Scalping
Scalping aims to take one or two ticks of profit at a time, over and over, in a market deep enough to get both sides matched quickly. It is the highest-frequency and lowest-margin approach in this section — the edge on any single trade is tiny, so it all comes down to execution, costs and repetition rather than to being right about the outcome.
What a tick is, and why its size changes
A tick is the smallest step a price can move on an exchange's odds ladder — the rungs between which nothing exists. Crucially, a tick is not a fixed proportion of the price. The steps get bigger as the odds get bigger, so one tick near even money is a far smaller proportional move than one tick out at long odds. Our terminology page gives the idea: a price near 2.0 may step 2.02 → 2.04, while a price near 20 steps in much larger increments.
The exact ladder differs between exchanges, and it is theirs, not ours — check the ladder on the exchange you are trading rather than assuming. What matters for scalping is the consequence: the same "one tick" is a different amount of money depending on where on the ladder you are working.
Where scalping is viable
Scalping needs a market that can absorb both halves of the trade almost immediately. In practice that means deep, well-matched, reasonably stable books — often the minutes shortly before an event starts, when money has arrived but nothing has happened yet to move the price sharply.
Thin markets defeat the approach outright. If the gap between the best back price and the best lay price is wider than the profit you are trying to take, there is nothing to scalp: you would have to cross that spread twice, paying more than the move is worth. A wide or flickering book is a signal to leave the market alone, not to trade it smaller.
A worked example
The arithmetic is the ordinary equalising calculation: the opposing stake is your original stake multiplied by your original odds and divided by the new odds, and the locked result is the difference between the two stakes.
| Back stake | Backed at | Lay at | Lay stake | Locked result |
|---|---|---|---|---|
| £20 | 3.00 | 2.98 | £20.13 | +£0.13 |
| £100 | 3.00 | 2.98 | £100.67 | +£0.67 |
| £100 | 20.0 | 19.5 | £102.56 | +£2.56 |
Take the middle row. £100 × 3.00 ÷ 2.98 = £100.67, so laying £100.67 leaves £0.67 either way. Check it: if the selection wins, the back bet returns £200 profit and the lay bet loses 1.98 × £100.67 = £199.33 — a net £0.67. If it loses, the back stake of £100 goes and the lay stake of £100.67 is won — again £0.67. Rounding the lay stake to the penny can shift the two outcomes by a small amount relative to each other — more at long odds, where a penny of stake carries more liability. That is normal and worth expecting rather than chasing.
The third row makes the tick point concrete. A move of 20.0 → 19.5 is a much larger proportional step than 3.00 → 2.98, which is why the same £100 locks in almost four times as much. All three figures are before commission.
Why costs and non-matching dominate
Exchanges charge commission on net winnings in a market, so a gross edge of 67p is not what reaches your balance. Rates differ materially — the exchange guides set out what each one charges. Both exchanges you can connect today charge 2%; the guides also cover the two that are coming, where the base rate runs as high as 5%. At the scale a scalper works, a difference of that size is a large fraction of the whole trade — which is the arithmetic reason scalpers care about which book they are on, and why price routing (Pro) exists.
The bigger risk is not cost, though. It is the trade that only half-fills. A scalp assumes both legs match; if the first matches and the second does not, you no longer hold a closed trade — you hold an ordinary open position on a selection you never intended to be exposed to, and its value moves with the market. One such position, left alone, can be worth many completed scalps.
Scalping is not a low-risk activity, and nothing here is a prediction of what you would make. Prices move against you as readily as with you, an unmatched second leg leaves real exposure, and commission applies to your winnings while your losses cost full price. Never stake more than you can afford to lose.
Execution is the whole discipline
Two mechanical facts govern whether a scalp fills. First, queue position: an order placed at a price joins the back of everything already resting there, so the money ahead of you must be matched before yours is. Second, the price can simply move through your order — the market trades past your rung, your bet never fills, and the opportunity you were pricing has gone.
Both are reasons to work prices deliberately rather than react late, and both are why scalpers care about the number of seconds a stake change costs them.
Work the trade on the Ladder view (Pro): every price level a market knows about sits as a fixed rung, with the money waiting on each side shown beside it, and a single click places a bet at exactly that price. That fixed layout is what lets you see where money is stacking up and place a resting order a rung or two ahead of the current price. Pair it with hotkeys and quick-stake presets — preset buttons change your stake in one tap, and Pro's keyboard hotkeys fire the same actions as the ladder's clicks, so no part of the trade waits on retyping a number. Hotkeys place real bets, so switch them off when you have finished trading.
Close a losing scalp as fast as a winning one. The approach only holds together if the size of a bad trade resembles the size of a good one, and a position held "until it comes back" is the single thing most likely to undo a long run of small gains. Decide the exit before the entry, keep an eye on Positions & P&L so you always know what is actually open, and treat any unintended position as something to close rather than something to trade.