Every AFL season has teams the betting market prices shorter than their results end up justifying. This isn't a knock on the bookmakers - their models are sophisticated. It's a predictable outcome of how bookmaker pricing interacts with public money flow. The teams with the largest, most active supporter bases get bet heavily at current prices, and the bookmakers have to adjust lines to balance their exposure. That adjustment is what produces the pricing bias.
This piece is an analytical observation about structural biases in AFL betting markets, not a prediction about any specific season or round. The pattern has held across multiple seasons of tracked data, which is the kind of pattern worth understanding if you're trying to find systematic AFL betting value.
The structural bias
AFL bookmakers run on thin margins in a highly competitive market. When a match has a popular team playing, public money flows heavily toward that team at whatever the current line is. To avoid accumulating one-sided exposure, the bookmaker has to shorten the popular team's price further than pure probability would suggest. The shortening continues until enough counter-action arrives on the unpopular side to balance the book.
The result is that popular teams consistently close at shorter prices than their win rates justify. Specifically, the teams with the biggest brand footprints - Collingwood, Richmond (historically), Carlton, Essendon - regularly end seasons with win percentages that are meaningfully worse than their average bookmaker implied probability.
A team that played at an average implied probability of 55% across the season should have ended with roughly a 55% win rate. When they end at 48% instead, you know the market consistently overpriced them. That gap is the pricing bias, and it's where the contrarian value lives.
Why it persists
This should close over time. Sharp bettors notice systematic biases and bet against them until the prices adjust. Why doesn't it close in AFL betting?
Three reasons.
First, the bias is small per-match but statistically robust over a season. An overpriced team might be off by 3-5% per match on average. That's inside the noise of any individual game, but across 23 rounds, it compounds. Sharp bettors who would otherwise arb or fade this might not notice the signal at the per-match level.
Second, betting against popular teams means consistently betting the less-popular side, which is uncomfortable. Most sharp AU punters still have emotional relationships with clubs. Systematically fading Collingwood every week for a season is hard to execute psychologically even when you know it's +EV.
Third, the bookmaker limiting system actively targets the punters who would close this edge. Someone who successfully bets against popular teams week after week produces the exact CLV pattern that triggers account restrictions fast. The edge persists partly because the bookmakers gub the punters who would otherwise remove it.
Which teams the bias applies to
The effect scales with supporter base size and media profile. In the AFL market, historically, the bias has been most observable on:
- Collingwood. The largest supporter base in the league and the most consistent overpriced team across multiple seasons of tracked data. The bias fluctuates with team quality - when Collingwood are genuinely good (2023 flag year), the bias narrows or inverts. When they're mid-table, the bias is largest.
- Richmond. Strong during the 2017-2020 flag era. More recently, as the team rebuilt, the market has been slower to adjust down than the results justified, creating persistent overpricing.
- Essendon. Classic overpriced-on-brand team. Wide supporter base, heavy media coverage, pricing that has historically flattered their actual performance.
- Carlton. Similar dynamic to Essendon. The Blues get heavy public action regardless of form, and the pricing has been reliably shorter than results over the last decade (excluding their strong 2023-2024 seasons when the pricing matched the results).
The opposite pattern - teams the market systematically under-rates - shows up on smaller-brand clubs. The Giants, Suns, Hawks in rebuild years, and Saints historically have all ended seasons with better win rates than their average implied probability suggested. This is the other side of the same coin.
Important caveats
Three things to be careful about before you go and bet against Collingwood 22 weeks in a row.
First, the bias is not stable year-to-year. In years when a “popular” team is genuinely one of the best teams in the league, the market can correctly price them or even underrate them. Betting against a team in their flag year is punishing. The bias shows up on average over seasons and over teams, not every week for every fixture.
Second, the bias is small. We're talking about 3-5% EV on the fade side of specific matches, not 20%. It's real but it's not going to change your life. Compounded across a full season with consistent sizing it adds up, but it requires patience and the discipline to keep going when you hit a streak of popular teams winning anyway (which happens regularly because this is still a small edge on probabilistic events).
Third, this is an observation about historical market behaviour, not a fixed law. AU bookmakers adjust. If enough sharp money starts fading Collingwood consistently, the market will re-price. The edge has been visible in tracked data across multiple seasons, but nothing guarantees it persists.
How to actually use this
Don't bet every single fade-the-popular-team opportunity. Instead, treat it as one input among several when evaluating AFL H2H markets.
A specific workflow that I've found useful: when a popular team is priced as a moderate favourite (between -20% implied probability edge and -5% implied probability edge, so roughly lines of -15 to -30), check whether the market consensus across AU bookmakers agrees on the line. If there's substantial dispersion, trust the longest price (bookmaker offering the least aggressive shortening) as being closer to fair value.
The Krok Odds +EV Finder does this comparison automatically across every AFL match and highlights the best-available prices where they diverge from market consensus. Not all those divergences will be on the fade-the-popular side, but in practice a disproportionate share are, for the reasons described above.
For market-consensus-based AFL tipping without the contrarian analysis layer, the AFL tips page publishes median-consensus picks weekly and tracks its running accuracy across the season. These picks will align with the popular side more often than not, because the market is usually roughly right. The value opportunities are in the specific matches where the market is wrong, which requires checking individual fixtures rather than blindly fading any team.
The principle to take away, beyond any specific team list: AU betting markets are not purely rational. They are shaped by the flow of public money, which is unevenly distributed across the league. Wherever the public money concentrates, the pricing gets stretched. Finding AFL betting value means looking at the matches where that stretching is largest, and staying disciplined enough to take the less-popular side when the numbers justify it.
Sarah covers the sport-by-sport pricing landscape and the wider betting culture. Reports on tipster schemes, social-media betting scams, and the specific market inefficiencies that show up in AFL, NRL, and NBL player props.