Market Making explained for betting

In one sentence

Market making means placing both a back order and a lay order around the current price, aiming to get both matched and pocket the small gap between them, many times over.

How it works

On Betfair, someone who wants to bet right now takes whatever price is on offer. A market maker is the person whose unmatched orders sit there waiting. If you offer to back at 2.02 and to lay at 1.98, and both get matched, you have locked in a small profit whatever the result.

The profit per round trip is tiny, often a percent or two of stake before commission, so market makers need high volume and tight control. The real skill is not placing orders but managing inventory: what you do when only one side gets matched and the price moves away.

That is the core risk, called adverse selection. The people most keen to take your price are often the ones who know something, such as late team news or a goal about to be shown on a faster feed. You get matched exactly when you would rather not be.

The maths

L = \frac{B \, O_b}{O_l}, \qquad \text{profit} = \frac{B\,(O_b - O_l)}{O_l}

\text{Expected profit per trade} \approx \text{spread earned} - \text{adverse selection losses} - \text{commission}

B: the back stake that was matched.
O b: the back price you got; O l: the lay price you got.
L: the lay stake that makes the profit the same whichever way the event goes.
Commission is charged on net winnings in each market.

In plain English: if you back higher than you lay, a correctly sized lay locks in a profit, but losses on the trades that go wrong can easily outweigh it.

Worked betting example

A pre-match football match odds market. The home team trades at 2.00 with a tight spread. You place a back order at 2.02 and a lay order at 1.98.

Both sides match (the good case):

Backed £100 at 2.02.
Lay stake to lock in: 100 × 2.02 ÷ 1.98 ≈ £102.02 at 1.98.
If home wins: +£102.00 from the back, −£99.98 on the lay (102.02 × 0.98), net +£2.02.
If not: −£100 on the back, +£102.02 on the lay, net +£2.02.
After 2% commission on net winnings: about £1.98.

Only one side matches (the bad case): someone backs heavily, taking your lay of £100 at 1.98 (liability £98). News breaks and the price shortens to 1.80 before your back order at 2.02 matches. To close, you back £110 at 1.80:

If home wins: +£88 from the back, −£98 on the lay, net −£10.
If not: −£110 on the back, +£100 on the lay, net −£10.

One adverse trade wipes out about five good round trips. Profitable market making depends on keeping that ratio in your favour.

Where it's good

Busy pre-match football markets such as Premier League Match Odds and Over/Under 2.5 goals, where spreads are tight and flow is steady.
Traders with fast, reliable API access and automated order management.
Markets where much of the flow is recreational, reducing the share of informed takers.
Earning from short-term fluctuations without needing a view on the result.

Limitations and pitfalls

Adverse selection is the main cost, and it is usually underestimated. Track your results on trades where only one side matched.
Queue position matters: at a popular price there may be thousands of pounds ahead of you, so you get matched mostly when the price is about to move through you.
Commission comes off every net win: at 2%, the £2.02 round trip above keeps £1.98, so build it into every quote.
In-play, bet delays and sudden suspensions make quoting risky; many professionals avoid in-play market making in football, where a goal moves the price in an instant.
Competition from well-resourced bots has narrowed spreads in the busiest markets.
Backtesting is hard: historical data rarely shows where your orders would have sat in the queue.

How to build it

Python: betfairlightweight for the Betfair API and flumine for an event-driven trading framework.
Data: streamed order book data with timestamps, your own order and match logs, and the commission rate on your account.
Tip: set hard limits on open exposure per market and pull all quotes automatically when volume or price moves suddenly.

Order book imbalance - signals of which way the price may move next.
Queue position - how likely your order is to be matched, and when.
Informed trader models - the theory behind adverse selection.
Back-lay hedging - the arithmetic of locking in a result.
Tick size maths - why the spread is worth more at some prices than others.