Reconciles two values of one KPI into quantified effects
A walk from one value to the other through named effects (population, aggregation, time, currency), each computed from the data in percentage points
Can AI do this?
Not settled yet The benchmark tested this, but the models' answers do not settle yet whether they master it. 0 of 4 models in the current benchmark master it.
| Model | Tier | Probability of mastery | Status |
|---|---|---|---|
| gemini-4-argon | Flagship | 53% | Not shown |
| gemini-3.5-flash-lite | Low cost | 16% | Not shown |
| gpt-6-luna | Low cost | 51% | Not shown |
| claude-haiku-5-5 | Low cost | 48% | Not shown |
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: A walk from one value to the other through named effects (population, aggregation, time, currency), each computed from the data in percentage points
How it is assessed: The effects sum to the difference and each is computed, not asserted
Related knobits
Builds on
- W-251 Explains the components of a metric definition
- W-252 Explains additivity of facts
- W-253 Explains the EBA definition of non-performing exposures
- W-1801 Explains the meaning of a percentage point
All 19 knobits of Define metrics and KPIs.
Source: Graded on its own by dwbi-metric-definitions