Compare n' Bet™
Compare n' Bet™
Where opening lines come from, how the market consensus forms, why the closing line is the most accurate prediction available, and where pockets of real inefficiency still exist for retail bettors.
People imagine sportsbooks setting lines in a windowless room with chalkboards and a bunch of guys arguing about the Patriots. That's not what happens. The reality: most retail US books don't originate any lines at all. They subscribe to data feeds from a small number of market makers and price-making operators, post those lines with slight modifications, and adjust based on their own customer action.
The actual line origination market is small. Pinnacle and a handful of offshore operators (Bookmaker, BetCris, Heritage, 5Dimes historically) plus syndicate-driven action at sharp Vegas books (Circa, the Westgate) are where most original prices come from. Recreational US books like DraftKings and FanDuel reference those lines, apply their own margin, and adjust based on how their book is positioned.
This matters because it tells you where information actually enters the market. If you want to know what the consensus probability of an event is, you don't average ten retail books. You look at the sharp originator and devig their line. Everyone else is downstream.
Modern sportsbook pricing has two distinct phases:
Phase 1: Origination at the market maker. A market maker like Pinnacle posts an opening line based on their internal model and accepts essentially all sharp action up to a high limit. The opening number is a hypothesis. As sharp money comes in, the market maker adjusts the line until incoming action is roughly balanced. The market maker's job is to find the price that produces equilibrium between sharp action on each side. They don't care which side wins. They earn their margin on the volume.
Phase 2: Distribution to retail. Retail books take the market maker's line, apply their own price filter (often shading toward the public's preferred side to extract more margin), and post it to their customers. The retail book's job is different. They don't want balanced action. They want to maximize hold from a recreational customer base while managing tail risk on big underdog hits.
The difference shows up in pricing. A line at Pinnacle might be -2.5 / +2.5 with -110 vig on both sides. The same game at DraftKings might be -2.5 (-115) / +2.5 (-105), shaded slightly toward the favorite because that's where the public is. Both books "have" the same line in the spread sense, but the implied probabilities and EV are different.
A typical line for a major NFL game opens at the market maker about a week before kickoff. It moves a couple of times in the early days as sharp action probes for mispricing. By the day of the game, the line has incorporated:
The closing line is the result of this process. It's the price at which the market is in (approximate) equilibrium between informed and uninformed action. Empirically, the closing line is the single most accurate prediction available for the event in question. Better than any individual handicapper, better than most public models, better than the opening line by a meaningful margin.
The opening-to-closing improvement in Brier score, while small in absolute terms, represents real information being incorporated by the market over the days leading up to the event. Bettors who can systematically beat the closing line are by definition pricing more accurately than the consensus, which is the entire premise of long-term profitability.
If you've spent any time in serious betting circles, you've heard the phrase "use Pinnacle as your fair line." Here's why that works.
Pinnacle operates on a low-margin, high-volume model. Their typical vig on major spreads and moneylines is 2% to 3%, compared to 4% to 5% at most retail books. They accept essentially all action up to high limits without limiting winners. Their pricing must therefore be accurate, because they don't have margin to absorb sharp customers' edge if they're consistently mispriced.
Empirically, the Pinnacle closing line on major US sports outperforms almost every other published source as a probability estimate. Multiple academic studies (Forrest, Goddard, Levitt) find closing lines at sharp books beat models, expert tipsters, and consensus aggregations across multiple sports. The "use Pinnacle as fair" rule isn't superstition. It's a useful heuristic backed by performance data.
The implication for finding value at retail books: you don't need to build your own model. You can use the sharp closing line (devigged) as your true probability estimate, then check whether retail books are offering prices that beat it. When they are, that's the +EV. When they aren't, no bet.
If a retail book is offering the away team at +135 or better, that's +EV against the Pinnacle-derived fair line. The exact percentage of edge depends on the price, but anything notably better than +127 is the kind of value that makes line shopping worthwhile.
The closing-line argument applies most strongly to liquid, high-volume markets at major books. The further you get from those, the more inefficiency creeps in.
Player props. The handle on a typical NBA player rebounds prop is a fraction of the handle on the moneyline. The market doesn't have time to converge to accurate pricing in the same way. Sharp models and tools that price props directly often find consistent value here, especially on second-tier players where the book's algorithmic pricing has less data to work with.
Niche sports. WNBA, college basketball outside the top 25, second-tier soccer leagues, lower-tier MMA cards, minor league baseball. The handle is small enough that sharp action moves prices meaningfully but the action isn't deep enough to fully correct mispricings.
Live betting. Live markets move on every play, every score, every momentum shift. The book's algorithm has to price in real time, and small mispricings happen continuously. Sharp live bettors can capture meaningful edge here, though the technical infrastructure required (fast data feeds, automated bet placement) is non-trivial.
Period and quarter markets. First-half NFL, first-quarter NBA, period-by-period NHL. These have lower handle than full-game markets and less attention from sharp players, so mispricings persist longer.
Promotional markets. Boosts, parlay insurance, profit boosts. The book is offering reduced or negative margin to acquire engagement, which sometimes produces +EV opportunities for customers who do the math. See the bonus EV guide for details.
Futures. Championship futures and award futures (MVP, ROY, etc.) move slowly and often have meaningful edge available, especially mid-season as new information becomes available faster than the book updates the line.
Two distinct ways to find +EV:
Information edge. You know something the market doesn't yet, or you process publicly available information better than the consensus. This is what professional sports models do. It's the hardest path. It requires real domain expertise and quantitative skill, and the edge erodes as more bettors converge on the same insights.
Structural edge. The market has predictable mispricings due to structural features rather than information failures. Examples: favorite-longshot bias, public team overpricing on Sunday Night Football, bias against road teams in college basketball, draws being slightly underbet in soccer markets. These edges persist because they're driven by recreational bettor preferences, not by ignorance of facts.
Most retail bettors who succeed do so on structural edge plus line shopping. The combination is enough to produce a small but real ROI without requiring a model that beats Pinnacle. The math: if the consensus is slightly biased in a known direction, betting against the bias at the best available line captures the structural edge plus the line-shopping edge.
The further you move from major-market, full-game, pre-game pricing, the less reliable the closing line is as a fair-odds proxy:
The practical implications stack up cleanly:
Line shopping is the highest-EV habit a retail bettor can develop. The market is efficient enough at the consensus level that beating the consensus by even 1% or 2% per bet over thousands of bets produces real ROI. Line shopping converts that consensus efficiency into your edge by getting you the best available price each time.
Building your own model is mostly unnecessary unless you're targeting markets where the closing line itself is inaccurate. For NFL spreads, your model needs to outperform Pinnacle's closing line consistently to be useful, which is a high bar. For props and niche markets, the bar is much lower and the upside is real.
The "use sharp book as fair" heuristic is good enough for most retail strategies. Compute fair odds from a sharp book's closing line, devig, and use that as your benchmark. Bet at retail books only when their line beats the benchmark. This is the structural backbone of any consistent retail betting strategy.
Where the market is efficient, edge comes from line shopping. Where the market is inefficient, edge comes from pricing accurately. Almost everything profitable in retail betting falls into one of those two categories.
The content on Compare n' Bet is published for educational and informational purposes. By reading this guide you acknowledge:
If you or someone you know has a problem with gambling, the National Council on Problem Gambling helpline is 1-800-522-4700 (US) or visit ncpgambling.org. International readers can find local resources at gamblingtherapy.org.