Footy Tip Jar tips Fremantle to beat Brisbane by 10 points in the 2026 Grand Final at the MCG on Saturday, at 62 percent, with a projected score of 111 to 101. That is the headline. The rest of this preview is about why, because the two sides arrive rated first and fourth in the model, both ratings jumped after the preliminary finals, and across the six things the model measures they are closer to level than a 10 point margin suggests.
If you joined us during the Brownlow count, welcome. This is the weekly preview, written from the same model that produced the vote guide, and this week it has one game to think about. The results report for the count stays free at footytipjar.com/brownlowresults.
The tip in one paragraph
Fremantle by 10, 62 percent, 111 to 101. The model's fair price is $1.53 for Fremantle and $2.53 for Brisbane. The confidence label is moderate, which is where a Grand Final between the two form sides of September should sit. The MCG is a neutral ground for both clubs, so neither side receives a home ground credit, and both named 23s on Thursday night are priced as named.
What the model is actually measuring
The match model rates every team on six domains, each out of 100, updated after every game. Contest is winning the ball at stoppages: contested possessions, clearances and hitouts. Territory is moving it forward: inside 50s and metres gained. Retention is keeping it, measured through turnovers. Chance creation is scoring shots generated. Finishing is conversion. Pressure is tackling and pressure acts.
They are not weighted equally. Contest, territory, retention and pressure carry the tip. Chance creation carries a little. Finishing carries none in the base rating, because conversion is the noisiest thing a team does from week to week. Since June the team power part of the score has read those domain levels together rather than adding them up, which is why two sides can look level on a game by game split and still be 10 points apart when the model prices the whole of what each one is.
Where the two sides stand entering the game
Fremantle is rated first in the competition for territory at 87 and third at the contest at 77. Brisbane is third for territory at 74 and sixth at the contest at 65. Those two domains are where the model finds most of its margin.
Brisbane's answers are retention and chance creation. It keeps the ball better, 60 against Fremantle's 51, which is fifth in the league against tenth. It creates chances at 78, the best in the competition, and Fremantle sits one place behind it on the same number.
Finishing is the quiet separator. Fremantle converts at 65, fifth in the league. Brisbane sits at 48, eleventh, and has been between 41 and 49 for the entire season. The model does not weight finishing in the base rating, but it does project scores from it, and 16.5 goals to 15.1 is where the 10 points ends up on the scoreboard. Neither side pressures the ball well by the model's measure, 26 and 25, fifteenth and sixteenth, so this is not a game either team wins by tackling.
The trajectories matter more than the snapshot
Fremantle's contest rating was 39 at the end of the home and away season. Four rounds later it is 77. Beating Geelong by 14 and Sydney by 12 in finals, against two of the three best contest sides in the league, has rebuilt that number faster than anything else on the ladder this year. Its territory rating has climbed from 75 to 87 over the same stretch. The one number moving the other way is retention, which peaked at 81 in Round 18 and has slid to 51.
Brisbane's season runs in the opposite shape. It began the year as the best contest side in the competition at 92 after seven rounds and has come down to 65. Its retention collapsed to 22 by Round 18 and has recovered to 60 through the finals. Its finishing has not moved. In the model's reading Brisbane is a side that has fixed its ball use and lost its edge at the source, and Fremantle is a side that has found its edge at the source and got looser with the ball. The Grand Final is those two arcs crossing.
How the game itself is projected to split
When the model simulates the 46 players rather than the two teams, the game comes out close to even. Projected disposals are 369 to 371. Contested possessions favour Fremantle 134 to 131, clearances favour Brisbane 41 to 39, inside 50s favour Brisbane 60 to 57, and Fremantle wins the hitouts 40 to 34 through Luke Jackson. Brisbane takes far more marks, 109 to 89, which is its ball retention showing up in the numbers. Fremantle lays more tackles, 61 to 52.
So the six domain shares for this specific game read 49 to 51 Brisbane at the contest, 51 to 49 Fremantle on finishing, and 50 to 50 on the other four. The margin does not come from Fremantle dominating any single phase on Saturday. It comes from the model rating Fremantle the better side in the phases that decide games, with an 83 to 74 gap in overall power, and from Fremantle turning a share of the ball into more goals.
The players the model leans on
Player power is the model's rating of a player's impact on winning across the season, out of 100. Luke Jackson is third in the competition at 93, and his hitout edge is the one clear structural advantage Fremantle carries into the game. Caleb Serong sits sixteenth at 79 and has risen 13 points this week alone. Shai Bolton is twentieth at 77. Josh Treacy has fallen 12 points to 75 and Hayden Young has risen 26 to 56, which tells you how the preliminary final was won. The projections have Andrew Brayshaw at 25 disposals, Serong at 24 with a goal, Jordan Clark at 23, and Jye Amiss at three goals, with Treacy, Patrick Voss and Michael Frederick at two each.
Brisbane's list is deeper through the middle. Will Ashcroft is seventh in the competition at 88 and is projected for 27 disposals and a goal, the most of anyone on the ground. Lachie Neale is at 26 and Dayne Zorko at 25. Logan Morris rates fifteenth at 79, the highest rated forward in the game. Kai Lohmann has climbed 16 points this week and Eric Hipwood 36, both off the Hawthorn game. The goals are spread: Lohmann, Morris and Charlie Cameron at two each. Fremantle's scoring is concentrated in Amiss. Brisbane's is spread across five, which is a strength in a wet Grand Final and a weakness if none of the five gets going.
