Do Longshots Really Offer Worse Value? What the Data Shows

A well-documented pattern in betting markets, known as the favourite-longshot bias, describes odds on unlikely outcomes carrying a heavier built-in margin than odds on likely ones. This piece is educational analysis of market structure, not betting advice or a tip on any specific outcome.

The Myth: "Longshots Are Where the Real Value Hides"

A recurring belief among casual bettors is that favorites are overpriced because everyone backs them, and that the real value sits with longer-odds outcomes the crowd supposedly ignores. The reasoning sounds intuitive: if a market reflects public opinion, and the public gravitates toward the obvious pick, then an underdog should be the place where a sharper eye finds an edge the crowd missed. It is a tidy story, and it is largely backwards.

Why the Idea Persists

Part of the appeal is structural rather than analytical. A single large payout from a long-odds outcome is memorable in a way that a string of small, unremarkable favorite wins is not, so recent history feels weighted toward "the longshot that came in" even when it represents a small share of total outcomes. Recreational betting volume also skews toward longer odds for reasons that have nothing to do with pricing efficiency: a small stake on a big number promises a life-changing return, which is a more exciting story to tell than a modest profit from repeatedly backing a likely outcome. That demand pattern shapes how odds compilers set prices in the first place, which is the actual mechanism behind the bias rather than any special insight held by longshot bettors.

Where the Pattern Was First Documented

The favourite-longshot bias was not originally identified in football betting at all. Early research documenting it looked at horse racing, a market with decades of dense historical price and result data available for study, long before football odds markets carried comparable trading volume. Once the pattern was established there, researchers went looking for it elsewhere and found a similar shape in a range of other betting markets, including football, though the size of the effect differs from one sport and market type to another. That history matters for how much weight to put on the pattern: it is not a quirk of one dataset or one bookmaker's pricing model, but a tendency that shows up repeatedly across genuinely different markets, which is part of why it is treated as a structural feature of how betting markets absorb demand rather than a coincidence in any single dataset.

What Documented Market Data Actually Shows

The favourite-longshot bias is one of the more consistently replicated findings in betting-market research, observed across a range of markets over long periods. The core finding is that the implied probability embedded in longshot odds tends to overstate the outcome's true chance by a wider margin than the implied probability embedded in favorite odds understates it. In practical terms, a bookmaker's overall margin is not spread evenly across every outcome in a market — it is loaded more heavily onto the least likely outcomes, where demand from bettors chasing a large payout is highest and where the true probability is hardest to estimate precisely in the first place.

• Longshot outcomes systematically attract more betting volume relative to their true chance of happening than favorites do.
• Odds compilers respond to that demand pattern by pricing in a larger margin on long-odds outcomes, since the extra volume itself does not make the outcome more likely.
• The effect compounds because a longshot's true probability is statistically harder to pin down precisely, giving compilers more room to price defensively without an obvious reference point.
• The gap is not fixed — it varies by market type, competition, and how liquid the specific betting market is.

The Nuance: The Bias Is Real but Not Uniform

Treating "longshots are always bad value" as an absolute rule overstates a pattern that is real on average but variable in size. The bias tends to be more pronounced in markets with lower overall liquidity, where fewer participants and less trading volume give compilers more room to skew pricing without it being corrected quickly. It also tends to vary by market type: a simple match-result market in a heavily traded competition generally shows a smaller gap than a niche market, like a specific correct-score outcome, precisely because more informed money flows through the liquid market and narrows any mispricing faster. None of this means every underdog price is inflated by the same amount, or that the bias disappears entirely in efficient markets — only that its size is not constant, and treating it as a fixed rule of thumb across every market ignores exactly the variation that determines whether it applies strongly in a given case.

The Draw as a Special Case in Football Markets

Football's three-way match-result market complicates the simple favorite-versus-longshot picture, because the draw does not sit neatly at either end of the spectrum. A draw is typically a mid-probability outcome rather than a true longshot, yet research on football-specific markets has found the draw price can behave differently from what a straightforward two-outcome favourite-longshot pattern would predict, partly because bettors evaluate a draw differently from how they evaluate an outright win for either side. Some studies of football markets have found draw prices priced closer to fair value than the shortest-priced favorite or the longest-priced underdog in the same match, which is a reminder that the bias documented in simpler markets does not automatically transfer, outcome for outcome, into every structure a betting market can take. Treating a three-way football market as if it were just "a favorite and a longshot" skips over the draw's distinct behavior entirely.

Why "Favorites Are Better Value" Isn't the Right Takeaway Either

The natural next assumption — that backing favorites is therefore the smarter structural choice — replaces one oversimplification with another. A favorite's price still contains a margin, just typically a smaller one on average than a longshot's, which is a different claim from saying a favorite is fairly priced or represents value in any given match. Value is a comparison between a price and an independently estimated true probability for that specific outcome, not a property that belongs automatically to one end of the odds spectrum. A heavily backed favorite in a match with genuine uncertainty can be just as overpriced, relative to its real chance, as a poorly modelled longshot — the favourite-longshot bias describes an average tendency across large samples of markets, not a guarantee about any single price on any single day.

How Data-Driven Analysis Actually Approaches This

The more defensible use of this pattern is as a caution about where mispricing is statistically more likely to cluster, not as a shortcut for picking outcomes. Analysts who work from independently built probability models — often anchored in underlying performance data such as expected goals rather than reputation or recent results alone — compare a model's own probability estimate against the market price across the full range of outcomes, favorites included, rather than assuming value is concentrated at one end of the odds spectrum by default. That comparison is only as good as the model behind it, and a model built on thin or stale inputs can misprice an outcome just as easily as a public market can. The favourite-longshot bias is useful mainly as a reminder that market prices are not a neutral mirror of true probability at any point on the spectrum, and that the direction of the average distortion happens to run against longshots rather than in their favor.

This comparison exercise also has to account for its own blind spots. A model that leans heavily on recent-form inputs can be slow to react to a genuine change in circumstances, such as a run of fixtures against unusually strong or weak opposition, and will misjudge probability at both ends of the spectrum in ways that have nothing to do with the favourite-longshot pattern itself. Treating any single model's output as ground truth risks replacing one systematic bias — the market's demand-driven skew — with another, less visible one baked into the model's own assumptions. The more defensible approach compares several independently built probability estimates against the market price, and treats persistent disagreement between the model and the market as a prompt for closer scrutiny rather than an automatic signal in either direction.

Match-level data such as expected-goals models and historical scoring patterns, which underlying probability estimates are typically built from, is available on rubiscore.com for readers who want to compare a model-based view against posted market prices themselves.

This article discusses betting-market structure for informational purposes only. If betting stops being enjoyable or starts to feel out of control, support is available through resources such as GambleAware or a local responsible-gambling helpline. Betting on these markets is intended for those aged 18 and over.