How to stake several bets on one day
Six picks on one board: why full Kelly on each would stake 167% of your bankroll, and how to size bets so the whole day's exposure stays small.
On this page
When you have several bets on the same day, size each one as a small, fixed share of your bankroll and then check the total you have at risk. Do not work out each stake as if it were your only bet. Six good bets staked as six separate "best" bets can add up to more than you own.
The problem with sizing each bet alone
Here are the six picks on the BAI8 Desk this morning, with our chance, the best price we saw, and the full Kelly stake for each as if it were the only bet you had. Kelly is explained in what Kelly actually tells you to do.
| Pick (Desk, 23 Oct) | Our chance | Best price | Full Kelly, alone |
|---|---|---|---|
| Arizona | 70.0% | 1.95 | 38.4% |
| Tennessee | 58.0% | 2.16 | 21.8% |
| Ole Miss | 54.5% | 2.46 | 23.3% |
| Northern Illinois | 55.5% | 2.68 | 29.0% |
| Virtus Bologna | 58.2% | 2.60 | 32.1% |
| Amed SK | 58.1% | 2.17 | 22.3% |
| Total | 166.9% |
Those per-bet stakes are already far too big, because they trust our numbers completely. Added together they ask for about 167% of the bankroll. That is the clearest sign that bets on one day must be sized together, not one at a time.
A simple way to do it
- Pick a flat stake per bet, such as 1% of your bankroll. The case for flat stakes is in flat stakes versus percentage stakes.
- Add up the day. Six bets at 1% is 6% at risk.
- Set a ceiling for one day before you look at the board, and keep to it. If the board has more bets than the ceiling allows, cut every stake equally rather than dropping the last bet you found.
- Count linked bets as one. Two bets that depend on the same game, the same weather or the same team news can both lose for one reason.
What a day of six bets looks like
Using our own chances for the six picks, one unit on each, and assuming the games do not affect one another:
| Outcome for the day | Chance on our numbers |
|---|---|
| Finish the day ahead | about 81% |
| Finish the day behind | about 19% |
| Lose all six | about 0.4% |
So even if our numbers are right, about one day in five like this ends in a loss. If our numbers are too generous, and edges this wide are where we check most carefully, as explained in why a very wide edge makes us look twice, losing days come more often. Small, fixed stakes are what let you keep going through them, the idea behind staking without going broke.
What this does not tell you
It does not tell you what your daily ceiling should be; that depends on your bankroll and how much a bad day costs you. The table assumes the six games are independent and that our chances are right; neither is certain. The six games are not all played today: we use them because they are on the board on the same day, which is when you would place the bets.
18+. Bet only what you can afford to lose.
Common questions
How much should I stake when I have several bets at once?
Size each bet as a small share of your bankroll, then check the total. Six bets at 1% each puts 6% at risk; if that total feels too big for one day, cut every stake, not just the last one.
Can I use the Kelly formula on several bets at once?
Not by adding up single-bet Kelly stakes. On our six Desk picks of 23 October, full Kelly per bet adds up to about 167% of the bankroll, which is impossible and far too risky.
What if my bets on one day are linked?
Treat linked bets, such as two legs in the same game or two teams hit by the same weather, as one larger bet and stake them less in total.
How likely is a losing day with six value bets?
On our own chances for today's six picks, and assuming the games are independent, about 19% of days with one unit on each would still lose money.