Here's a puzzle every rating site has to solve, and most solve badly.
Track A has five votes, all of them SLAP. A perfect 100%. Track B has a thousand votes, and 900 of them are SLAP — 90%. Sort the leaderboard by percentage and Track A wins. Does anyone actually believe Track A is the better song?
Five people is a group chat, not a verdict. A thousand people is a crowd. The whole value of a crowd rating is confidence, and a raw percentage throws confidence away — it treats "we barely asked anyone" and "we asked everyone" as the same thing.
What we do instead
Our leaderboards rank tracks by something called the Wilson score lower bound. The name is dry, but the idea is intuitive: instead of asking "what percentage did this track get?", we ask "given how many votes it has, what's the worst its true approval could plausibly be?" — and we rank by that pessimistic answer.
- Track A, 5 for 5: the honest answer is "we don't really know — could be anywhere from mediocre to great." Its lower bound lands around 57%.
- Track B, 900 of 1,000: the answer is "we know pretty precisely — it's around 90%, give or take." Its lower bound is roughly 88%.
Track B wins, and it should. The magic is what happens next: if Track A keeps collecting SLAPs, its uncertainty shrinks and its score climbs. Nothing is being held against the new track except the fact that we haven't heard from enough people yet. Evidence is the only currency.
The Hall of Slop is harder to enter than you'd think
The same logic runs our Hall of Slop — the chart of tracks the crowd rejects — and this is where it matters most. The Hall of Slop is not the slap chart read bottom-up. It runs the same confidence math with SLOP as the "success."
That means a brand-new track that catches two early SLOPs can't crash into the Hall of Slop on bad luck. To earn a spot there, a track has to be confidently disliked: lots of votes, mostly negative. Getting into the Hall of Slop is, in its own miserable way, an achievement.
The other guardrails
The Wilson score does the heavy lifting, but two simpler rules back it up:
- A ten-vote floor. No track appears on an all-time chart until at least ten people have weighed in, full stop.
- One vote per person per track. Coming back to smash SLAP again doesn't count twice — you'll just see your existing vote.
None of this is exotic. Reddit uses the same Wilson ranking for comments, and it's the textbook fix for exactly this problem. What's rarer is saying so out loud: most rating sites won't tell you how their charts are sorted. Ours is documented in full, and the votes behind every number came from people like you.
So when a track sits at the top of our charts, it means something specific: a lot of listeners heard it, and most of them hit SLAP. That's the whole promise of this site — and it's why the percentage you see is the start of the story, not the end of it.