Staking is the most boring and most important part of betting: the best selections in the world lose money with reckless stakes, while modest selections survive with disciplined ones. A staking plan answers the only question that matters after you pick a bet, which is how much of your bankroll to risk, and the right plan is the one you can follow through a losing run without abandoning it.

Flat staking means betting the same amount on every selection, typically 1 to 2 percent of your starting bankroll. Its virtue is simplicity: it survives losing runs without the stake spiralling, it makes record-keeping easy, and it prevents the emotional stake increases that destroy most bankrolls. Its drawback is that it does not distinguish between a strong fancy and a speculative one, treating both the same.

For most recreational bettors, flat staking at 1 percent is the right default: a run of ten consecutive losers, which happens to good bettors, costs only 10 percent of the bankroll, leaving plenty of room to recover.

Percentage staking adjusts the stake to your current bankroll, betting, say, 2 percent of whatever remains: stakes shrink during losing runs, protecting the bankroll, and grow during winning runs, compounding profits. It is slightly more complex than flat staking but more resilient, and it is the natural step up for bettors who track their balance carefully.

The Kelly criterion goes further, calculating the optimal stake from your estimated edge and the odds, and in theory it maximises long-term growth. In practice, full Kelly is far too aggressive for football, where edge estimates are uncertain, and most professionals use fractional Kelly, often a quarter or a half of the calculated stake, to smooth the wild swings.

Whichever plan you choose, write it down before you bet, including your starting bankroll, your stake size and the conditions for changing it: for example, review stakes only after 100 bets or when the bankroll moves by 25 percent. Never increase stakes to chase losses, never bet money you need for living costs, and treat the bankroll as a separate pot, not as part of your everyday spending.

Review your plan quarterly: if you are consistently profitable, a modest stake increase is reasonable; if you are losing, the answer is smaller stakes and better selection, not bigger stakes and hope.

Every bettor faces losing runs, and understanding their maths removes much of their fear: even a bettor who wins 55 percent of even-money bets will, over a long season, experience runs of eight or more consecutive losses, and a bettor winning only 50 percent will face longer ones. These runs are not evidence that your method has broken, but a normal feature of probabilistic betting, and the staking plan exists precisely to survive them. A flat stake of 1 percent means ten straight losses cost only a tenth of the bankroll, leaving you fully able to continue.

The psychological danger is greater than the mathematical one: after a losing run, most bettors either increase stakes to recover quickly, which risks ruin, or abandon a sound method for a new one, which resets the learning process. The disciplined response is to review your bets objectively, checking whether your selections were sound and the results merely unlucky, or whether your method genuinely needs adjustment. Keep a betting diary with reasons for each bet, because it turns a painful losing run into data, and data is far easier to act on than frustration.