Create standardized decision rules for teams and repeatability.
Grade Or Hold Decision Rules
A repeatable rulebook helps you stop guessing. Instead of debating every card, your team can route candidates into grade, hold, or sell-raw queues using shared thresholds.
Why most grading decisions drift over time
Inconsistent calls usually come from changing standards: one reviewer is strict on surface, another overweights centering, and everyone handles borderline cards differently.
A written rulebook protects margin by enforcing the same criteria across cards, reviewers, and submission windows.
Rulebook framework
- Define pass/fail thresholds per defect class.Set explicit limits for centering, corners, edges, and surface.
- Map confidence bands to actions. High-confidence upside goes to grade, mixed signals to hold, low EV to raw.
- Add economic guardrails. Require break-even and expected-value checks before any card enters a paid submission queue.
- Create an escalation path. Any uncertain card is second-reviewed, not decided by gut feel in the moment.
- Review outcomes monthly. Compare predicted vs. actual grades and tighten rules where misses cluster.
Example policy snippets
- Grade: clean surface + acceptable centering + strong expected value after fees.
- Hold: one uncertain attribute or weak market timing, pending re-check.
- Raw: clear grade ceiling or negative EV at current pricing.
FAQ
How often should we update the rulebook?
Monthly is a solid default. Update faster if fee structures, market comps, or your observed miss patterns change.
Should every reviewer use the same thresholds?
Yes. Personal style creates noise. One shared standard is the point of a team decision system.
What is the biggest mistake in grade/hold rules?
Leaving "exceptions" undefined. If exceptions are frequent and undocumented, the rulebook no longer controls outcomes.
Take action
Build your rulebook with real card data, confidence ranges, and EV checks in Pregrade.