AFL Live Ladder

What is Luck? First, it's the thing that the Live Ladders model attempts to subtract from the scoreboard result, to find an underlying margin. Second, it's what propelled the tipping model to 156 this year, 20 ahead of the bookies. 24 of 29 @SquiggleAFL models beat them. [1/3]

@MatterOfStats Yes. Motivated by the error off the boot following some distribution with that angle as the parameter. Fatter tails than normal, not quite Cauchy! Interested to see what transformation works best for you, I'll let you know mine soon

To my xScore-creating friends, for a set shot, is xScore directly proportional to the apparent goal angle or some transformation of it?

@EmlynBreese @FootyTips *Joly, and yes Kornes is not a journalist or a voice I would put trust in

Have now had Christian Jolly from CD confirm Kane’s story is bullshit on @FootyTips pod. Decision initiated by CD because the value of the data in the feed didn’t match the complexity of recording it (esp multi-umpire input into decision).

11 teams currently sitting on four premiership points. The last time we had more than 10 teams on the same points, Gold Coast didn't exist but the Bulldogs and Swans were the top two, and Brisbane led the small contingent of winless teams on percentage #2" target="_blank" rel="nofollow noopener">afltables.com/afl/seas/1998.…

Bit more work on this project, inside 50 visualiser - loaded up with the data from #AFLDeesSaints

@EmlynBreese If you exclude intercept marks, 34 is the equal highest since 2010 with that WBD v WCE game. R9 2009 RICH v ESS there were 35 and Nathan Brown took 8 inside-50 marks for 3 scores?! afltables.com/afl/stats/game…

Back to 2017 (as far as I've got this data for) WBD v WCE 2018 R02 - 35 (18 to WBD) NTH v STK 2018 R23 - 35 (23 to NTH) Most of these have been blowouts. NTH v STK was a bit of a shootout 117-94 PTA v SYD 2025 was a ludicrous (given total 37 f50 marks) 52-71 (inc 4 rushed)

Meanwhile, Sydney genuinely spifflicate Carlton and Brisbane, the latter despite losing clearances by 22. Ranking improves from 11th to 3rd in two weeks, the biggest re-evaluation we've seen in a long time

Bulldogs 2025: KPIs have them top 4 but lose narrowly to top teams 2026 OR: concede 124 points worth of shots and win because Bris can't kick straight R1: win by 81. Only +25 equity despite +19 clearances including 5-0 F50 throw-ins for 3.1 Computer downgrades them each week

@maxlaughton Round 1 is also when defensive structural flaws get exposed. Clubs then spend the next week or two intensively patching them to limit the damage. Let's see what Round 4 looks like

Still very early so I wouldn’t want to overreact, and we need to see more good team vs bad team games. But surely there’s a chance all the rule changes to try and make the game faster are just making the good teams better and the bad teams worse? By definition speed should make defence harder, and defence is usually the first thing bad teams try and fix. The NRL had something similar in 2021 as a few years of rule changes to try and speed the game up, when added together, saw record levels of blowouts.

Four out of the six games have been won by the team with the lower shot equity. The sample size is in! Luck is the new hottest stat teams are training to win. -- afl. com. au

@dowhah @SquiggleAFL Don't make me tap the pinned tweet

@AflLadder @SquiggleAFL The bookies usually get it right.

SCG, Thursday March 5: Day Zero of Round Zero. What a time to be alive! Curiously on Sunday, Collingwood is a $2.02 outsider against St Kilda. Every @SquiggleAFL computer model tips Collingwood, forecasting them as a top-8 team. St Kilda's highest projection is 12th

Nick Daicos (95 games) is the 17th most experienced player in Collingwood's named team, and 20th on their list

@simonsbargains @SquiggleAFL Nah, you never know when your edge will actually pay off. Typically 70%+ of our profits would come from a seven-round block when the computer had evidence that wasn't priced into markets. That block could be start, middle or end of season; you never knew when the luck would come

@AflLadder @SquiggleAFL Form in the first 3-4 weeks of a season is a raffle. Smart gamblers would be best to hold off until at least April, when some real form lines start to emerge.

@AFLLab Some of us are thrilled by a team scoring 70 points in a quarter

AI just gets footy, a new era has arrived

Urk looking at this in the daylight. Why is Edge dark mode so bad, and Windows Snipping Tool worse? Better screenshot here

19 days to go, and we have Wildcard projections

‘You need 11+ wins to make the ten’ is the new ‘you need 12+ wins to make the eight’. This article exists so my fellow media members, and coaches, can update our cliches. @FOXFOOTY foxsports.com.au/afl/afl-2026-h…

ICYMI the game started at 5pm and Canberra didn't score until after 9pm. We were there at kickoff and final whistle, but managed a four-course meal in between down the road #truestory

C'mon @NYTGames, it's the off-season and your puzzles shouldn't disturb us like this

And Sydney win the flag from 10th - now possible in the 2026 Squiggle Ladder Predictor. predictor.squiggle.com.au