An AFL betting strategy is a set of price checks built on how footy is scored and selected: set your own number first, and bet only where a bookmaker's price beats it. AFL gives you different inputs from other codes: margins have no key numbers, totals move with the ground and the weather, team news lands to a timetable, and player lines can trail a change of role.
Each piece is measured from the AFL's own results and player statistics: no winning margin from 1 to 30 points decided more than 2.6% of home-and-away games from 2012 to 2026, 2020 aside. None of it removes the bookmaker's margin; AFL betting covers the markets.
An AFL betting model: start with a fair price
Value betting means taking a price above the fair price your own estimate gives a result. Per $1 staked, expected value = chance x price - 1 (the expected value calculator works it out), so a bet is worth having only when your chance beats 1 / price.
The estimate is where an AFL betting model comes in, and the simplest is a margin rating. Each club carries a rating in points, the forecast margin is the gap between two ratings plus a home advantage, and each result shifts both ratings by a share of the forecast's error.
A version with its settings fitted to 2013 to 2019 missed the final margin by 27.4 points on average over the AFL's 2013 to 2026 home-and-away results, with a standard deviation of 34.7 points. The 2020 season is left out because its quarters were cut to 16 minutes.
Home advantage is the statistic to measure first. From 2012 to 2026 (2020 left out), home sides hosting an interstate visitor in their own state won 58.3% of 1,659 games, by 8.2 points on average. When both clubs came from the same state, home sides won 53.3% of 955 games, by 2.7 points.
Example: Illustrative numbers, not a real game. Your rating has the home side by 12 points. With results spread normally around the forecast, at the 34.7-point standard deviation, that is a 63.5% chance, a fair price of 1 / 0.635 = $1.57. At $1.68 the expected value is 0.635 x 1.68 - 1 = +6.7% per $1; at $1.48 it is 0.635 x 1.48 - 1 = -6.0%.
A rating knows nothing of team news or weather and the market's line does, so treat a gap between the two as a question, and let the closing price show whose number was better.
Risk: Betting involves risk. Your model's number is a judgement, not a known chance, and a bet that beats it still loses often. See responsible gambling for limits and support.
AFL margins have no key numbers
The AFL's results for the 2012 to 2026 home-and-away seasons cover 2,807 games, 2020 left out. Every exact winning margin from 1 to 30 points came up in between 35 games (1.2%, a margin of 21) and 73 games (2.6%, a margin of 2). A goal is worth 6 and a behind 1, so a game can finish on almost any number.
NFL scoring works in 3s and 7s: in 1,359 regular-season games from 2021 to 2025, exactly 3 points separated the teams 199 times (14.6%) and exactly 7 points 105 times (7.7%) (NFL key numbers).
Near the line, each AFL point is worth about 1.2 percentage points of chance, measured against a forecast in AFL line middles; the normal model in AFL line betting gives about 1.1. Three rules follow:
- A point of line at the same price is worth about 1.2% x 1.90 = 2.3c per $1 at $1.90, so compare lines on every game, not only near a favourite number.
- A half point onto a whole number turns about 1.2% of results from losses into pushes, worth about 1.2c per $1. The NFL key numbers guide's model puts 7.3% of results on exactly 3 for a -3 favourite, so the half point onto 3 is worth about 7.3c there, roughly six times as much. Never pay an NFL-style premium for an AFL half point.
- Alternate lines should step evenly. If the main line is $1.90 at a fair 50%, a line one point better covers about 51.2% of the time, worth 0.95 / 0.512 = $1.86 at the same margin, and two points better $1.81. A steeper drop means the extra points cost more than they are worth.
Risk: Betting involves risk. A better line raises what a bet is worth on average, not whether it wins. See responsible gambling for limits and support.
Totals: measure the roof and the rain before you use them
Docklands, the Melbourne ground the AFL plays as Marvel Stadium, sits under a retractable roof. From 2012 to 2026 (2020 left out), its 632 home-and-away games averaged 177.7 points against 169.4 for 639 at the MCG, about 8 points more, and it outscored the MCG in 12 of the 14 seasons.
The exceptions, 2022 and 2023, show how a short sample hides a real gap. At about 45 games a season, with totals spread about 28 points either side of a ground's season average, one season's average can wander about 4.2 points either way (28 / 6.7, the square root of 45).
Rain shows the same pattern. The AFL's match records give each game a weather type, more forecast than rain gauge. Open-air games recorded as rain or thunderstorms, 547 of them, averaged 163.2 points against 172.8 for the other 1,628.
After allowing for each season's scoring level the gap was about 7 points, and it pointed the same way in 13 of 14 seasons. AFL total points betting measures these effects for 2022 to 2026 and found the grounds' dimensions did not line up with their scoring.
Use them as a check, not a bet: start from the season's scoring level, adjust for the ground and the forecast, and expect the posted total to have moved already for anything public.
Team news: price after the final teams
Under the AFL Regulations (11 February 2026 version), teams for Friday and Saturday games go to the AFL by 5:45 pm on Thursday, Melbourne time (Regulation 11.2). Sunday teams are due then too, with the bench settled by 4:30 pm on the Friday.
