House Edge vs Hold: Why the Casino's Number and Yours Never Match

Colorado's Division of Gaming specifies an explicit accounting standard in Section 11 of its regulatory manual: casino accounting departments must calculate the hold percentage for each…

Colorado's Division of Gaming specifies an explicit accounting standard in Section 11 of its regulatory manual: casino accounting departments must calculate the hold percentage for each table by dividing net win by drop. That mandatory calculation explains why state revenue reports routinely post table retentions of 20 percent on games that carry theoretical house advantages of less than two percent.

The two metrics answer entirely different mathematical questions. House edge measures the mathematical rate of loss on an individual wager, calculated against the total volume of money placed on the felt or spun through the reels. Hold percentage measures the proportion of cash dropped into a game's locked box or bill validator that the casino keeps when the counting room opens. Confusing the two leads players to assume games are rigged or operating far outside their mathematical design.

Net Win, Drop, and Total Handle

The Nevada statewide gaming data collection, compiled by the UNLV Center for Gaming Research, separates financial results into three fundamental components: total handle (money wagered), drop (cash or credit exchanged for chips or vouchers), and win (retained revenue).

Drop represents the physical cash, credit slips, and converted vouchers dropped into the table drop box or slot bill acceptor. Win is what remains in the casino's possession after all payouts have cleared. Handle reflects the gross sum of every single bet made during a session, including initial bets, raises, splits, and re-wagered payouts.

Colorado gaming regulations require casino accounting staff to audit hold percentages continuously, comparing net win against drop to verify the reasonableness of table performance. A table game's drop box might collect $10,000 in paper currency over an eight-hour shift. If the casino finishes that shift with $2,000 more than its starting chip inventory, the table hold is 20 percent.

That 20 percent figure says nothing about the house edge on any individual hand. A player who puts $100 into the drop box does not make one single $100 bet. They buy $100 in chips and proceed to wager, win, lose, and re-bet their winnings across dozens of consecutive rounds.

How Slot and Table Holds Diverge in State Records

State regulatory agencies do not treat gaming machines and table pits as a uniform category. Nevada gaming statistics reports publish slot hold percentage and table hold percentage as separate line items, tracking each alongside its corresponding win and drop numbers.

The commercial gaming sector follows this regulatory division. The American Gaming Association's Commercial Gaming Revenue Tracker reported that national slot hold averaged roughly nine percent between 2022 and 2024. In the same reporting series, the AGA tracked national table hold at 20.0 percent in the third quarter of 2024, noting that this figure represented the fourth consecutive quarter of declining table hold across commercial gaming properties.

Nevada monthly and six-month statistics reflect this division across regional sub-markets. A Nevada six-month report for April 2024 documented slot hold values across reporting districts ranging between 6.34 percent, 7.17 percent, 7.41 percent, and 8.10 percent. The Nevada six-month report for October 2024 showed a similar distribution, listing slot hold values of 6.87 percent, 7.19 percent, and 7.57 percent.

Looking back further into historical filings, a Nevada Gaming Statistics report for 2010 listed slot hold and table hold as distinct statewide entries. It recorded that the Boulder Strip generated a 4.49 percent drop-in slot win over that period.

Because slot machines track every unit of currency entered as coin-in alongside physical cash inserted into bill validators, slot hold figures often run closer to the underlying statistical programming of the machine. Table games, by contrast, rarely track individual wager handle outside of dedicated electronic tracking systems, leaving drop as the only reliable denominator for regulatory compliance.

Bankroll Recycling and the Conversion Math

The mathematical bridge connecting house edge to hold percentage is bankroll recycling. When a visitor brings $100 to a table and receives $100 in chips, that $100 constitutes the table drop. If that visitor wagers $10 per hand on a game with a two percent house edge, the theoretical expected loss on the first wager is 20 cents.

If the visitor wins the first hand, their chip stack increases to $110. They wager $10 again. Over the course of two hours, the player might play 80 hands at $10 each.

The total volume of action generated—the handle—equals $800:

$$80\text{ hands} \times \$10 = \$800\text{ total handle}$$

The expected casino win against that $800 handle at a two percent house edge is calculated as:

$$\$800 \times 0.02 = \$16$$

When the player colours up and leaves the table with the remaining $84, the casino's accounting department logs the interaction according to Colorado and Nevada regulatory definitions:

• Drop: $100
• Casino Win: $16
• Table Hold Percentage: $\frac{\$16}{\$100} = 16\%$

The game's mathematical edge was two percent per wager throughout the session. The casino's accounting hold was 16 percent of the buy-in.

To convert between reported hold and underlying player volume, divide the hold percentage by the game's theoretical house edge:

$$\frac{\text{Hold Percentage}}{\text{House Edge}} = \text{Recycling Multiplier}$$

$$\frac{16\%}{2\%} = 8$$

A multiplier of 8 means the initial buy-in was wagered eight times over through the natural churn of wins and losses before the session concluded.

If a casino table operates with an average hold of 20.0 percent—matching the AGA national third-quarter benchmark—and the underlying rules give the house a theoretical edge of 1.25 percent, the implied wagering volume is sixteen times the original cash drop.

Regulatory Measures Are Revenue Totals, Not Single-Bet Odds

UNLV gaming industry statistics show that published win percentages reflect aggregate performance across entire reporting districts rather than static mathematical probabilities. When state filings show table hold fluctuating between quarters, those shifts reflect changing table velocity, average session lengths, and player behaviour rather than sudden alterations to game rules.

A game with a high hold percentage is not necessarily worse for a player than a game with a low hold percentage. A player sitting at a single-deck table with a 0.5 percent house edge who wagers for six hours straight can surrender 100 percent of their $50 buy-in. The table logs a 100 percent hold on that drop box transaction, yet the underlying game offered some of the most favourable mathematical rules on the gaming floor.

Conversely, a player who drops $1,000 into a machine with an eight percent hold, wagers $10 once, and immediately cashes out $9.20 leaves behind an eight-dollar win on a $1,000 drop. The machine logs a hold of 0.8 percent for that specific transaction.

State gaming regulators do not publish hold percentages to advise players on game selection. Regulators compile net win and drop figures to audit gross receipts, ensure accounting controls meet statutory standards, and verify that reported revenue matches physical cash flows across audited properties.