Grand Final, 2026

Congratulations Drop Kick Data for winning the Squiggle Overall Comp in 2026 with 163 correct tips. This is the same number as runner-up Holy Grail Ratings, but here at Squiggle we tie-break on Bits, and Grail is very conservative, always happy to call a contest a 50/50 toss-up, which means it doesn’t get penalised much for bad tips, but doesn’t get rewarded much for good ones. Drop Kick Data won by backing itself in.

Of course, Tips is the crowd-pleaser award; what the very serious models care about is MAE and Bits, where it’s harder to fluke a lucky year (or be cruelled by an unlucky one). Here, wouldn’t you know it, it’s Wheelo Ratings again! That makes it back-to-back MAE victories for Wheelo, to go with a third-place finish in 2024. A heck of a track record. Our MAE runner-up (and the only other model to beat the hive mind’s collective wisdom in s10) is AFLalytics.

On Bits — not as luck-based as Tips, not as boring as MAE — our 2026 champion is Don’t Blame the Data, with the venerable Matter of Stats runner-up. This is the same 1-2 as last year’s Overall Comp… which is enough to make you think there’s not that much luck involved in Tips after all.

Squiggle Leaderboard 2026

Thank you to all 28 models who competed in the Squiggle comp this year!

I’m a big fan of the new Final Ten. I want to acknowledge this, because when the AFL gets something wrong, I’m right there, ready to tip the bucket. But I like the Final Ten a lot.

• The Final Ten is a lot like the Final Eight, plus a special introductory fun extra week. When I heard the AFL wanted to rejigger the finals system, I had concerns. Because the Final Eight is fantastic. I thought the AFL would do that thing where they change a thing and then it turns out that thing doesn’t work in practice quite like they thought and now it needs another thing, and then another, until it’s hard to see how the end result is substantially better than the original problem that allegedly needed fixing. There are many bad finals systems around, but the Final Eight isn’t one of them. It avoids dead rubbers; it effectively rewards teams for Home & Away achievements without tilting the balance too far; it offers a series of matches between closely-matched teams; it lets fans know what their teams are playing for before they play for it. Terrific stuff. I’m glad we kept all that.
  
• It is genuinely more interesting for low-mid-table teams. The knock, of course, was that letting 9th and 10th into finals — excuse me, into “wildcard finals;” more on the naming scheme later — would devalue the whole thing. But we were already scoffing at 7th and 8th, saying they were just making up the numbers, so extending that to 9th and 10th is no real difference. In return, we give teams around 10-13th something to play for late in the season. Sure, it’s the chance to make a I-can’t-believe-it’s-not-really-butter final, and winning that will only offer the opportunity to be beaten by better teams later — but we’re delusional football fans; we will believe anything is possible. And if it’s not, well, our team is still alive while other teams’ seasons are over. That’s exciting all by itself, and worth achieving, so it grants meaning to what would otherwise be dead rubber late-season games.
  
• We had to do something about the pre-finals bye. It was a terrible lull just when the season should be building to a climax. At least now there are two games of football in it.

Obviously the naming is stupid. They’re “wildcard” finals even though they have no apparent connection to wildcards, and the AFL is trying to claim simultaneously they’re finals but not part of “the finals.” The official AFL ladder now has a dividing line between the Top Ten and the rest, but also a special line after the Top Six, apparently to show those teams have the benefit of no knockout final first-up — even though the Top Four have the exact same benefit relative to the teams below them via the double chance, but they don’t get a line.

But the system! The system is good.

A few little changes to the Squiggle API, which is used by bots and scripts who want basic data on teams, games, and scores. Because sometimes they want it very badly, and aren’t afraid to crash my site to get it.

The Event API, which is the part that gets you updated score info as quickly as possible, has been shifted to a new sse.squiggle.com.au endpoint that puts it out of the way of regular site traffic, so bots can’t sit there occupying all the site’s available connections. It is also more generous with connection limits, so more people can use it.

And the regular API has a cache in front of it, which means it’s slightly less useful for up-to-the-second info — which is fine, since it’s not the Event API — but also less likely to crash my site when a bot decides it needs to redownload my entire database over and over every few seconds.

Two more new models sneaking in just ahead of the bounce (bounce?): Holy Grail Ratings and In The Game.

That makes 28 models + 2 meta-models (Aggregate and S10) + Punters. This competition is so hot right now.

This year’s newest model is HBin, and if you look up the website, you’ll see this stands for Human Behaviour Insights. It is just a coincidence that the author’s name is Hamish Binnie.

If you believe that, you are probably in need of the kind of insights you can get from following HBin.

We’re in the final hours before the start of season 2026! We have 21 models good to go, plus a handful either putting on the final touches or having crashed and burned in the off-season. Time will tell.

You can’t develop a game all by yourself, they said. The market for video games based on Australian Rules Football is too small, they said. Don’t you have a real job you’re supposed to be doing, they said. Well HA! Check it out!

As of Aug 14, 2025, the game is now available to purchase on PC, Mac, Linux for A$24.95. If you get in quick, there’s a launch discount.

Buy Squiggle Football on Steam

Please show your support by buying the game, playing the demo, or wishlisting! And tell your friends.

Special Thanks: This game has had exactly 1 play-tester, my friend Wild Pete. I just checked and he has 208 hours played. So either he’s very dedicated or he really likes the game. Maybe both. Thanks Wild Pete! I made lots of great changes thanks to your constant criticism.

One week out! Wishlist now for launch discount.

You can now play the first two seasons of Squiggle Football on PC, Mac & Linux.

Or you can just wishlist it! Then you get notified when the full version drops with a launch discount in August, while also helping my stats. Everybody’s happy.

Squiggle Football on Steam – demo

Coming August 2025 for PC, Mac + Linux!

WISHLIST NOW ON STEAM