review manipulation. the math protects the integrity of the list.
badges + tiers.
your ranking earns recognition. public, fun, a little competitive.
the math behind the rank
the core of rankd is the elo rating system, originally designed for chess. when an item wins, it gains points; when it loses, it drops. defeating a highly-ranked item yields more points than beating a low-ranked one. this self-correcting math ensures the best items naturally float to the top, regardless of when they were added to the system.
what is next
core algorithm
the math engine and database schema are locked in and tested.
Flutter client
building out the native Android interface for fast, swipe-based voting.
public beta
opening the platform for users to submit their own datasets for ranking.