The head to head and the road here
These sides met once this year, in Round 12 at the Gabba. The model had Brisbane at 53 percent. Fremantle won by 25, 103 to 78, and it remains one of the more emphatic away wins of the season by a side the model was not tipping.
Both arrive the long way. Fremantle lost its qualifying final to Hawthorn by 32 at Optus Stadium, beat Geelong by 14 in the semi and beat Sydney by 12 at the SCG in the preliminary final. Brisbane lost its qualifying final to Sydney by 53, beat Adelaide by 53 and then beat Hawthorn by 9 at the MCG. The model tipped four of those six. It had Fremantle over Hawthorn and Brisbane over Sydney in the qualifying finals, and Hawthorn over Brisbane last week, and was wrong all three times. It has been wrong about Brisbane twice in three finals, in both directions, which is worth saying plainly before backing against them.
Before the finals began the model's 50,000 season simulations gave Fremantle the best chance of the premiership at 23 percent, with Hawthorn at 21, Brisbane at 21 and Sydney at 20. Two of those four are in the game.
Value on the day
The strongest leg on the board is Jaspa Fletcher for 15 or more disposals at 80 percent. He projects to 18 and averages 19. Jordan Clark for 20 or more disposals sits at 72 percent, projecting 23 against a 24 average. Josh Dunkley and Zac Bailey are each at 66 percent for 20 or more. The full board with live prices is on the match page. The Multi of the Week locks 25 minutes before the first bounce and will appear on the site then.
How the model has gone in 2026
159 correct from 217 tips, 73.3 percent, including finals. The finals themselves have been harder: 5 from 10, with both qualifying finals wrong and the Hawthorn tip wrong last week. That is the same record shown on the Footy Tip Jar home page and season tracker, and it is the context for a 62 percent tip. Six times in ten this call is right. The other four are why they play the game.
Power rankings (after R28)
| # | Team | Rating | Chg | Move |
|---|
| 1 | | | +7 | ▲2 |
| 4 | | | +6 | ▲2 |
Footy Tip Jar has Fremantle by 10 at the MCG.
FremantlevBrisbane Lions· MCG
Fremantle rates first for territory and third at the contest; Brisbane keeps the ball better and creates as many chances as anyone but converts at 48, eleventh in the league. The game by game split is close to level, and the 10 points comes from the model rating Fremantle the better side in the phases that decide games and projecting 16.5 goals to 15.1. Luke Jackson, third in the competition on player power, gives Fremantle its one clear structural edge in the ruck. Will Ashcroft is projected for 27 disposals, the most on the ground.
Tip: Fremantle by 10. 62% win probability
Value bets · as at team announcement
| Player | Market | Proj | Avg | Edge |
|---|
| Jaspa Fletcher BRL | Disposals 15+ | 18 | 19 | +6% |
| Jordan Clark FRE | Disposals 20+ | 23 | 24 | +2% |
| Josh Dunkley BRL | Disposals 20+ | 21 | 22 | +2% |
| Zac Bailey BRL | Disposals 20+ | 21 | 22 | +3% |
| Darcy Wilmot BRL | Disposals 15+ | 18 | 18 | +4% |
Frequently Asked Questions
Who does the model tip to win the 2026 AFL Grand Final?
Fremantle, by 10 points, at 62 percent. The projected score is Fremantle 111 to Brisbane 101, and the model's fair price is $1.53 for Fremantle and $2.53 for Brisbane.
Why Fremantle when the two sides look so even?
On a game by game split they are close to level. The margin comes from the model rating Fremantle the better side in the domains that carry the tip: first in the competition for territory at 87 against Brisbane's 74, and third at the contest at 77 against 65. Fremantle also converts better, 65 against 48, so an even share of the ball projects to more goals.
What is Brisbane's case?
Ball retention and chance creation. Brisbane keeps the ball better, 60 against 51, and rates the best side in the competition for creating chances at 78. It projects to take 109 marks to Fremantle's 89 and to win the inside 50 count. Its midfield is deeper, with Will Ashcroft, Lachie Neale and Dayne Zorko all projected for 25 or more disposals.
Does the MCG favour either side?
No. The model's home ground credit only applies at a team's true home venue, and the MCG is neutral for both clubs. Brisbane did win there last week against Hawthorn.
Have Fremantle and Brisbane played this year?
Once, in Round 12 at the Gabba. The model had Brisbane at 53 percent and Fremantle won by 25 points, 103 to 78.
How accurate has the model been in 2026?
159 correct from 217 tips, 73.3 percent, including finals. The finals have gone 5 from 10, with both qualifying finals wrong and last week's Hawthorn tip wrong. The figure is the one shown on the Footy Tip Jar home page and season tracker.
Where is the Brownlow results report?
At footytipjar.com/brownlowresults, free and ungated. It scores the guide's 3-2-1 for every game against the actual count and settles every market.
See the full season record on the season tracker and the 2025 backtest on the model proof page.
