A limit is a decision about your money
A limit looks like a setting on a player's account. It is really a statement about how much of your own money you are willing to put behind that player this week. Every limit you hand out is a small loan against your bankroll, and the limits added together are the most your week can cost you.
That is why the right starting point is your bankroll and not the player. Most agents do it the other way round: a player asks for a number, it sounds reasonable, and it goes in. Thirty reasonable numbers later the total is more than the book could pay.
Work back from what you can pay
The useful question is simple. If every player won the most they could this week, could you pay all of them in full and on time?
Take straight bets at the usual price, where a player risks 110 to win 100. A player allowed to risk 500 in a week can win at most about 455 from you, because 500 divided by 1.10 is 454.55. Do that for every player and add it up. That total is your worst week on straight bets, and it needs to sit below the money you have set aside for the book.
It will almost never happen that everyone wins everything. The point of the exercise is that you know the ceiling, and that the ceiling is one you chose.
There is real mathematics behind being this careful. In 1956 a Bell Labs researcher, John Kelly, described a formula for sizing bets against a bankroll, and the best known finding from that work is that risking more than the formula allows increases the risk of ruin, even when the odds favor you. A book is in the same position as any bettor with an edge: the margin is on your side, and you can still go broke by putting too much behind any one week. The related result, gambler's ruin, says a finite bankroll that keeps playing against a much deeper one eventually loses even in a fair game. Limits are how a small book avoids being the shallow pocket.
Two limits, not one
A single weekly number is not enough. You want at least two.
The first is the most a player may risk on any one bet. This stops a single game from deciding your week.
The second is the most he may risk across the week. This stops a player from chasing a bad Saturday with a worse Sunday, which hurts him and, when he cannot pay, hurts you.
Parlays need a third: a maximum payout per ticket. A parlay risks little and pays a lot, so a player well inside his weekly limit can still win a sum you did not plan for.
Start low and raise on evidence
A new player gets the lowest limit you use. Watch three or four weeks before you change anything. You are looking for two things: how he bets, and whether he pays on the day you close the week.
Raise a limit when the evidence supports it, in steps, and only after checking that the new total across all players still fits your bankroll. A winning streak is not evidence that a player deserves more credit. Paying on time for a month is.
Lowering a limit is harder, so do it early and plainly. Tell the player before his next bet. An awkward message on a Tuesday costs less than an unpaid balance on a Monday.
Split limits by sport and bet type
Players are not the same across the board. One is cautious on football and reckless on parlays. Another only bets a sport you follow less closely than he does.
On the platform you can set limits by player, by sport and by bet type. That lets you tighten the one place you are exposed without closing the account, which is usually what both of you want.
Signals worth checking every week
- A player who always takes the same side just before the line moves. He may simply know more than your number does.
- A player whose usual bet suddenly doubles.
- Several players loading the same side of one game, which concentrates your risk on a single result.
- A balance that grows for three weeks without a payment.
Read these from the weekly figures and not from memory. Memory keeps the dramatic weeks and forgets the pattern. If you are unsure why a balanced week still makes the book money while a lopsided one may not, where a book's profit comes from explains the margin.
A worked example
Take an agent, call her Marisol, running thirty players in Puerto Sereno with 15,000 dollars set aside for the book.
She uses three tiers. Eighteen newer players may risk 300 a week, ten established players 500, and two long-standing players 1,000. The total her players can risk in a week is 5,400 plus 5,000 plus 2,000, which is 12,400.
At the usual price of 110 to win 100, the most they could win on straight bets is 12,400 divided by 1.10, or about 11,273. That is under her 15,000, so even a week in which every player wins everything is one she can pay.
Then one of the established players asks to go from 500 to 2,000. On its own it sounds fine, since he always pays. Marisol runs the total again: 13,900 at risk, about 12,636 in possible winnings. Still under 15,000, but the cushion has dropped from roughly 3,700 to roughly 2,400, and parlay payouts are not in that figure at all. She offers 1,000, with a cap on parlay payouts, and says they will look again in a month.
Nothing in that conversation required software. It required her knowing the number before he asked.
What limits do not do
Limits control how much you can lose in a week. They do not make an individual bet profitable, they do not replace paying attention, and they do not answer whether taking bets is lawful where you operate. That last question depends on where you and your players are, and it belongs with a lawyer.
BookiePerHead is software and a support line. It takes no wagers and holds no player money, so the credit you extend is yours, and so is the decision about how much of it to give. For the wider set of controls a book has over its lines and maximums, see the limits and controls on the sportsbook.