Chips & Truths No spin. Just the math.
Home/Back of House/Table Games Operations/BOH 315: Table Game Performance Metrics: Results, Theo, Labor, and Control

BOH 315: Table Game Performance Metrics: Results, Theo, Labor, and Control

How casino managers read table games through drop, win, hold, decisions per hour, labor, volatility, and risk controls.

A table can finish the night with a large win and still have been operated poorly. Another can lose money while the dealers, ratings, limits, staffing, and procedures were exactly right. Table-game performance metrics must separate result, expected value, operating efficiency, and control quality.

That is the central discipline. Actual win contains short-term gambling variance. Management decisions should not reward or punish employees for luck.

Layer 1: What the accounting records say happened

Statistical drop

Statistical drop is the approved activity measure used in the table-games performance report. Its detailed treatment of cash, markers, pit credit issuances, and pit repayments follows local rules.

Drop is not total wagering handle. A player can buy in once and recycle the same chips through many decisions.

Statistical win

Statistical win is the approved table-games result for the period. It is built from controlled table inventories, drop, fills, credits, and applicable credit adjustments.

A simplified table equation is:

Table Win = Closing Inventory + Credits + Drop - Opening Inventory - Fills

If opening inventory is $25,000, closing inventory is $31,000, drop is $18,000, fills are $7,000, and credits are $2,000:

Table Win = $31,000 + $2,000 + $18,000 - $25,000 - $7,000
          = $19,000

The exact report may handle markers and specialty-game transactions separately. The point is that win comes from a reconciled accounting chain, not from estimating what players appeared to lose.

Hold percentage

Hold % = Statistical Win / Statistical Drop

With $19,000 win and $80,000 statistical drop:

Hold = $19,000 / $80,000 = 23.75%

Hold is not house edge. Hold uses drop as the denominator and is affected by chip recycling, session duration, player behavior, credit movement, and short-term results.

Layer 2: What the wagering activity was expected to produce

Theoretical win

For a player or table group, a simplified model is:

Theoretical Win = Aggregate Average Wager × Decisions per Hour × Hours × House Edge

Suppose a blackjack table averages $420 in total wagers per round, deals 70 rounds per hour, operates at that activity level for 8 hours, and uses a 1.2% expected edge for the measured mix:

Theoretical Win = $420 × 70 × 8 × 0.012
                = $2,822.40

Every input needs a definition. Aggregate wager is not the same as one player's average bet. The edge depends on rules, wager mix, and player decisions. The active period may be less than the table's total open hours.

Actual-to-theoretical difference

Result Variance = Actual Win - Theoretical Win

If the table wins $7,200 against $2,822.40 theo:

Result Variance = $7,200 - $2,822.40 = +$4,377.60

That is a result difference, not evidence that the dealer created extra advantage. Managers should review it across game type and rolling periods.

Nevada's current Table Games MICS require statistical reports by table and game type and management review of unusual fluctuations against a defined base level; see the Nevada Table Games Minimum Internal Control Standards, Version 9.

Layer 3: Was the capacity used well?

Table hours

Table hours count how long tables were open. They are a capacity measure, not proof of demand.

Average Open Tables = Total Table Hours / Operating Hours

If the pit records 96 table hours during a 12-hour shift:

Average Open Tables = 96 / 12 = 8 tables

Seat occupancy

Seat Occupancy = Occupied Seat-Hours / Available Seat-Hours

A six-seat table open for eight hours has 48 available seat-hours. If observed occupancy totals 31 seat-hours:

Seat Occupancy = 31 / 48 = 64.58%

Properties may estimate occupancy from ratings, observations, sensors, or sampled counts. The source affects accuracy, especially for uncarded play.

Decisions or rounds per hour

Pace can increase wagering volume, but only while game protection, customer service, and accuracy remain intact.

Rounds per Hour = Completed Rounds / Active Game Hours

Compare like games with similar player counts and procedures. A full baccarat table and heads-up blackjack cannot share one pace target.

Wager volume per table hour

Estimated Handle per Table Hour = Aggregate Average Wager × Decisions per Hour

This is an estimate unless every wager is captured. It helps connect occupancy, limits, and pace to theoretical value.

Layer 4: What did the operation cost?

Table games are labor-intensive. A useful view includes dealers, relief, supervisors, dual-rate coverage, and any directly assigned support.

