A non-discretionary comp is a benefit issued under a predefined rule. The player qualifies because a published or approved condition has been met, not because a host or supervisor makes a case-by-case judgment.
Examples can include:
- loyalty points earned per unit of rated play;
- tier benefits attached to a status level;
- free play triggered by a campaign rule;
- a meal, room, or parking benefit earned at a defined threshold;
- a birthday or anniversary offer sent to an eligible segment;
- a rebate or cashback amount calculated under program terms.
The benefit is still a casino cost and control event. “Automatic” does not mean unmonitored.
How it differs from a discretionary comp
A discretionary comp allows an authorized employee to decide whether to provide a benefit and how much, within limits. It may respond to a service failure, relationship need, unusual trip, or guest circumstance.
A non-discretionary comp follows a rule such as:
[ \text{Award}=\text{qualifying activity}\times\text{earning rate} ]
If a program awards 1 point for every $5 of eligible coin-in and the player records $2,000 of eligible coin-in:
[ \frac{2{,}000}{5}=400\text{ points} ]
The employee does not decide whether 400 points feels generous. The system applies the approved rule.
For benefits generally, see Comp. For the broader valuation process, read How Comps Are Calculated.
Rules must define the earning base
A comp rule is incomplete unless it defines what activity qualifies. Depending on the program, the base may be:
- coin-in;
- rated table-game theo;
- time played;
- wager count;
- hotel or dining spend;
- net gaming revenue;
- trips or visits;
- a combination of gaming and non-gaming activity.
Exclusions also matter. Promotional credits, voided wagers, refunded transactions, employee play, duplicate accounts, self-excluded activity, or certain low-margin games may not qualify.
A phrase such as “earn points while you play” is marketing language, not a complete accounting rule.
Face value, cost, and liability
A $100 benefit can have different meanings:
- face value: the amount communicated to the player;
- redemption value: what can actually be purchased or wagered;
- incremental cost: the additional cost to provide it;
- accounting liability: the amount recognized for outstanding earned benefits;
- expected cost: value adjusted for likely redemption or breakage.
Suppose 10,000 members each earn $10 in dining credit. The face value issued is $100,000. If historical redemption is 60% and the average internal cost of redeemed food is 35% of menu value, a simple expected incremental cost estimate is:
[ 100{,}000\times0.60\times0.35=$21{,}000 ]
That estimate depends on assumptions and accounting policy. It should not be confused with the player’s $100,000 total face value or the program’s recorded liability.
Non-discretionary does not mean unconditional
Rules can include lawful conditions such as:
- valid membership and identity;
- minimum age;
- qualifying dates or products;
- expiration;
- one award per person or household;
- no transfer or resale;
- account in good standing;
- exclusion of fraudulent or duplicate activity;
- responsible-gambling and self-exclusion restrictions;
- tax or reporting requirements;
- availability limits stated in the offer.
Conditions should be disclosed clearly enough for players to understand eligibility. A casino should not call a benefit “earned” and then rely on hidden terms to deny it.
System and control requirements
Rules-based programs need strong configuration governance because one error can affect thousands of accounts. Controls should address:
- documented earning and redemption logic;
- approved effective dates;
- testing before launch;
- role-based access to rates and balances;
- change logs;
- duplicate-account detection;
- exception and manual-adjustment reporting;
- reconciliation between source activity and awarded value;
- expiration processing;
- liability and redemption reporting;
- complaint and correction procedures.
The Nevada Gaming Control Board publishes current Minimum Internal Control Standards covering computerized player-tracking and promotional-account controls. Exact requirements vary by jurisdiction, but authorization, transaction history, and auditability are common principles.
Manual adjustments are the main exception risk
Even an automatic program needs corrections. A card reader can fail, a rating can be entered late, a campaign file can omit a player, or a service issue can justify a separate benefit.
The adjustment process should record:
- player and account;
- original transaction;
- amount added or removed;
- reason code and explanation;
- employee and approver;
- date and time;
- supporting evidence;
- whether the adjustment changes liability or campaign results.
Repeated corrections by one employee, large round-number additions, expired balances repeatedly restored, or awards without qualifying activity should be reviewed.
Measuring program accuracy
Useful measures include:
[ \text{Award accuracy}=\frac{\text{correctly calculated awards}}{\text{awards tested}} ]
[ \text{Manual adjustment rate}=\frac{\text{manual value changes}}{\text{total award transactions}} ]
[ \text{Redemption rate}=\frac{\text{value redeemed}}{\text{value made available}} ]
A high redemption rate is not automatically good or bad. It can mean the benefit is attractive, the target is accurate, the cost is high, or the offer is too easy. Management should pair redemption with incremental visitation, contribution, complaints, and control exceptions.
Common misunderstandings
“The system awarded it, so it must be correct.” Configuration, source data, and account matching can all be wrong.
“Rules-based benefits never require approval.” The standard award may not, but rate changes, corrections, and exceptions should.
“Every member gets the same value.” Rules can differ by tier, product, date, segment, or qualifying activity.
“Non-discretionary means guaranteed forever.” Programs can change prospectively under valid terms, law, and notice requirements.
“Points are free.” They are a financial and marketing obligation that should be measured and reconciled.
The operational definition
A non-discretionary comp is earned under a rule, not granted as a favor. The property should be able to show the rule, qualifying activity, calculation, account posting, redemption, and any adjustment.
That consistency can improve fairness and reduce favoritism. It works only when the rules are clear, the systems are tested, and exceptions remain visible.## Consistent rules reduce favoritism but can still be unfair
Automation applies the configured rule consistently, but the rule itself can create unequal or unintended results. A program based only on coin-in may over-reward low-margin play or ignore valuable non-gaming spend. A tier threshold can produce a large benefit difference between two players separated by one point.
Management should test the effect across games, customer groups, channels, and accessibility needs. The objective is not to make every outcome equal; it is to confirm that differences follow an approved business rule rather than an unnoticed data or design bias.
A practical award and correction example
Suppose a member earns 2 points per $10 of eligible play and records $3,500. The expected award is:
[ \frac{3{,}500}{10}\times2=700\text{ points} ]
The system posts only 500 because $1,000 of activity occurred while the card reader was offline. A correction should not simply add “about 200.” Staff should identify the missing period, verify the eligible amount, calculate 200 points under the same rule, and link the adjustment to the evidence.
If the offline play cannot be supported, the property may use a separate discretionary service decision within authority. It should not label an estimate as an earned automatic award.
Campaign overlap and double funding
A player can qualify for points, tier credit, free play, a hotel offer, and a host benefit from the same activity. That may be intentional, but the total reinvestment should be visible.
Program reporting should identify stacked benefits and prevent duplicate files, repeated household offers, or two systems issuing the same entitlement. A strong control does not remove every overlap; it distinguishes planned stacking from accidental double funding.