A baccarat rating is the casino’s record of a player’s measured action. It normally includes an estimated average wager, rated time, game or table, and enough information for the property to estimate theoretical value. Comps may then be issued from that value under the casino’s reinvestment rules.
The rating is not a promise, an exact transcript, or a calculation based only on whether the player won or lost. It is an operating estimate used by the floor, hosts, marketing, and accounting systems.
What the floor is trying to record
A useful rating answers four questions:
- How much was normally at risk per decision?
- How long was the player actively rated?
- How quickly did the game produce decisions?
- Which wagers—and therefore which house advantages—made up the action?
The first two may be observed directly but still require judgment. The third is often a system assumption. The fourth can be difficult when a player changes among Banker, Player, Tie, pairs, and other side bets.
A $500 buy-in does not establish a $500 average bet. A single $2,000 wager does not necessarily make the session’s average $2,000. The floor should observe enough representative play to estimate the normal exposure, then update the rating when the pattern changes materially.
A simple baccarat-theo calculation
The standard simplified model is:
Theoretical loss = average wager × hands per hour × hours played × house edge
Where:
- average wager is the rated average amount per hand;
- hands per hour is the assumed or measured game pace;
- hours played is the rated duration;
- house edge is the expected casino advantage for the wager or weighted wager mix.
Suppose a player is rated at:
- $250 average wager;
- 65 hands per hour;
- 2.5 hours;
- 1.06% house edge, used here as an illustrative Banker-bet assumption.
Total action is:
$250 × 65 × 2.5 = $40,625
Estimated theoretical loss is:
$40,625 × 0.0106 = $430.63
If the property’s illustrative reinvestment rate were 18%:
Estimated comp budget = $430.63 × 0.18 = $77.51
The 18% is not an industry rule. Reinvestment varies by casino, market, customer segment, benefit type, trip pattern, and management policy. The calculation shows the relationship, not an entitlement.
The baccarat comp-value guide examines the cost-benefit side in more detail.
Bet mix can matter more than it appears
A player described as “$200 average” may not create one uniform edge. Consider this illustrative average hand:
- $175 on Banker at 1.06%;
- $25 on a side bet at 9%.
The weighted expected loss per hand is calculated separately:
Expected loss per hand = ($175 × 0.0106) + ($25 × 0.09)
= $1.855 + $2.25 = $4.105
Weighted house edge on the full $200 action is:
$4.105 ÷ $200 = 2.0525%
The $25 side bet represents only 12.5% of the money wagered but contributes more expected loss than the $175 main wager in this example.
A rating system that assigns one generic baccarat edge may not capture that mix precisely. Some properties rate the main game only, some have configured wager categories, and some rely on floor judgment. Players should not make a high-edge side bet merely to appear more valuable. The expected loss created to earn an extra comp usually exceeds the extra benefit.
Rated time is active time, not time near the table
A player can be seated for three hours but absent for meals, phone calls, or long breaks. A disciplined rating should reflect the property’s definition of active play, not the time the chair was occupied.
The opening and closing time can also be wrong. Common causes include:
- the player begins before presenting a card;
- the floor opens the rating after several hands;
- the player moves tables without a clean transfer;
- a rating remains open after departure;
- a system interruption requires manual reconstruction;
- a crowded pit delays updates.
These errors matter because time multiplies every other input. At a $300 average, 70 hands per hour, and 1.06% edge, an extra half-hour adds:
$300 × 70 × 0.5 × 0.0106 = $111.30 theoretical loss
At a 15% reinvestment rate, that timing error changes the illustrative comp budget by about:
$111.30 × 0.15 = $16.70
That may be small on one rating but material across many ratings or a high-limit player population.
The separate article on why time played matters for comps focuses on this multiplier.
Actual loss and theoretical loss answer different questions
Actual result is what happened during the session: the player’s net win or loss under the casino’s accounting definition.
Theoretical loss is the long-run expected casino win from the rated action.
A player can win $20,000 and still generate $500 in theoretical value. Another can lose $20,000 during a short, low-theo session. Actual loss is volatile; theo is designed to normalize that volatility.
Casinos may consider both. Theo often supports routine reinvestment because it is more stable. Actual loss may affect discretionary service, trip review, or a host decision, particularly when the result is unusually large. The balance is property-specific.
This is why two players with the same final loss can receive different treatment, and two players with the same theoretical value can still receive different offers based on market, room demand, historical response, credit relationship, or service considerations.
What can be a comp
Baccarat comps may include:
- food or beverage credits;
- hotel rooms or room discounts;
- free play or promotional chips;
- event or tournament invitations;
- transportation;
- host-arranged discretionary benefits;
- tier points or benefits under a loyalty program.
Those items do not have the same cost to the casino. A room with spare capacity may have a lower incremental cost than its public rate, while free play has rules that affect expected redemption cost. A host may also have authority limits and documentation requirements.
The glossary entry comps separates earned, promotional, discretionary, and service-recovery benefits.
Why the rating needs controls
A rating can create financial value. Inflated average bets, false time, duplicate accounts, or unsupported point adjustments can cause overcomping. Understated ratings can damage legitimate customer service and create disputes.
Nevada’s Table Games Minimum Internal Control Standards distinguish point-accumulation systems from rating-only systems in which employee judgment drives the complimentary decision. They also require certain manual point changes to be documented and authorized. The exact controls differ elsewhere, but the operational lesson is broad: ratings and redeemable benefits need traceable access, adjustments, and review.
Useful checks include:
- ratings open for implausibly long periods;
- average wagers inconsistent with table limits or observed buy-in;
- repeated manual point additions;
- the same person rated simultaneously at distant tables;
- large host adjustments without supporting reason;
- frequent player complaints about missing time or action;
- unexplained differences between system assumptions and actual table pace.
These checks should trigger review, not automatic accusations.
How a player should use the rating
A player who already intends to play can make the record more accurate by presenting the card before the session, confirming that the rating is open, and notifying the floor when moving tables. At the end, the player may ask whether the session time and approximate average were captured; disclosure policies vary.
Do not increase a wager, switch to a poor side bet, or remain at the table solely to earn a meal or room. The expected cost of creating the comp is usually larger than the value returned.
A simple personal test is:
Net comp value = value you will actually use − extra expected loss created to earn it
Suppose staying one additional hour creates $300 × 65 × 1 × 1.06% = $206.70 in theoretical loss. Even if the extra hour produces an additional $35 comp under the property’s rules:
$35 − $206.70 = −$171.70
The session can still win because actual outcomes vary. The calculation shows that extending play for the comp is a poor exchange under the stated assumptions.
Use the expected-loss calculator to test wager size, pace, time, and edge. For the game-specific action model, continue with baccarat hands per hour and total action and expected loss per hour.