Why Odds Drift: Analyzing Market Movements and Bookmaker Liability

You place a bet at 5.00. Ten minutes later, the horse is 8.00.
This is called a drift.

Conversely, you bet at 5.00, and it drops to 3.50.
This is a steam (or shortening).

Why do prices move? Is a drifter always a "bad" horse? And what should you actually do when it happens to your bet?

TL;DR: Prices move for three reasons: informed money arriving (steam), bookmakers balancing their liabilities (mechanical drift on everything else), and public sentiment (noise). Steamers as a group outperform their morning prices; big drifters as a group underperform. But an individual drift often means nothing more than "a rival was backed" — and if nothing real has changed, a drifting price on a horse you rated is better value, not worse. The skill is telling the two apart.

The three drivers of price movement

1. Information (smart money)

Syndicates and pro bettors move the market. If they bet heavily on a horse, the bookmaker cuts the odds to protect themselves.

Steam usually indicates the horse is expected to run well.

The information arriving isn't always form-book stuff. It's gallop reports, stable confidence, a jockey booking that means more than it looks, or simply a sharper model than the compiler's. You usually can't know what the market learned — you can only see that it learned something, in the price.

2. Liability management (balancing the book)

Bookmakers want a balanced book where they make profit regardless of who wins. If everyone is betting on the favourite, the bookie has too much liability on it.

They cut the odds on the favourite to discourage more bets.
They drift the odds on the outsiders to attract bets and balance the risk.

In this case, the outsider drifting isn't "bad" — it's just a mathematical necessity for the bookie. Every market's probabilities have to sum to (a bit over) 100%; when one price shortens, others must lengthen, whether or not anything changed about those horses.

3. Public sentiment (dumb money)

On big race days (like the Grand National), the public bets on names, colours, and jockeys. This irrational money can distort the market, creating amazing value on the ignored horses.

There's a fourth mover worth naming, though it's really information in disguise: conditions. A going change redraws half the prices on a card in an hour — mudlarks steam, fast-ground horses drift, and some of those drifters become non-runners by lunchtime. A drift on a morning of changing ground is a different, more literal signal than a drift on a settled day.

The "drifter" myth

Myth: "The horse is drifting, so it must be lame/injured/not trying."

Reality: often, a horse drifts simply because another horse is being backed. If Horse A is backed heavily, Horse B must drift mathematically.

Drifters are often the best value bets. If you thought a horse was value at 5.00, and it drifts to 7.00 (and no news has changed), it is now even better value — same chance, bigger price.

But hold both truths at once, because the statistics are real: as a group, horses that drift markedly underperform their SPs, and steamers outperform theirs. Market moves carry information on average. The resolution of the paradox is the phrase as a group. The average drifter is drifting for a reason. Yours might be drifting because of pure book-balancing around a gamble on a rival — and if you can point to that cause, the statistics about average drifters don't apply to your case.

The practical test when your fancy drifts:

1. Is there identifiable cause elsewhere? A well-backed rival forcing everything else out is benign. Check whether the steam is concentrated on one horse.
2. Has anything real changed? Going update, non-runner, negative stable news, jockey switch. If yes, your original assessment is stale — redo it before deciding the bigger price is "value".
3. Is the drift extreme and friendless? A horse out from 5.00 to 9.00 with no rival gamble in sight is the market telling you something you don't know. Respect that more than a gentle 5.00 → 5.50.

Reading moves like a trader

A few patterns worth knowing beyond the basics:

Exchange first, books second. Informed money usually hits the exchange before high-street prices react. When the exchange price and a bookmaker's price disagree sharply, the exchange is usually ahead — that gap is precisely what our value analysis measures, and why a book price sitting above the exchange price is the classic value signature.
Late steam is the strongest steam. Money in the final minutes is placed by people confident enough not to need a better early price. Morning moves are opinions; ring moves are commitments.
The bounce-back. Book-balancing drifts often over-shoot and come back in the last minutes as the market corrects. Catching a horse at the top of a mechanical drift is one of the neater timing edges available — our early price vs SP guide covers the timing decision in full, and Best Odds Guaranteed means a drift after you bet costs you nothing anyway.

Each-way knock-on. A drifting win price drags the derived place price with it (the place part is a fixed fraction of win odds). A mechanical drift can therefore make the

each-way value hunting

How to react

Don't panic: if your pick drifts, don't cash out. Re-evaluate. Has the going changed? Is there a non-runner? If not, enjoy the bigger price.
Beat the closing line: as covered in our Bet Tracker Guide, your goal is to take a price that is higher than the starting price (SP). Our guide to early prices vs SP covers when to strike.

vibeodds tracks these movements in real-time. Use the Live Odds page and Value view to spot the steamers and catch the drifters before they bounce back, then check the Results view to see whether your read on the market was right.

Frequently asked questions

Are drifters worth opposing automatically?
No — the group-level underperformance is priced in by the time you see it. Blindly laying drifters or backing steamers at the moved price captures nothing; the edge belonged to whoever moved the price.

My horse drifted and won. Was my bet good or lucky?
Check what you beat: if you took 5.00 and it went off 8.00, you lost the price battle and got bailed out by the result. Winning bets can be bad bets — that's the whole CLV argument.

How big does a drift have to be to mean something?
Think in probability, not ticks. 2.00 → 2.20 is a bigger statement (4.5 percentage points of implied chance) than 15.0 → 19.0 (about 1.4). Small moves on short prices outrank big moves on long ones.

Do bookmakers drift a horse when they know something?
Books react to money and information like everyone else; a lone book drifting a horse ahead of the market is more often managing its own liabilities than signalling secret knowledge. When every book and the exchange drift together, the market as a whole has learned something.

Is a drift ever good news?
For a backer holding a BOG ticket, always: you'll be paid at the bigger SP if it wins. For value, a mechanically-caused drift on an unchanged horse is genuinely good news — it's the market handing you a better price for the same opinion.