How to approach betting: a marathon, not a sprint
Most people who follow a profitable betting model still lose money. Not because the model is wrong — because of how they bet it. They stake too much, chase losses, bail out during a bad month, or quietly turn a value strategy into an accumulator. This is what actually separates the people who end up ahead from the people who don't.
1. Decide your bankroll before you place a single bet
A bankroll is money you have set aside for betting and can afford to lose entirely. Not your rent. Not money you'll need next month. Decide the number first, write it down, and treat it as the whole world of your betting.
Everything after that is expressed in units. One unit = 1% of your bankroll. We stake between 1u and 3u depending on how much the model likes a bet.
| Your bankroll | 1 unit | Typical bet (1–3u) |
|---|---|---|
| £200 | £2 | £2 – £6 |
| £500 | £5 | £5 – £15 |
| £1,000 | £10 | £10 – £30 |
This is the single most important thing on this page. Percentage staking means a bad run shrinks your bets automatically, and a good run grows them. It is what makes surviving a losing month arithmetically possible.
2. Stake what the model says — not what your gut says
The stake is part of the bet. A 3u bet and a 1u bet are different instructions, and picking which ones to "back heavier" based on a feeling undoes the maths that produced the edge in the first place.
The temptation always arrives at the same moment: after a couple of losses, when you want to win it back in one go. That is precisely when raising your stake is most expensive.
3. You will lose more bets than you win. That's the design.
Our average odds are 2.83. At those odds you only need to win 35.3% of the time to break even. Everything above that is profit. So a losing night is not evidence of anything going wrong.
If you find yourself judging a strategy by how many bets won rather than what the bankroll did, that's the instinct to unlearn.
4. Drawdowns are the price of the edge
A drawdown is a fall from your highest point. They are not a malfunction — they are the cost of doing this at all, and everyone who lasts has sat through several.
Between 14 April and 19 May 2026 our bankroll fell 43 units and stayed down for over a month. Anyone who joined at the top and quit at the bottom locked in that loss. Anyone who kept staking the same way rode it out and finished far ahead. Same bets, opposite outcomes — decided entirely by behaviour.
5. Set expectations you can actually live with
In mid-July our monthly yield was running at +51.8%. We said publicly at the time that this was roughly four times our long-term average and should not be expected to hold. July finished at +26.5% — an excellent month, and half the mid-month figure.
Judge any strategy over hundreds of bets, not days. Weeks are noise. Months are barely signal. If a service is showing you a single spectacular week, you're being sold the exception.
The don'ts
Don't build accumulators out of value bets
Combining selections multiplies the bookmaker's margin against you at every leg. Bookmakers make more from accumulators than from any other product, which is exactly why they're promoted so heavily. Four "value" bets stacked into one slip is no longer a value bet.
Don't chase the price when the game turns
If the score goes against a bet and the live odds drift out, the temptation is to top up at the "better" price. Don't. The odds moved because the bet is now genuinely less likely to win — the price is not a bargain, it's an updated assessment. Stake once, at the price published before kick-off, and let it run.
Don't expect to profit every day, week or month
You will have losing days, losing weeks and losing months. Our own record has all three in it, publicly, because pretending otherwise would make the record useless.
The mindset in one line
Treat it like investing, not like gambling. You are looking for a small statistical edge applied consistently over a very large number of bets, and your job is to still be doing the same thing in six months' time. It is deliberately, necessarily boring — and boring is what compounds.