| name | tag-grading-system |
| description | Use when submitting modern cards for AI-precise sub-grade reading, or when customer wants algorithmic transparency.
|
TAG Grading System
TAG uses computer vision to assign 1-1000 sub-grade per axis. Publishes every score on the slab. Best for 'highest TAG' listing titles.
The 1000-scale
TAG scores centering, corners, edges, surface on 1-1000 (rounded to 10s). A card scores 950+ on all axes for TAG 10 (Pristine). Collectors can see if your TAG 10 is a 950 or a 998 — the 998 commands a unique listing premium.
Market position 2024-2026
- Modern Pokemon TAG 10 ≈ 0.6–0.9× PSA 10 (closing gap)
- Modern sports TAG 10 ≈ 0.5–0.8× PSA 10
- TAG 998+ titles outperform: unique 'highest TAG' premium
When to choose TAG
- PSA 10 candidate but customer wants second-opinion data
- Marketing the card as 'algorithmically verified'
- Consigning to a TAG-promoting auction house
- Expecting a sub-grade flaw and wanting transparency over ambiguity
Where this fits in the X3 empire
Powers our 'AI grader comparison' marketing — TAG validates the broader thesis that ML can grade. Strategic to recommend in our customer surfaces.