There's a HUGE problem in sports betting in Australia.

Not because there's too little information. There's too much of it, and almost none of it is shown in a way that tells you what it actually means.

Every bookie app has a stat next to the price. Every group chat has a screenshot of a player's last five. Every tipster has a hit rate. And somehow, with more numbers in front of us than any generation of punters before, most of us are still reading them like a form guide from 1995.

If you've ever missed an absolute dreamer of a multi by one leg, you're in good company. We aren't here to talk down to anyone. We've been there more times than we'd like to admit. We've tailed the mate who was “on a heater”. We've backed the leg that “hasn't missed in weeks”. And when it fell over, we did what everyone does: blamed the player, blamed the tipper, blamed the umpires, and went looking for the next one.

What we didn't do, for a long time, was ask whether we'd read the number properly in the first place.

$33 billion

Australians lost $33.0 billion gambling in 2024–25. That's $1,528 for every adult in the country, according to the national figures every state and territory reports into.

Sit with that for a second. We're a country at the bottom of the map, a long way from almost everyone, and Australia is regularly reported as the world's biggest gambling loser per person. That's not a stat about bad people. It's a stat about a lot of good people, doing something together, mostly blind.

Most of that is pokie machines: $17.4 billion. But wagering, which is betting on racing and sport, was $8.5 billion of it, or about $391 per adult. That's the part we're talking about here.

We're not quoting that to wag a finger at anyone. We're quoting it because a number that big means a lot of people are making a lot of decisions, and the tools most of them use to make those decisions are a headline stat and a hunch.

The group chat

Australians punt the way we do most things: together. A multi gets built in the group chat, a screenshot gets passed around, somebody's slip gets copied five times before the first bounce. It doesn't matter where your family's from or which team you go for; the group chat is where the footy happens. That's half the fun, and there's nothing wrong with it.

Where it goes wrong is tailing without looking, and then blaming the player or the tipper when it misses. That's not how you treat a mate. The player didn't promise you anything. The tipper showed you a number. Nobody in that chain checked what the number was actually saying.

And most of the time, the number is some version of this one.

“8 of the last 10”

It's the most convincing sentence in Australian punting. It sounds like 80%. It sounds like a lock.

Here's a real one. Going into Round 24 this year, Nick Daicos had gone over 34.5 disposals in 8 of the last 10 games. The last ten read 35, 41, 37, 31, 28, 36, 39, 35, 39, 36. That's an average of 35.7 and a medianThe middle game once the ten are sorted. of 36, comfortably over the line.

Daicos had 26.

Now stretch the window, and watch what 8 of 10 turns into.

Same player, same line, same numbers. Nothing's made up. The only thing that changed is how many games we looked at. 8 of 10 was true. It just wasn't the whole truth.

And the game itself? Collingwood lost to the Brisbane Lions by 63 points, and Daicos spent 62% of the game on the ground. In 2026 before that night, the lowest share of any game Daicos had played was 80%. A last-10 hit rate can't see that coming. It only knows what already happened.

It wasn't just Daicos

One player can be bad luck. So we checked every line.

From Round 22 through the preliminary finals this year, bookmakers offered 1,623 AFL player lines where the player had cleared that exact line in 8 of the previous 10 games. That's disposals, kicks, handballs, marks, tackles, clearances and goals, from 27 games.

If 8 of 10 meant 80%, about 1,300 of them would have come in.

1,119 did. That's 68.9%.

The prices didn't flag it either. Before the bounce, the bookmakers' prices on those same 1,623 lines implied, on average, an 81.6% chance. Part of any price is the bookmaker's margin, so a price always reads a little higher than the real chance. But 81.6% against 68.9% is the whole gap between what 8 of 10 looks like and what it did.

On disposals, the market most multis are built from, it was 18 of 46.

Sometimes it's just cruel. In a preliminary final, Will Ashcroft had cleared 28.5 disposals in 8 of the previous 10 games. Ashcroft had 28. One touch.

Three ways to read a number

None of this is a system. It's how to read what's already in front of you, and it's the stuff we wish someone had told us years ago.

Stretch the window

Ten games is a small sample. A player's last ten includes a hot patch almost by definition, because hot patches are what get noticed. The longer the record, the harder it is for one good month to fake it.

Look at the spread, not the average

An average is one number trying to describe twenty games. One 43-disposal day drags it up; a 62% time-on-ground night drags it down. Seeing every game as its own bar shows you how often the line gets cleared, and by how much, which is the actual question.

Turn the price into a percentage

Divide 1 by the decimal odds. Daicos over 34.5 was around $1.81 across four bookmakers. 1 divided by 1.81 is about 55%. So before a ball was bounced, the price was describing a 55% chance, not 80%. It's worth knowing what the price is saying before deciding what the stat is saying.

What this doesn't tell you

These numbers show what happened. They can't tell you what happens next, and nothing can. Our line-by-line check covers six weeks of AFL, 27 games from Round 22 to the preliminary finals, which is a small window of its own. Next, we want to run the same test across a full season, and across the NRL and the WNBA, to see whether 8 of 10 says more in some sports than others.

Why we built Project Punt

We didn't start Project Punt because we'd figured betting out. In fact, it's the opposite: we WANTED to figure out betting. We started it because we kept making the same reading mistakes. We made it because we were sick and tired of flicking from tab to tab and eventually coming back to the same conclusion we had at the very start. We made it because we felt that Australians are victims of their own judgement, clouded by the “ball knowledge” we're all so proud of.

The information to catch those mistakes was already out there. It just wasn't in one place, and it was never drawn so you could see it.

So that's what we do. Every game, every line, every window, shown as evidence. No tips, no locks, no “nuke this”. We're punters too, and we're on your side of the counter. We'll show you the whole picture, and what you do with it is your call.