The phrases "sharp money" and "public money" are the most-overused terms in betting media. Both are real concepts. Both are also used as marketing fodder by tout services who'd love you to buy their picks. Here's how the actual mechanic works, what the public data does and doesn't tell you, and how to use the signal without buying the hype.
The core distinction
"Public money" refers to bets placed by recreational customers — typically smaller stakes, often parlayed, frequently on favorites and overs, and usually placed in the hours leading up to a game. "Sharp money" refers to bets placed by sophisticated, profitable bettors — typically larger stakes, more often on underdogs and unders, more often placed at line-opening or in response to specific information, and frequently spread across multiple books.
The two groups have measurably different long-term outcomes against the closing line. Public money tends to underperform CLV. Sharp money tends to beat CLV. Sportsbooks know this and use it for risk management.
How books separate the two
Every major sportsbook now classifies customer accounts on a spectrum from "recreational" to "sharp." Classification is automatic, based on a few signals: deposit and bet size patterns, time-of-bet patterns (sharps bet earlier), market diversity (sharps spread bets across more markets), CLV history, and account age. A customer who consistently beats the closing line is flagged sharp regardless of their raw W/L record.
The book's response to a sharp customer varies: lower max-bet limits on certain markets, slower bonus offers, exclusion from certain promotions, and in extreme cases, full account closure. We've documented this dynamic in our player-prop limits coverage.
What public data tells you
Public consensus services (Action Network, BettingPros, Vegas Insider) report two numbers per market: bet percentage (what share of bets are on each side) and handle percentage (what share of money). Bet % tells you what the public is doing. Handle % tells you what the bigger bets are doing. The diff between them is the most usable signal:
- If 78% of bets are on Team A but only 40% of handle is on Team A, that means the bigger bets are on Team B. That's a sharp/public divergence in favor of Team B.
- If 65% of bets and 70% of handle are both on Team A, money is uniformly on A — no sharp signal.
- If bet % and handle % are roughly equal across both sides, there's no strong public-money tilt at all.
Reverse line movement is a corollary
Reverse line movement (RLM) — when the line moves against the public bet majority — is the cleanest sharp-action signal because it's an action by the book, not a self-reported number. If 75% of bets are on Team A and the line moves toward Team B, the book has decided that the money on B is more credible than the count on A. That's sharp signal.
How to use the data
- Use sharp/public divergence as confirmation, not as a thesis. If you already like Team B based on your model, and the public is heavy on Team A, the public-fade adds confidence. Don't bet Team B because the public is on Team A — that's circular reasoning that doesn't compound.
- Be skeptical of "sharp money" claims that don't cite handle data. Tout services frequently call any line move "sharp action" to justify their picks. Real sharp action shows up as handle/bet divergence or RLM, not as a free Twitter graphic.
- Sample size matters. Sharp signals are weakest in obscure markets (small-conference college games, second-tier soccer leagues). They're strongest in NFL/NBA/MLB game lines.
- Don't fade the public on principle. The public is correct on majority of bets — that's why they're betting them. Fading the public works only when public sentiment is paired with line action that confirms a sharp position.
Things sharp/public data does NOT tell you
- Whether the line is +EV at the current price. Sharp/public data tells you which side has the money; it doesn't tell you whether the price is fair.
- How long the signal will hold. Sharp action that hits at line-open and sits for hours is more durable than late-public action that arrives in the final 30 minutes.
- What the sharps actually know. Sharp action could be informational (an injury they've identified earlier than the market) or model-based (they think the price is mispriced) — the data doesn't distinguish.
The bigger frame
Sharp/public divergence is one signal in a betting process — not the whole process. The bettors who use it best are the ones who already have a number, already have a process, and use sharp/public as a context check. The bettors who use it worst are the ones who outsource their thinking to a "sharp moves" Twitter account. As always: build your number first, check the market second.