The Odds Gap

Hey IH. I'm Dan, solo founder of The Odds Gap, a free sports betting odds comparison and tools site. I made the classic mistake: I built for months and shipped 30+ features before doing any real distribution. This week I am finally doing the scary part, and it starts here.

Here is the one thing I did not expect to learn, and it is the whole reason the product exists: the number on the screen is not the price.

When people compare sportsbooks, they read the odds and pick the biggest number. But that number quietly hides costs that make a fair comparison almost impossible by the eye test. A few of them:

1) The vig, the juice, the sportsbook's cut. A -110 line looks like a coin flip. It is not. It is a coin flip with the book skimming about 4.5% off the top, win or lose, whether you notice it or not. 4.5%! I know plenty of boomers that let their money sit all year in bonds for a 4.5% return. I know these aren't sportsbooks' clientele, but that's probably partially why most look past it. Those gambling on sports are looking for a quick buck, a lotto ticket, a way out of the rat race, or maybe just a quick fix of adrenaline. Whatever the reason may be, the majority of sports bettors start off behind in the race. Back to the analysis - two books can post the same -110 and still not be the same bet once you look at both sides of the market.

2) The exchange fine print. Prediction and betting exchanges like Kalshi, Polymarket, Novig and ProphetX often show a better raw number than any sportsbook, because there is no vig baked in. But each one takes its cut differently, and that decides who is actually cheapest:

• Kalshi charges a per-contract taker fee, roughly 7 x P x (1 - P) cents, about 1.75 cents on a coin flip. A price that looks great before the fee can finish behind a sportsbook after it. Let's be honest - the majority of sports bettors aren't placing limit orders and saving on taker fees. The odds Kalshi shows are not THE odds you actually get.

• Polymarket also runs no vig, but it charges a per-trade taker fee just like Kalshi, currently about 2.5 cents on the dollar on a coin flip and steeper the longer the shot. Same problem: the raw price you see is not the price you actually get, so we bake the fee in before we show it. Even after that, Polymarket is the exchange that most often posts the best moneyline of the majors in our data, which is why it earns its spot. The one real tradeoff is breadth: it is mostly game lines with light prop coverage, so it wins you the best number on a marquee moneyline more than it covers your whole card.

• Novig charges no vig and no commission at all. In my matched-line data, where I only count spots where every book quoted the exact same line, Novig usually posts the best price of anyone. The catch is liquidity and market offer breadth. It is often the best number you cannot actually get filled at real size and they don't have nearly the number of betting options that traditional sportsbooks do.

• ProphetX is a newer CFTC-regulated exchange with better breadth than Novig, more markets actually quoted, but the same liquidity reality and a 2% commission on your net winnings for straight bets, zero on parlays. On a coin flip it sits around -102 where a sportsbook is -110.

So the honest ranking flips depending on what you care about. Best posted price is often Novig. Best price you can actually bet at size is frequently a sportsbook. That tension is invisible until a tool puts them side by side, post-fee, and tells you.

3) Profit boosts, the value people leave on the table. Sportsbooks hand out profit boost tokens, a 30% boost here, a 50% one there, and most people forget they have them or misjudge what they are worth. A boost can quietly make a worse-priced book the best book for one specific bet. So I built a Profit Boost Wallet: you save your boosts with the fine print, and the bet slip applies the eligible ones automatically and re-ranks which book pays the most for your exact bet.

The takeaway that shaped the whole build: you cannot line shop honestly by eye. The tool has to bake in the vig, the per-exchange fees, the liquidity reality, and your boosts, then flag the real best number. So it does. I do not sell picks and never will. It just tells you where the same bet pays more. Math, not picks.

Now the ask, since this is IH: I spent almost all my energy building and almost none getting it in front of people, and I know that is backwards. If you poked at a free odds tool, what would make you actually come back? And is there a betting-math thing you wish a tool just showed you? I will build it.