Public Betting Percentages and Betting Splits Explained

Betting splits show how the public is betting on a game. When a site says "78% of bets are on the Chiefs -3," that's a betting split. When they say "65% of money is on the Chiefs -3," that's a separate data point. The gap between those two numbers is where things get interesting.

What betting splits actually measure

There are two types of splits, and they tell you different things.

Bet percentage is the count of individual bets placed on each side. If 100 people bet on a game and 78 took the Chiefs, the bet percentage is 78% Chiefs / 22% opponent. This mostly reflects recreational bettor behavior, since most individual bets are small.

Money percentage (also called handle percentage) is the dollar volume on each side. If $1 million was wagered on the game and $550,000 is on the Chiefs, the money percentage is 55% Chiefs / 45% opponent.

When these two numbers diverge, it signals something. If 78% of bets are on the Chiefs but only 55% of money, it means the small bettors are heavily on the Chiefs while the bigger bettors are taking the other side. In many cases, bigger bettors are sharper bettors.

Where to find betting splits data

Several sites publish betting splits, though the data quality varies:

• Sportsbook-specific data: Some books (DraftKings, for example) show their own bet and handle percentages. This is real data from one book, but it only reflects that book's customer base.
• Aggregated data: Sites that compile data from multiple sportsbooks provide a broader picture, but the methodology isn't always transparent.
• Consensus data: Shows the average line across many books, which helps you identify where individual books deviate.

No single source has complete data. Every sportsbook has different customers with different tendencies. DraftKings betting splits skew more recreational than Pinnacle's handle data would (if Pinnacle published it). Use multiple sources and treat the numbers as directional, not precise.

How to interpret betting splits

One-sided public action

When 75%+ of bets and money are on one side, the public has a strong opinion. This happens regularly with:

• Popular teams (Cowboys, Lakers, Yankees) regardless of the matchup
• Teams on winning streaks
• Primetime/nationally televised games where casual bettors participate more heavily
• Favorites in general (the public gravitates toward favorites and overs)

One-sided public action doesn't automatically make the other side a good bet. The public can be right. But it does tell you where the market might be inflated by recreational demand.

Splits divergence (the key signal)

The most useful signal from betting splits is when bet percentage and money percentage point in opposite directions.

Example:

• 72% of bets on the Cowboys -6.5
• But only 40% of money on the Cowboys -6.5

This means the large majority of individual bettors took Dallas, but the money favors the other side. Since larger bets tend to come from sharper bettors, this divergence suggests that sharp money is on the underdog while the public piled on the favorite.

Reverse line movement

This is the strongest signal from betting split data. When the line moves against the public side, it usually means the sportsbook is adjusting to sharp action, not public action.

Example:

• 80% of bets on the Over 47.5
• The line moves from 47.5 to 48

If the public is hammering the over, you'd expect the line to move UP to balance the book's exposure. Instead it moved up, making the over harder to hit. Wait, that's the expected direction.

Now flip it:

• 80% of bets on the Over 47.5
• The line moves from 47.5 to 46.5

The book moved the total DOWN despite 80% of bets being on the over. This means sharp money hit the under hard enough to move the line against the public weight. That's reverse line movement, and it's one of the strongest indicators of smart money. It's also evidence against the efficient-market hypothesis applying perfectly to sports betting: markets don't instantly reflect all available information, which is what creates the opportunity.

For more on interpreting line movement and how it connects to closing line value, read our guide on identifying sharp action.

Fading the public: does it work?

"Fade the public" is one of the most popular betting strategies on the internet. The idea is simple: if the public is usually wrong, bet the opposite. Take the unpopular side.

The reality is more nuanced.

Where fading works

The public genuinely overvalues certain things:

• Big-name teams. The Cowboys, Lakers, and Yankees get disproportionate action regardless of the spread. This can inflate the line past fair value.
• Recent performance. A team on a 5-game winning streak gets more public money than their current form justifies. The line adjusts partially, but recreational sentiment can push it further.
• Overs. The public slightly prefers overs because scoring is more fun to watch. Historically, unders have had a very slight edge in certain markets, though this is well-known enough that the market has partially corrected. This is a form of the wisdom of crowds breaking down: when the crowd has a systematic bias rather than independent opinions, the aggregate is wrong.
• Favorites. The public leans toward favorites, especially large favorites. Underdogs have historically shown a small positive return in certain sports and line ranges.

Where fading doesn't work

Blindly fading every public side is not profitable. The public is right more often than contrarian bettors want to admit. Popular sides win at roughly the rate the line implies. The edge from fading is small, inconsistent, and depends heavily on game selection.

The profitable approach isn't to fade the public mechanically. It's to use public betting data as one input alongside sharper signals: closing line value, sharp book pricing, and expected value calculations.

Using betting splits with +EV betting

Betting splits alone don't tell you whether a bet has positive expected value. They tell you where the public is, which is useful context but not a complete strategy.

Here's how to combine splits with a data-driven approach:

1. Identify a +EV opportunity using sharp line comparison (e.g., a retail book offers better odds than Pinnacle's devigged line).
2. Check the betting splits. If the +EV side also happens to be the unpopular side, that's confirming evidence. The retail book may be slow to adjust because public demand is on the other side.
3. Check for reverse line movement. If the line is moving toward your side despite public money going the other way, sharp money agrees with your edge.

None of these signals alone is sufficient. Together, they increase your confidence that the edge is real.

Bet Hero does the sharp line comparison automatically. You can layer betting split data on top to confirm whether the public vs. sharp dynamic supports the opportunity.

Sport-specific notes

NFL betting splits

NFL has the most public action of any sport. Primetime games (Sunday Night, Monday Night, Thursday Night) draw massive casual volume. Splits divergence is most pronounced in these games. Regular-season Sunday 1pm games tend to have more balanced, less exploitable splits.

NBA betting splits

NBA splits are useful during the playoffs when casual bettors pile onto favorites and star-driven teams. Regular-season NBA gets less recreational attention, so splits are less extreme. The total (over/under) market tends to show more public bias than the spread.

MLB betting splits

Baseball splits can be misleading because the public heavily backs starting pitchers with name recognition. A game with a well-known ace attracts lopsided public money, but the market typically accounts for this. NRFI and first-five-innings markets often have less efficient splits than the full game.