Player props now generate the largest share of US sportsbook revenue, taking over the role parlays played a decade ago. They carry the same vig structure, worse liquidity, and an additional psychological hook — the illusion that watching the player gives the bettor an edge. A look at why the format has won, who is actually making money on it, and what the math says about its expected value for the recreational user.
A decade ago the marquee product at every US sportsbook was the parlay. Books loved them because parlays compound the vig on every leg into an aggregate expected return that is much worse for the bettor than any single leg would be. Customers loved them because the payouts looked dramatic and the cost of entry was small. The math was unforgiving but visible: a 4-leg parlay at -110 a side paid roughly 13:1 against true odds of about 16:1, and any reasonably attentive bettor could see it.
The player-props market is the spiritual successor to the parlay, and it is structurally worse. The vig is wider (6-10% per market versus 4-5% on spreads), the liquidity is thinner, the limits are lower, and the customer interpretation is more confused. This is not an accident — it is product design. The properly-skeptical reader of this article should not take away that player props are always wrong to bet, but that the format is built to favor the book more than any prior US sportsbook product.
How the format won
Three forces lined up to make player props the dominant US sportsbook revenue source between 2020 and 2024. The first is mobile-first product design: the prop menu, with its 30-80 individual stat lines per game, is built for scrolling on a phone in a way that the spread/total/moneyline interface never was. The second is the same-game-parlay (SGP) builder: a UI that lets the user combine 3-12 props from the same game into a single ticket, with implied correlation adjustments hidden inside the book's pricing engine. The third is fantasy-style player familiarity: the average US sports fan in 2026 knows player names and projected stat lines from fantasy and DFS in a way the average 2010 fan did not. The book is meeting the customer where their attention already is.
The product designers know what they built. The SGP builder is a parlay UI dressed up as a personalized prediction tool. The customer feels like an analyst constructing a thesis; the math underneath is the same multiplied-vig compounding that made parlays unattractive in their original form, with an additional correlation adjustment that the book sets in its own favor.
The vig structure: wider, less visible, equally extractive
A standard NFL spread carries about 4.5-5% vig (the overround above 100% on both sides at -110/-110 is 4.76%). A standard player prop — over/under on a single player's passing yards or points scored — carries 6-10% vig depending on the book and the stat. The reason is structural: the market for any single player prop is much thinner than the market for the side or total of the same game, so the book widens the price to compensate for the higher per-ticket variance it absorbs.
The wider vig is not advertised. Both sides of a typical prop show -115 or -120, which is visually similar to the -110 of a spread but mathematically meaningful: -120/-120 sums to 109.09% implied probability, an overround of 9.1% — almost exactly double the 4.76% of -110/-110. A bettor who treats the two as roughly equivalent in cost is making a systematic 4-5 percentage point error on every prop ticket relative to a side ticket on the same game.
| Spread (Chiefs -3.5 / Broncos +3.5) | -110 / -110 → 4.76% vig |
|---|---|
| Total (47.5 over / under) | -110 / -110 → 4.76% vig |
| Moneyline (Chiefs / Broncos) | -180 / +160 → ~4.5% vig |
| Mahomes pass yards over/under 270.5 | -118 / -114 → 7.5% vig |
| Kelce receiving yards over/under 70.5 | -120 / -120 → 9.1% vig |
| 4-leg SGP (spread + total + 2 props) | Combined implied vig ≈ 18-24% |
Liquidity, limits, and the account-restriction asymmetry
The thinness of the prop market is a customer-facing problem in two ways. First, the price the book is willing to honor scales down with market depth: a side bet on the Chiefs spread might accept $5,000 from a single account, while the same account might be capped at $500 on Patrick Mahomes' passing yards. The smaller limits make it impractical to bet props at sizes that materially affect a recreational bankroll.
Second and more important, books are quicker to limit accounts that win on props than accounts that win on sides. The reason is that the prop market is where the book's own pricing model is weakest, so any bettor consistently beating prop closing lines is a bigger signal-to-noise problem for the trading desk than a bettor beating spread closing lines (where the desk is more confident in its model). The empirical pattern is clear in published bettor-community data: account limits applied for prop CLV happen 2-3 times faster than limits for side CLV at the same magnitude.
