A sportsbook does not charge you a fee. It charges you in the price. Both sides of a coin flip pay −110 instead of +100, and the difference is the house's cut. You never see it as a line item, so most people never count it.
This page counts it. Then it shows what happens when you take the same bet at the best of several books instead of the first one you opened. That is the whole idea: paying less for the identical bet. It is not a prediction, it is not a system, and it is not an edge over the books.
The number below is the hold: the share of every dollar bet on a market that the book keeps once both sides are added up. On a market priced the way books price them, it is also, near enough, what a bet costs you on average — 4.29% hold means about $2.15 of a $50 bet, every time, win or lose.
Books do not post the same number. On any given game one is a point better on the favorite and another is a point better on the dog. Take the best available price on the side you already wanted, and the cut you pay drops — same game, same bet, same stake.
Read the last column as: on 200 bets of $50, that is what stays in your pocket instead of the book's. On the moneyline the cut falls by about a third (4.29% → 2.84%). On a no-run-first-inning bet it falls by half (6.00% → 3.03%) — the markets books price laziest are the ones where shopping is worth the most.
We built models for all of it — moneyline, totals, run line, first five innings, first-inning runs — and then measured them against the closing market over thousands of games. None of them beat the market. The market price was as good as or better than our number, which is the ordinary result and the one most sites do not publish.
So we are not going to tell you we found an edge. What survived measurement was this: the price you pay is not fixed, and the difference between the best price and a typical one is real, repeatable, and provable from the books' own posted numbers. It requires no forecast of any kind. That is why it is the thing we put in front of you.
The 15 games on 2026-09-13, priced across 5 books we can see. These are the games where the books disagree most right now — one market per game, the one where shopping is worth the most.
The median game on today's board, market by market. These are today's numbers — a single day is noisy, and they will not match the multi-season figures above exactly. Both are shown so you can see the difference.
Books priced on this board right now:
Measured on today's own moneyline prices, across 8 games quoted by all 4 books. Each row is what you would pay holding exactly that many accounts, averaged over every combination of books — because nobody knows in advance which book will be best, and it changes game to game.
The shape is always the same: the second account does most of the work, the third adds a useful slice, and after that each one buys less than the one before. Two is worth doing today, three or four is the practical ceiling for most people, and past that you are opening accounts for a shrinking return.
A price you saw is not a price you have. Lines move on lineup news, weather and money. Everything on this page is a snapshot of when we last saw each book, and we print that time next to it for exactly this reason. Confirm the number in the book's own app before you put money on it.
Cheaper is not the same as good. Getting a better price on a bet you should not be making is still a bet you should not be making. Shopping changes the cost, never the merit.
The same bet means the same number. A total of 8.5 and a total of 9 are different bets, and comparing them invents a saving that does not exist. Every comparison on this page is made at one number, and a book quoting a different number simply does not appear in it.
We cannot see every book. We show the ones we have prices for. There may be a better number somewhere we do not cover, and there are markets where too few books quote the same number for any comparison to be honest — on those we say nothing rather than guess.