One of the most durable patterns in betting markets is the favourite-longshot bias: as a group, bettors overbet favourites and underbet longshots. The result is that short-priced favourites tend to be slightly *over*priced, and underdogs slightly *under*priced.
Why it happens
It's mostly psychology. Longshots offer a cheap ticket to a big payout, so they attract more money than their true odds justify — the same instinct that sells lottery tickets. Favourites feel 'safe', so bettors pile in even when the price has already moved past fair. Both pressures push the market away from the true probabilities in a predictable direction.
Where that leaves the value
If favourites are overbet and longshots underbet, the mispricing — and the value — sits disproportionately on the underdog side. That doesn't mean blindly backing underdogs; it means the market is more often wrong there, so a good model finds more edges there.
It shows up in the data
When we back-tested Momus's value bets, the returns on underdog value bets were roughly double those on favourite value bets — a direct fingerprint of this bias. It's why the discipline isn't 'back the favourite'; it's 'back wherever the market is most wrong'.
See how that fits the wider method in how Momus finds value on Polymarket, or the maths of a good bet in expected value, explained.