WeBall.
Open the app

Flat Betting

WeBall3 min readUpdated Aug 15, 2026Reviewed Aug 15, 2026

Flat betting means the same stake on every wager regardless of how good it looks. It gives up the theoretical growth of edge-proportional staking in exchange for two real advantages: it is impossible to execute badly, and it makes your results interpretable because every bet counted equally.

Flat betting is one unit on everything. The best team you have ever seen and the marginal lean you almost passed on get exactly the same stake.

It is the least sophisticated staking method available, and it is the correct one for most people betting today.

What it gives you

It cannot be executed badly. Every other staking method requires a probability estimate as an input, and amplifies whatever error is in that estimate. Flat betting has no input to get wrong.

It makes results interpretable. If every bet was the same size, your record means what it appears to mean. With variable staking, a 55–45 record can be a large loss and a 45–55 record can be a profit, which makes it almost impossible to learn anything from your own history.

It caps the damage from any single mistake. One badly misjudged bet costs one unit, not a quarter of your bankroll.

What it costs you

Edge-proportional staking genuinely does grow a bankroll faster when the probabilities are right. A bettor who can reliably distinguish a 2% edge from a 6% edge is leaving real money on the table by staking them identically.

The question is whether you are that bettor, and the honest answer for almost everyone starting out is no — not because of a lack of ability, but because you do not yet have enough history to know.

How to do it

Pick a unit — 1% of bankroll is standard — and do not move it.

The discipline is not in choosing the number, it is in holding it. Two failure modes destroy flat betting:

Stake creep after wins. A good month makes 1% feel timid. Recalculating your unit from a temporarily inflated bankroll is how a normal drawdown becomes a serious one.

"Best bet" exceptions. The moment you allow yourself to double up on the ones you really like, you are doing variable staking with your feelings as the model. If some bets deserve more, use a staking method that says so explicitly.

A flat-betting season

  1. 01Bankroll of $2,000, unit of 1% = $20 per bet.
  2. 02You place 400 bets over the season at −110.
  3. 03You go 214–186, a 53.5% win rate — a genuinely good result.
  4. 04Profit is roughly 28 units, or $560, on $8,000 wagered.
A 7% return on turnover from a 53.5% win rate. Note how modest the profit looks next to how good the win rate is — that gap is what the vig costs you.

When to move on

Not until you have evidence that your probability estimates are calibrated. That means a few hundred bets with consistently positive closing line value, which is a much stronger signal than a winning record.

At that point fractional Kelly at a quarter stake becomes reasonable — it will feel almost identical to flat betting most of the time, which is rather the point.

Common questions

Isn't flat betting leaving money on the table?

In theory, yes — edge-proportional staking grows a bankroll faster if your probability estimates are accurate. In practice most estimates are overconfident, and badly executed Kelly underperforms flat betting by a wide margin. The theoretical loss is real; the practical one usually isn't.

What size should a flat bet be?

One percent of bankroll is the standard. Between 0.5% and 2% is defensible depending on your confidence and tolerance for swings. What matters far more than the number is that it does not move.