The official team sheet must be in 60 minutes before the scheduled start, after which only one of its listed emergencies can come in (Regulations 12.1 and 12.3). There has been no substitute since 2026: the AFL announced on 1 October 2025 that a fifth interchange player would replace it, in a team of 23.
For player markets, the default is to bet after the teams. An all-in bet stands and loses if the player is left out, so an early edge has to beat the chance of that:
expected value of an early all-in bet = (1 - chance he is left out) x chance x price - 1
Example: Illustrative numbers. A player bet you rate a 55% chance at $1.95 is worth 0.55 x 1.95 - 1 = +7.25% if he plays. With a 5% chance he is left out, it is worth 0.95 x 0.55 x 1.95 - 1 = +1.9%, and at 8% it turns negative: 0.92 x 0.55 x 1.95 - 1 = -1.3%. The break-even risk is 1 - 1 / 1.0725 = 6.8%.
Team bets carry a different risk, a price that moves on the news; AFL team selections and betting sets out the weekly timetable and what each change does to a bet.
Risk: Betting involves risk. Waiting for team news avoids one risk but can leave you a worse price, and a late change can still land after you bet. See responsible gambling for limits and support.
Disposal lines trail a change of role
A disposals line built on a season average prices the role a player used to have. Centre bounce attendances, the AFL's count of a player's attendances at the centre restarts that start each quarter and follow each goal, show a midfield role changing. In the AFL's 2025 and 2026 home-and-away player statistics, taking every player game after at least six that season:
| Change over the player's last three games | Player games | Next game's disposals | Season average before it | Last-three average | A line at the season average went over |
|---|---|---|---|---|---|
| At least 10 fewer centre bounce attendances a game | 143 | 16.4 | 17.9 | 15.9 | 37.1% |
| At least 10 more centre bounce attendances a game | 161 | 19.6 | 17.7 | 20.1 | 64.6% |
| All player games | 11,770 | 16.4 | 16.5 | 16.5 | 48.6% |
The line is the half-point line nearest the season average, so 17.9 becomes 17.5. After a move out of the centre the next game landed about 1.5 disposals below the season average, and after a move into it about 1.9 above.
A bookmaker may have moved already, so compare its line with the player's current role, his recent attendances and time on ground. Tags, an opponent shadowing one ball-winner, are not counted publicly, so judge that match-up yourself. AFL disposals betting covers game time and milestones.
Risk: Betting involves risk. These are rates across hundreds of player games, not a forecast for one player, and a role can change back. Past results are no guarantee of future results. See responsible gambling for limits and support.
Quarter markets are not a quarter of the match
A favourite's edge does not split evenly into four quarters. Against the same margin forecast, over 2,609 home-and-away games from 2013 to 2026 (2020 left out), counting a level quarter as not won:
| Forecast margin for the favourite | Games | Favourite won the match | Favourite won the average quarter |
|---|---|---|---|
| Under 10 points | 889 | 53.1% | 51.2% |
| 10 to under 20 | 706 | 66.1% | 57.3% |
| 20 to under 30 | 480 | 74.8% | 62.8% |
| 30 or more | 534 | 89.0% | 74.0% |
A favourite that wins two matches in three wins only about four quarters in seven, so a quarter price scaled straight from the match price is too short. Wind and ends matter too: the teams change ends at every break (Law 12.1 of the 2026 Laws of Australian Football), so a steady breeze favours each team in alternate quarters. AFL quarter betting prices those markets from the match.
Keep score with the closing price
Profit takes hundreds of bets to say anything. Closing line value (CLV) answers sooner: your price / closing price - 1, so at illustrative prices, $2.20 taken on a team that closes at $2.02 beats the close by 2.20 / 2.02 - 1 = 8.9%.
Beat the close over a few hundred bets and that is early evidence your prices are sharper than the market's; fall short of it and the market's are sharper than yours. Closing line value covers taking the margin out of the close. Stake from a fixed bank you can afford to lose, as bankroll management sets out.
Mistakes that cost AFL punters money
| Mistake | What it costs |
|---|---|
| Paying extra for an AFL half point as if it were an NFL key number | A one-point-better line at $1.80 instead of $1.86 returns 0.512 x 1.80 = 92c per $1, not 95c |
| Backing a disposals over at the season average after a move out of the centre | At 37.1%, $1.90 returns 0.371 x 1.90 = 70c per $1 |
| Scaling a quarter price straight from the match | 66.1% for the match is $1.51; 57.3% for a quarter is $1.75, or about $1.71 under a dead heat rule |
Where the screeners fit in an AFL method
B337's +EV screener compares price against fair price across sports markets, but its fair price comes from the market, not a rating: prices across the market with the margin removed. That is an estimate, not a known chance, as positive EV betting explains. The screener lists bookmaker prices above it, the most out-of-line first, and every bet stays yours to place in your own bookmaker account.
Terminal Pro and Full Automation carry the screeners, as does the separate Screeners plan, and a free account shows the live counts while the rows stay locked. Bets placed through B337 use credits, a per-bet usage charge. For free and confidential support call 1800 858 858 or visit gamblinghelponline.org.au.
Risk: Betting involves risk. A +EV bet can still lose, and often does; the price on screen may be gone by the time you bet, and bookmakers can restrict or close accounts and void bets. See responsible gambling for limits and support.