Who actually makes money on a prediction market?

“The easiest money on the internet.” That’s the pitch, and it’s everywhere right now — billboards, timelines, cable news, some guy insisting he makes all his money on one of these apps.

So I went and read what actually happens to the people who show up. Not the ads — the research. The studies of who wins and who doesn’t on the biggest real-money prediction markets.

It’s not what the billboards say.

Almost nobody comes out ahead

Start with the number that matters: on the biggest of these markets, about seven in ten people lose money.

If that still sounds better than a casino — you’re right, it is. But look closer at the three in ten who come away with anything and it falls apart. The money isn’t spread across them. It’s stacked at the very top: a sliver of people — the top one percent — take more than three-quarters of the money that gets made. Everyone else splits what’s left.

So the pitch isn’t a lie, exactly. Money is being made. Just not by most of the people making it possible.

So who are the few?

The researchers sort the winners into three kinds, and none of them is you on a Tuesday.

People who already know. Someone holding a piece of information the rest of the room doesn’t have yet — the result that hasn’t gone public, the thing about to be announced. They show up, put their money on the sure thing, and leave. It’s a small share of the whole, but it’s real, and it’s exactly what it looks like.

A small group who are simply good. Around three in a hundred. They read faster, they move on news before the room does, and — this is the part worth sitting with — they win partly by not doing what everyone else does. The average person leans on the obvious side: the one coloured green, the one listed first, the cheap long shot that feels like a steal. Those habits quietly bleed money. The good ones don’t have them.

The professionals on the other side. Every market needs someone willing to take the other side of whoever walks in, or there’s no market at all — and firms do that as a job, quoting both sides all day. A fraction of a percent of the accounts, a slice far bigger than that: the going rate for a room that’s always open.

Add it up and it’s the picture the studies keep finding: a tiny group at the top, and everyone else feeding it.

The part that stuck with me

Here’s the twist I didn’t see coming.

Prediction markets really are pretty accurate — that half of the pitch holds up. But when researchers went looking for where the accuracy comes from, it wasn’t the crowd. It was that same small group of sharp readers, dragging the number toward the truth.

Which means the majority isn’t producing the accuracy everyone quotes. It’s funding it. Every time an average player reads it wrong and loses, they’ve paid a sharper one to be right — and paid for the tidy, accurate number the news repeats the next morning.

That’s the whole machine. A school where most people pay tuition, a few collect it, and nobody ever hands the payers a grade. You lose, you decide you got unlucky, you go again. The most expensive way to learn you’re average is one loss at a time.

So we left the money out

None of this is a knock on those markets. They work. They’re accurate. But you should know which seat you’re in before you sit down.

And it’s most of the reason majjha.fun is built the way it is. There’s no money in it — none in, none out. Which quietly deletes the whole trap: if there’s nothing to lose, nobody’s paying anyone’s tuition, and nobody across the table needs you wrong. Nobody’s rooting against you because there’s nothing to root for.

What’s left is the only part that was ever worth having — the skill. Reading a real question a little better than the room is learnable, and here it’s the one thing you get to keep. We just hand you the grade the real markets never do. It’s the record, and it’s yours.

So find out how well you actually read the world. It won’t cost you a thing to learn the answer.

Don’t pay tuition. Go build a record.