afl-tips.com

AFL Predictions 2027

One chart shows ratings for all 18 clubs, with attack across and defence up. Pick any season from 2017 onward and slide through it to watch the contenders pull away.

2027 · Pre-season
-40-40-30-30-20-20-10-1000+10+10+20+20+30+30Attack: points scored above average →↑ Defence: points kept outStrong both endsWeak both endsADEAdelaide: attack +6.2, defence +0.5BRIBrisbane Lions: attack +17.4, defence -2.6CARCarlton: attack +2.2, defence +3.1COLCollingwood: attack +3, defence -2.1ESSEssendon: attack -10.7, defence -11.7FREFremantle: attack +5.7, defence +7.1GEEGeelong: attack +13.4, defence +0.6GCSGold Coast: attack +0.5, defence -2.4GWSGreater Western Sydney: attack +3.1, defence -3.3HAWHawthorn: attack +7.1, defence +4.4MELMelbourne: attack +7.3, defence -2.3NTHNorth Melbourne: attack -1.1, defence -6.2PTAPort Adelaide: attack -6.2, defence -5RICRichmond: attack -13.1, defence -8.7STKSt Kilda: attack -0.5, defence -1.6SYDSydney: attack +13, defence +1.5WCEWest Coast: attack -10.8, defence -10WBDWestern Bulldogs: attack +0.1, defence +1.7

Since 2015, every AFL result in the Squiggle API has fed into our rebuilt ratings. Each club has a dot, and a dot higher and further right is better.

How the flag race shifted in 2026

After the fact, our model replayed the 2026 season round by round. It shows the shape of the race rather than the tips we published.

The flag chance

0%10%20%30%40%PreR3R7R11R15R19R23ADEBRIFREGEEHAWSYD

The ladder tower

141018PreR3R7R11R15R19R23ADEBRIFREGEEHAWSYD

Round winners

Each result changed a club's season: its chances of a top 4 finish, the finals and the flag, plus projected wins and projected finish. The largest gain leads the list.

ClubResultTop 4FinalsFlagProj. winsProj. finish
BrisbaneBeat Collingwood 159–96100% ▲ 47.4100% –22.2% ▲ 12.116.0 ▲ 0.53.0 ▲ 1.5
HawthornBeat West Coast 107–45100% ▲ 17.2100% –15.3% ▲ 2.816.0 ▲ 0.24.0 ▼ 0.2
AdelaideBeat GWS 157–1010% ▼ 0.8100% –2.8% ▲ 1.815.0 ▲ 0.36.0 ▲ 0.6
SydneyBeat North Melbourne 123–70100% –100% –26.8% ▲ 0.818.0 ▲ 0.12.0 –
CarltonBeat Fremantle 119–820% –100% –0.6% ▲ 0.212.5 ▲ 0.610.0 ▼ 0.3
EssendonLost to Port Adelaide 95–1040% –0% –0.0% –2.0 ▼ 0.418.0 ▼ 0.4
Gold CoastBeat St Kilda 103–800% –0% –0.0% –9.0 ▲ 0.713.0 ▲ 0.7
GWSLost to Adelaide 101–1570% –0% –0.0% –10.0 ▼ 0.312.0 ▼ 0.1
North MelbourneLost to Sydney 70–1230% –0% –0.0% –9.0 ▼ 0.114.0 ▼ 0.7
Port AdelaideBeat Essendon 104–950% –0% –0.0% –7.0 ▲ 0.415.0 –
RichmondLost to Geelong 76–1330% –0% –0.0% –3.0 –17.0 ▲ 0.4
St KildaLost to Gold Coast 80–1030% –0% –0.0% –10.0 ▼ 0.711.0 ▲ 0.1
West CoastLost to Hawthorn 45–1070% –0% –0.0% –4.0 ▼ 0.216.0 –
Western BulldogsBeat Melbourne 108–870% –100% –0.2% –13.5 ▲ 0.78.0 ▲ 1.1
CollingwoodLost to Brisbane 96–1590% –100% –0.2% ▼ 0.613.0 ▼ 0.59.0 ▼ 0.9
FremantleLost to Carlton 82–119100% –100% –23.7% ▼ 7.519.0 ▼ 0.61.0 –
GeelongBeat Richmond 133–760% ▼ 22.3100% –7.7% ▼ 3.715.0 –5.0 ▲ 0.1
MelbourneLost to Western Bulldogs 87–1080% ▼ 41.5100% –0.4% ▼ 5.915.0 ▼ 0.77.0 ▼ 2.0
Our model replayed 2026 Round 24. Arrows show how things changed since before the round. For finish, an upward arrow means a higher position.

How the model makes tips

Each club has two numbers. One shows points scored above an average side. The other shows points kept out. To get a game's expected score, add one side's attack to the league average, then subtract the other side's defence.

The home side gains an edge by playing where it plays most often, and by making the other club travel interstate. Once a game is done, both ratings shift a little toward what really happened; in the off-season each rating is drawn back toward the middle, since lists change.

An expected margin becomes a win chance. The ladder and flag odds come from 20,000 simulated seasons of the fixture. It was tested on 2,476 games played between 2015 and 2026, picking 68.3% of winners correctly, with an average margin error of 27.0 points.

Scoring for tipsters

Tips

Each correctly tipped winner earns one point, using the same method the Squiggle API applies to every model.

Bits

Correct confidence earns rewards, while incorrect confidence is punished. A tip at 50% brings zero points; a right call at 80% earns about 0.68, and a wrong call at 80% loses about 1.32. Across a season, this shows who knows how confident they should be.

The margin error

Margin error is the point gap between the tipped margin and the actual margin. This is averaged across the season. A lower error is better. In 2026, the strongest models were near 24.6 points.

See every model graded