The illusion of skill — the format's marketing edge
The deepest reason player props have won is psychological. A bettor placing a spread bet on the Chiefs is implicitly admitting they are taking a view on a complex system (offensive line, defensive scheme, weather, coaching, etc.) that they cannot fully evaluate. A bettor placing an over on Patrick Mahomes' passing yards is taking a view on a single player whose game they can watch, whose recent form they can recite, whose matchup-by-matchup history they feel they understand. The bet feels like research.
In practice, the book's prop pricing model already incorporates everything the recreational bettor knows and most of what they do not. The model uses the same recent-form data, the same matchup history, the same injury reports, plus proprietary inputs (opponent-adjusted snap-share projections, weather-adjusted pass-attempt models, in-week practice reports the public does not see). The marginal information the recreational bettor brings to the prop is small relative to what the model already prices in, but the experience of bringing that information feels like an edge.
The empirical test is straightforward: track 100 prop tickets across a season and compute closing-line value the same way you would on side bets. The vast majority of recreational prop bettors find their CLV is meaningfully negative — usually -2% to -4% on average, which is roughly the vig the format builds in. The 'edge' the bettor felt while researching the prop is not visible in the CLV data because it was not real.
Who is actually winning on player props
The bettor population making money on player props in 2026 is small and mostly invisible to the typical user. It consists of three rough segments: (1) bettors with proprietary projection models tuned to specific stat categories (typically operated by former DFS pros who built infrastructure during the 2014-2020 daily-fantasy era), (2) bettors with information advantages on specific players or coaching staffs (typically beat writers, former players, or coaches who have moved to media), and (3) match-bettors arbitraging cross-book prop pricing gaps. The first two segments are capped at maybe 200-500 accounts nationwide; the third is larger but constrained by book-shopping logistics and limit caps.
What is conspicuously absent from the winning population: the recreational bettor who watches a lot of games. The 'I watch every Chiefs game' bettor has no advantage over the book's model, because the book's model is built by people who also watch every Chiefs game, with more data and faster reaction times. Watching the games is necessary to enjoy props as entertainment; it is not sufficient to win at them.
What this means for the recreational reader
If you enjoy player props as entertainment, the cost is roughly 6-10 cents on the dollar per ticket in expectation, which compares with 4-5 cents on a spread. The premium is not exorbitant for an entertainment product, but it is real and it compounds across a season. If you bet $2,000 in prop volume per season, the format costs you an extra $40-80 versus the same volume in spread bets.
If you are betting props with the belief that you have an edge, the burden of proof is high. Track your CLV for 200+ tickets, calculate your hit rate against the implied probability of the prices you took, and compare to the implied vig. If your CLV is consistently above the vig (positive 1%+ after the book's margin), you may have a real edge worth scaling — though the book will probably limit you before the edge becomes meaningful to your bankroll. If your CLV is negative or zero, you are paying the vig the format was designed to extract.
What about SGPs — are they a worse version of the same problem?
Yes, by construction. A 4-leg SGP combines 4 markets with their individual vigs, then applies a correlation adjustment that the book chooses in its own favor. The combined implied vig on a typical 4-leg SGP runs 18-24%, depending on how aggressive the correlation discount is. The format's appeal is the dramatic potential payout; the underlying math is closer to a lottery ticket than a sports bet.
Are prop pricing models really better than what a careful bettor can build?
For mainstream stats on major-market players, almost always yes. The book has more data, better projection infrastructure, and more reaction speed. For thin-market stats on lower-profile players, sometimes the gap narrows enough that an attentive bettor with a basic model can find edge. The edge is real but small, the limits cap the position size, and the time-to-account-restriction is short. It is not a sustainable bankroll strategy for most recreational bettors.
Why are the books offering this format at all if a small population of bettors can beat it?
Because the small winning population is more than offset by the much larger losing population that the format attracts. The book's net economics on props are strongly positive in aggregate, even after limiting the few accounts that find edge. The product designer's job is to maximize the gap between the losing and winning populations' size, which is exactly what the SGP builder and the mobile-first prop UI accomplish.
The point of this article is not to discourage anyone from betting props — they are a legitimate entertainment product and the cost is reasonable on entertainment-budget grounds. The point is that the format's marketing presents it as an analytical opportunity (you watch the game, you bring information, you have an edge), and the math does not support that framing. Bet props because you enjoy props. Do not bet props because you think you are getting paid for your fantasy-football-style intuition. The book's pricing model already has your intuition baked in.