Recognises blocking keeping true matches apart
Blocking on a field that changes.
Can AI do this?
AI does this well Most of the tested AI models master this. 3 of 4 models in the current benchmark master it.
| Model | Tier | Probability of mastery | Status |
|---|---|---|---|
| gemini-4-argon | Flagship | 92% | Mastered |
| gemini-3.5-flash-lite | Low cost | 84% | Not shown |
| gpt-6-luna | Low cost | 93% | Mastered |
| claude-haiku-5-5 | Low cost | 92% | Mastered |
Mastered means a probability of at least 90% that the model has this knobit, estimated from its graded answers. “Not shown” means the answers do not settle it either way. How the models are tested.
What it looks like in practice
Expected result: Blocking passes that keep true matches together
How it is assessed: Flags the deficient case and passes the acceptable one. Deficient: BK-01 on surname; a married woman's records never compared Acceptable: Several passes: date of birth + first name; personal code; phone
Also known as: Blocking that misses true matches
Related knobits
Builds on
All 16 knobits of Define the matching rules.
Source: Review checklist 03-matching, row 5 (DMBOK)