Hold

Win over handle

Handle is all the money bet, stakes only. Win is what the book has left after paying the winners. Hold is win divided by handle, so a book that writes $500,000 in a week and keeps $25,000 held 5 percent.

On straight bets at -110 the theory says 4.55 percent. That is the vig, and the hold drifts toward it over a long enough stretch if the lines are good. In any one week it can be anywhere.

Paid in lumps

A book is never perfectly even on a game, and on most Sundays it needs certain results, usually the underdogs and the unders, because the public bets favorites and overs. When the favorites all cover, the book has a losing week with every price still in its favor. When the favorites fail, it holds far more than the vig says it should.

So I never judged a book by one week, and you shouldn't either. A season starts to tell you something.

This is the argument every bookmaker has with the people upstairs. The accountants see slot machines hold the same percentage month after month, and they want to know why the sportsbook lost money in a week when it took more bets than ever, and the only honest answer is that a sportsbook is paid in lumps. Nobody in accounts likes it. (Nobody in accounts has ever liked anything.)

What the players bet matters more than the price

Hold depends heavily on the mix. Say that $500,000 week breaks down like this, using the straight-bet figure of about 4.5 percent and a parlay figure of 20 percent (about what a five-teamer costs).

Bet type	Handle	Hold	Win
Straight bets	$400,000	4.5%	$18,000
Parlays	$100,000	20%	$20,000
All bets	$500,000	7.6%	$38,000

The parlays are a fifth of the handle and more than half the win. That table is the reason for the parlay boosts, the same-game parlay button and the prop menu that gets longer every season. A book raises its hold by changing what you bet, not by changing the price of a side, and if I were still running one I would be pushing parlays too. The parlays page has the cut leg by leg.

High hold or high handle

A book can chase one or the other. It is hard to have both.

Take two books dealing straight bets only, both perfectly balanced. One deals -110 and writes $500,000; at 4.55 percent it keeps about $22,700. The other deals -105 and writes $1,000,000; at 2.38 percent it keeps about $23,800. The reduced-juice book has to write twice the business to make the same money, and it has to take bets from sharper players, because they are the ones who come for the price.

That is a real business, but a harder one. I respect the books that run it. I also understand why most would sooner keep the price at -110 and hand out bonuses.

The bookie's sheet

A local bookie using a per-head office does the same sum with two more lines in it. His win is what his players lost for the week, and out of that comes the fee to the office for every player who bet, and whatever he could not collect. A bookie whose players lost $6,000 on paper, who paid $400 in fees and is still waiting on $1,500, has $4,100 so far.

Paper hold and money in hand are different things on credit.

Reading it from your side

You can't see a book's hold. You can work out what each market is built to hold, and that is the part that concerns you: add up the two prices on anything before you bet it, or let the no-vig calculator give you the cut. Notice what a book promotes, too, because it is promoting where it holds the most. And don't read anything into one week, yours or the book's.

The hold is also a fair guide to your own costs. If you bet $200 a weekend on straight bets at -110, expect it to cost about $9 a week over time, win or lose in any one week. Put the same $200 into five-team parlays and the expected cost is about $40. I know which of those I'd pick. Decide which you can afford before the season, and set your loss limit to match.