How Big Should a Betting Unit Be? A Bankroll and Kelly Staking Reference (2026)
Selection decides your edge; staking decides whether you last long enough to realize it. This reference sets out the arithmetic of bet sizing: what a unit is, why flat staking is the sane default, and exactly what the Kelly criterion prescribes when you genuinely have an edge — plus why full Kelly is so rarely used in practice. Every figure is re-derivable. Reference material only, not financial advice.
This is an educational reference document (PDF / Word), not an app or a betting product to download. WagerLex takes no wagers, publishes no picks, and runs no affiliate links.
What's inside
- The unit as a fixed fraction of bankroll, with sizing examples at 1% and 2%
- Flat staking and why it is the correct default for a bettor who cannot measure their edge
- The Kelly formula f* = (b·p − q) / b, with each term explained in plain language
- A fully worked Kelly example, and why it stakes nothing when you have no edge
- Why full Kelly is volatile and unforgiving, and what fractional Kelly buys you
- The overestimation trap: why an inflated edge estimate makes full Kelly overstake
Frequently asked
How big should one betting unit be?
A common figure is around 1% of your bankroll, so about $20 on a $2,000 balance. Smaller units survive variance longer; larger ones grow or shrink a bankroll faster. The right fraction depends on your measured edge and tolerance for swings, not on how confident a single bet feels. This is a reference note, not betting advice.
What does the Kelly criterion tell me to stake?
Kelly stakes a fraction f* = (b·p − q) / b, where b is the decimal odds minus one, p is your estimated win probability and q is 1 − p. At +120 (b = 1.20) with an estimated p of 0.50, f* = 0.10 / 1.20 = 8.33% of bankroll. It is positive only when your estimate beats the price's implied probability.
Why do people bet a fraction of Kelly instead of the full amount?
Full Kelly maximizes long-run growth on paper but is highly volatile and assumes your probability estimate is exactly right. Because edges are usually overrated, full Kelly tends to overstake. Betting a half or quarter of the figure sharply cuts swings and cushions an over-optimistic edge, at a small cost to theoretical growth.
Is bankroll management the same as financial advice?
No. This is a mathematical framework for managing risk on money already set aside for entertainment wagering. It is not investment guidance, not a recommendation to bet, and no staking method removes the house edge or turns a losing approach into a winning one.