BOH 825: Game Profitability Ranking

A casino cannot rank games intelligently by house edge or yesterday’s win. It has to ask a broader question:

What does this game contribute after demand, capacity, labor, space, volatility, promotions, controls, and customer value are considered?

That answer changes by property and by time period. Baccarat may lead in a destination casino with international premium play. Slots may dominate a locals property. Blackjack may justify more space on weekends than on weekday mornings. A carnival table with a high mathematical edge can still rank last if it attracts little play.

A useful profitability ranking is therefore a decision framework, not a universal list of “best casino games.”

Start by separating four different numbers

Managers often create bad rankings because they mix unlike measures.

Actual win answers, “What happened?” Theoretical win answers, “What should the action have been worth on average?” Contribution asks, “What remained after the costs required to produce it?” Strategic value asks whether the game affects the rest of the property.

None should be used as a substitute for the others.

For table accounting terms, read Table Win, Drop, and Hold Explained. A high hold month may reflect favorable player results for the casino rather than a durable improvement in the game.

The core ranking dimensions

A robust comparison normally includes at least these dimensions.

1. Normalized gaming value

For a table game, a simplified theoretical-win estimate is:

Theoretical win = average bet × decisions per hour × open hours × occupancy factor × house advantage

The formula must match the game. A baccarat table may track total action per coup. Blackjack may require hands played and number of occupied spots. A roulette spin can carry many simultaneous chips. Side bets may need separate treatment because their edge and frequency differ from the base wager.

For slots, normalized value is usually built from coin-in and the approved hold percentage:

Theoretical slot win = coin-in × theoretical hold

These figures are models. Bad ratings, unrecorded play, incorrect speed assumptions, and unusual rules can distort them.

2. Demand and occupancy

A game cannot monetize an attractive edge without players. Track:

• occupied positions by daypart;
• rejected demand when the game is full;
• average waiting time;
• open-table hours versus staffed hours;
• minimum-bet sensitivity;
• repeat and rated-player participation;
• migration when the game is closed or moved.

Occupancy needs context. A six-seat table at 100% occupancy may produce less than a half-full high-limit table. A game can also be “full” because the minimum is too low relative to demand, leaving profitable action unserved.

3. Direct operating cost

Live tables carry dealer, relief, supervision, cards or dice, fills, credit, surveillance, equipment, training, and error-correction costs. Some games require specialized dealers or slower settlement. Progressive and proprietary games may add licensing or jackpot-accounting costs.

Slots require less direct labor per position but add capital cost, maintenance, systems, participation fees in some arrangements, jackpot liability, and floor-network support.

A simplified direct contribution measure is:

Direct contribution = normalized gaming win
                    - direct labor
                    - game-specific fees
                    - direct promotions
                    - consumables and variable support cost

This is not full property profit. It is a comparison layer. Rent, utilities, security, marketing, finance, and executive overhead still exist, but allocating every shared cost too early can hide the difference between avoidable and unavoidable expense.

4. Capacity productivity

Prime floor space is scarce. A ranking should show what each constrained resource produces.

Useful ratios include:

Win per open hour = normalized win / open hours
Contribution per table = direct contribution / tables open
Contribution per position = direct contribution / playable positions
Floor yield = direct contribution / square metres or square feet used
Labor productivity = direct contribution / paid labor hour

No single denominator is correct for every decision. A staffing decision needs contribution per labor hour. A floor redesign needs yield per unit of space. A game-mix decision may need contribution per position and rejected-demand data together.

A worked comparison

Consider an illustrative one-month result. The numbers are not industry benchmarks; they show the method.

A ranking based on actual win would put baccarat last because a few players won heavily during the month. A normalized contribution ranking puts baccarat second overall and first per square foot.

That does not automatically mean the casino should expand baccarat. Management still needs to test whether demand exists, whether the loss was ordinary variance or a control issue, whether suitable dealers and credit capacity are available, and whether added baccarat would cannibalize the current table.

Volatility changes the confidence in a ranking

Games with concentrated high-limit action can swing far around expectation. A monthly ranking based on actual win may reverse simply because one player had a strong trip.

Use several views:

• actual results for cash accountability;
• theoretical results for normalization;
• rolling 12-month performance for stability;
• player-level review for concentrated exposure;
• confidence ranges or variance indicators where data supports them.

The goal is not to ignore actual losses. It is to avoid interpreting random fluctuation as a structural failure.

Nevada’s table-games minimum internal control standards require statistical reports showing drop, win, and win-to-drop hold by table and game type across multiple periods, followed by management review and investigation of material fluctuations. That control logic supports the same principle: compare results with a base level and investigate before acting.

Strategic value belongs in the model—but not as an excuse

A game may contribute beyond its own win:

• a premium baccarat room can support hotel, food, credit, and host relationships;
• poker can create long visits and cross-play;
• a low-limit blackjack pit can supply energy and accessible entry to the floor;
• a distinctive game can support marketing or brand positioning;
• a slot bank can draw traffic toward a quieter zone.

Indirect value should be measured where possible. Track cross-play, incremental visits, room nights, offer redemption, nearby machine lift, and customer migration. Avoid statements such as “the game creates atmosphere” without evidence.

A weak game can survive for strategic reasons, but the reason should be explicit, owned by management, and reviewed against an alternative use of the same resources.

Why house edge is a poor ranking shortcut

House edge measures price per unit wagered. Profitability also depends on how many units are wagered and what it costs to offer the game.

A 10% edge on $100,000 of annual action has $10,000 of theoretical win. A 1% edge on $20 million of action has $200,000. The lower-edge game is worth twenty times more before costs.

Speed also matters. A game with a modest edge and high decisions per hour can generate more theoretical win than a slow game with a larger edge. But increasing speed may raise errors, reduce service quality, or shorten desirable player sessions. “Faster” is not automatically “better” when customer value and control risk are included.

Rank by daypart, not only by annual average

One annual leaderboard can hide operational opportunities.

A useful ranking matrix separates:

• weekday morning;
• weekday evening;
• weekend peak;
• special events;
• high-limit and mass-market areas;
• seasonal periods;
• promotional versus non-promotional days.

A game may rank poorly overall but be the best use of one table on Saturday night. Another may justify continuous operation because reopening cost, staffing complexity, or guest expectations make intermittent closure inefficient.

This is why why some games disappear from the floor is a portfolio question rather than a simple response to one bad month.

An operating scorecard

A practical monthly scorecard can show:

Weights should reflect the decision. A closure model may emphasize avoidable contribution. A floor-allocation model may emphasize space yield and rejected demand. A staffing model may emphasize contribution per paid hour and service levels.

Do not hide judgment inside a single unexplained score. Show the component metrics, assumptions, and tradeoffs.

Common ranking errors

• Using one month of actual win as the main performance measure.
• Ranking by house edge without action volume.
• Comparing slot and table economics without labor and capital differences.
• Treating all overhead as avoidable when a game closes.
• Ignoring full-table turnaways and unmet demand.
• Giving strategic value credit without measuring it.
• Removing a game without tracking where its players migrate.
• Expanding a game that is full only because its minimum is underpriced.
• Comparing a premium relationship game with a grind game on one ratio.

The strongest ranking does not produce a permanent order. It tells management what to open, close, resize, reprice, move, or test under current demand and constraints. For the wider vertical comparison, continue to Table Games vs Slots Profit. For customer-level economics, low-roller economics shows why volume and service cost must be interpreted together.