Bankroll management for casual bettors

Picking winners gets all the attention. Staking decides whether you are still around to enjoy them. I have watched sharper judges than me go broke because they bet too much on the days they felt sure, and I have watched ordinary tipsters last for decades because they never did. This is the plain version of how to manage a betting bank if you bet for fun and want it to stay that way.

Start with a bank you can afford to lose

A betting bank is a separate pot of money set aside only for betting. It is not your current account, and it is not "whatever is left this month".

The right size is the amount you could lose in full without it affecting your bills, your plans or your mood. For some people that is £100. For others it is £1,000. The number matters far less than the rule: when the bank is gone, you stop, and you do not top it up from elsewhere on impulse.

Write the figure down. It sounds trivial. It is the single most effective thing a casual bettor can do.

Units: 1% of your bank

Every tip on GameTipster is staked in units, and 1 unit is 1% of your bank. On a £200 bank, 1 unit is £2. On a £1,000 bank, it is £10.

Units let people with very different banks follow the same advice in proportion. They also stop you thinking in pounds, which is where emotion creeps in. A 2-unit bet is a 2-unit bet whether you are up or down for the week.

Our stakes run from 0.25 units to 3 units on singles, set by the confidence grade. Accumulators are a flat 0.25 units. That means the most we ever recommend on a single bet is 3% of your bank.

Flat vs proportional staking

There are two sensible ways to turn units into pounds.

Flat staking

You fix the unit size from your starting bank and keep it. Start with £500, and 1 unit is £5 until you decide to review it, whatever happens in between.

It is simple, and it makes your results easy to read: profit in units is profit in units. The weakness is that a long losing run takes the same amount out of a shrinking bank, so each loss is a slightly bigger share of what is left.

Proportional staking

You recalculate the unit as 1% of your current bank before every bet. Win and the stakes grow. Lose and they shrink.

The advantage is protection. Here is what a run of straight 1-unit losers does under each method:

The proportional figures come from 0.99 multiplied by itself 8 times (0.923) and 15 times (0.860). The difference is small over short runs, but proportional staking can never quite reach zero, while flat staking can.

The cost is that proportional stakes are smaller on the way back up, so recoveries are slower.

What a 55% strike rate really feels like

Here is where casual bettors get caught out. They assume that a good strike rate means short losing runs. It does not.

Take a bettor who wins 55% of their bets at average odds of 1.91. That is a genuinely good record. The break-even strike rate at 1.91 is 1 ÷ 1.91 = 52.4%, and at 55% the expected profit is 0.55 × 1.91 − 1 = +5.05% per unit staked.

That bettor loses 45% of the time. The chance of any particular eight bets all losing is 0.45 to the power of 8, about 1 in 595. That sounds remote. But over hundreds of bets, there are hundreds of places for a run to start. These are the probabilities of seeing at least one losing run of a given length:

Read the 1,000-bet column again. A bettor with a real, long-term edge has better than a 60% chance of losing at least eight in a row at some point, and roughly a one-in-six chance of losing ten in a row. On average, the longest losing run is a little over five bets in a 100-bet sample, about seven in 500 bets, and about eight in 1,000.

Losing runs are only part of it. Wins and losses interleave, so the drop from a high point to a low point is usually bigger than the longest streak. We simulated 20,000 sequences of 500 bets at 55% and odds of 1.91, staking 1 unit flat. The median worst drawdown was about 16 units. One sequence in ten had a drawdown of about 27 units or more. All of those bettors were profitable in expectation.

A 27-unit drawdown on a 1-unit stake is survivable. The same run at 5 units a bet is the end of the bank.

That is the whole case for small stakes. Not because the tips are bad, but because even good tips come in lumps.

Five rules worth keeping

1. Keep the bank separate. Decide the amount in advance and do not add to it after a bad week.
2. Never stake more than 3% on one bet. Most bets should be 1% or less.
3. Do not raise stakes to win back losses. Chasing turns an ordinary losing run into a serious one. We wrote more about this in why we never chase losses.
4. Record every bet. Date, selection, odds, stake, result. Memory keeps the winners and loses the rest.
5. Judge yourself over hundreds of bets, not dozens. The table above shows why a bad fortnight tells you almost nothing.

The same standards apply to us. Our record begins with the relaunch on 15 September 2026, and every tip goes on the Results page as it settles, at the odds and stakes we published. There is no track record yet to quote. Judge us the same way you would judge yourself: over a large sample, and with the losing runs included.

If betting stops being fun, or the bank starts to feel like money you need, step away. Our responsible gambling page lists free, confidential support.