Labor Cost per Table Hour = Allocated Labor Cost / Table Hours
Theoretical Win per Labor Dollar = Theoretical Win / Allocated Labor Cost

Suppose the example table carries $760 in allocated labor cost:

Theoretical Win per Labor Dollar = $2,822.40 / $760 = 3.71

This is not final profit. It excludes many shared costs and may not allocate supervisors perfectly. It is a comparable operating measure when the same method is used consistently.

Low labor cost is not automatically good. Understaffing can increase closed games, missed ratings, errors, fatigue, disputes, and weak supervision.

Layer 5: Did the controls survive the pace?

Revenue without procedural integrity is not healthy performance. Add control measures such as:

Dealer Error Rate = Verified Dealer Errors / Decisions Observed
Dispute Rate = Substantiated or Reviewed Disputes / Table Hours
Rating Capture Rate = Rated Eligible Play / Estimated Eligible Play
Fill Accuracy Rate = Correct Fill Transactions / Fill Transactions Reviewed
Exception Closure Time = Total Time to Resolve Exceptions / Exceptions Resolved

These denominators must be chosen carefully. A property that reports more incidents after improving detection may look worse even while control improves.

Useful qualitative indicators include:

  • repeated late or missing ratings;
  • frequent unnecessary fills and credits;
  • unusual voids or corrections;
  • weak handover documentation;
  • supervisor zones too large for meaningful observation;
  • speed increases accompanied by payout errors;
  • unresolved negative or extreme hold results;
  • recurring procedure deviations by game or shift.

A compact table review

Consider this eight-hour blackjack table:

Measure Result
Statistical drop $48,000
Actual win $7,200
Hold 15.00%
Aggregate average wager $420 per round
Rounds per hour 70
Theoretical win $2,822.40
Seat occupancy 64.58%
Allocated labor cost $760
Verified payout errors 1
Reviewed decisions 560

The table had strong actual win, but management should not stop there. Questions include:

  • Is the 15% hold normal for the game's rolling base level?
  • Was the aggregate wager captured correctly?
  • Did the table remain open after occupancy fell?
  • Did the payout error expose more value than the pace gained?
  • Were all premium players rated?
  • Is the labor allocation comparable with other shifts?

A good dashboard turns each answer into a decision, owner, and review date.

Game-type context

Metrics cannot be compared blindly across blackjack, baccarat, roulette, craps, and carnival games.

  • Blackjack: player count, rules, penetration, side bets, and decision quality affect pace and edge.
  • Baccarat: high-limit concentration can dominate monthly actual win and credit exposure.
  • Roulette: table layout, chip handling, call bets, and player density affect speed and dispute risk.
  • Craps: several wager types, staffing positions, and complex payouts make simple decisions-per-hour measures less useful.
  • Carnival games: higher edges can coexist with lower demand or slower adoption.

Use game-specific theoretical assumptions and control measures rather than forcing every table into one ranking.

Rolling periods beat isolated shifts

Daily and shift reports are necessary for accountability, but strategy should use month-to-date, year-to-date, and rolling comparisons. A single premium player can move baccarat hold dramatically. A large roulette result can make a weakly occupied table appear exceptional.

A practical fluctuation review compares:

Hold Deviation = Current Hold % - Rolling Base Hold %

The threshold for investigation should come from the property's approved controls and game characteristics, not an improvised percentage after the result is known.

Management uses, with limits

Decision Useful metrics Required caution
Open or close a table occupancy, wait time, theo per table hour, labor protect game availability and service
Change limits average wager, refused demand, volatility, credit risk high limits can concentrate result risk
Coach a dealer pace, verified error rate, procedure observations never judge by table win alone
Change game mix demand, theo, contribution, floor space allow for cannibalization and learning period
Review a promotion incremental drop/handle, theo, cost, player overlap separate promotional instruments in reporting
Investigate result actual vs base hold, theo, ratings, surveillance, records variance is possible but does not excuse weak data

Continue with Table Win, Drop, and Hold Explained, Dealer Speed and Revenue, Table Minimums and Floor Yield, and Game Profitability Ranking.

The dashboard rule

Actual win answers what happened. Theo estimates what the recorded action should produce over time. Occupancy and pace explain use of capacity. Labor measures operating burden. Error, dispute, rating, and exception measures show whether control survived. A table-games decision is credible only when all five layers point in a defensible direction.

Play smart. Gambling involves real financial risk. If the game stops being entertainment, it's time to stop playing.