We settled 1.2 million each-way place halves. Here's when the terms work for you — and when they're a tax

Every each-way bet is really two bets: a win bet, and a place bet priced by a formula — a fixed fraction (1/4 or 1/5) of the win odds, paying a fixed number of places set only by field size and race type. The formula never asks whether those terms are fair for the race in front of you. So we asked instead.

We joined ten years of Betfair win markets to their place markets — 1.23 million UK and Irish runners, 2013–2024 — and settled the place half of a standard each-way bet at the industry-standard terms on every one of them, using actual place-market results. No model, no assumptions about finishing-order maths: just what the place part would have returned, race by race.

First, the trap we nearly fell into (and what it did to the numbers)

Our first run priced everything at Betfair's starting price — the closing number. It showed the place half returning +13.2% overall. Lovely. Also fiction: nobody can bet at a price that doesn't exist until the race starts. Re-run at morning prices — numbers you could actually act on — the overall return collapsed to −1.8%.

We publish both because that gap is the entire story of most betting-system marketing: measure at prices you couldn't get, and everything looks like an edge. It's also why every number on this site is logged before the off.

The finding that survives: big-field handicaps at double-digit prices

One segment held up in discovery (2013–2022) and in a held-out validation era (2023–2024), at real morning prices:

Handicaps with 16+ runners — where standard terms pay 4 places at 1/4 odds — with the horse priced between 11.0 and 21.0: the place half returned +17.0% across 26,000 runners in discovery, and +17.3% across 4,900 in validation.

Why would bookmakers give this away? Terms rigidity. "Four places, quarter odds" in a 20-runner handicap is genuinely generous maths, and it persists because it's an acquisition tool — and because firms manage the cost by restricting accounts that exploit it too well (industry surveys put restricted punters at 44% and rising). The overlay is structural, not a secret.

Two honest caveats before anyone gets excited: our morning prices are exchange prices, and on extreme outsiders bookmaker boards sit well below the exchange — so we've deliberately ignored the even bigger numbers our study shows above 21.0. And a structural overlay on the place half is not a promise your each-way bet profits: the win half still has to carry its own weight.

The finding you should tattoo somewhere: small fields are an each-way tax

The mirror result is stronger than the overlay, and it's the one that will save most punters real money:

In 5–7 runner races — 2 places at 1/4 odds — the place half lost between 14% and 22% in every single price band. Roughly 100,000 runners, a decade of data, confidence intervals nowhere near zero. And the everyday habit of backing a 6/1–10/1 shot each-way in an 8–11 runner handicap? −9.2% on the place half, across 88,000 runners.

Two places at a quarter of the odds in a six-runner race is arithmetic working for the bookmaker, dressed as insurance. If you like the horse in a small field, the data's advice is blunt: back it to win, or don't back it.

The rules of thumb, in one table

From today, every racecard on the site shows a small each-way terms context note computed from the day's actual field — red when the terms are against you, and a structural note when you're in the one shape where they lean your way.

Method notes: standard terms table applied by field size and handicap status; settled on actual Betfair place-market results; discovery/validation eras kept separate and 2025+ data untouched as a sealed holdout. This is analysis of the terms' structure, not betting advice, and past patterns can decay. 18+ · Bet responsibly · BeGambleAware.org.