> For the complete documentation index, see [llms.txt](https://docs.rootcause.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.rootcause.ai/more-details/digital-twin/model-quality.md). # Reviewing Model Quality Causal discovery finds relationships; evaluation tells you how trustworthy the model actually is. The **Evaluation** tab shows overall predictive accuracy, per-variable performance, and side-by-side version comparisons. Read this before you trust simulation results. A model that predicts poorly will also simulate poorly. For context, see [Step 5: Build Digital Twin](/user-guide/creating-digital-twin.md) and [Exploring the Causal Model](/more-details/digital-twin/exploring-causal-model.md). *** ## Opening the Evaluation tab The twin's Home shows the headline accuracy figure on its **Model evaluation** card; open the card for the full view, which fills the page: the best-predicted charts and per-node metrics on the left, with a **Configuration** panel on the right holding the **Version comparison** selector (and, on multi-environment twins, an Environment scope summary).
The Model Evaluation page for a 21-variable, 33-relationship churn model: the Best Predicted Categories and Best Predicted Numeric Variables charts on the left, and the Configuration panel with its Version comparison selector on the right

The full Evaluation view. Everything on it is per variable; the charts and the table below them are the whole verdict.

*** ## Where the headline number lives The Evaluation view carries no aggregate grade: it is per variable throughout. The single headline figure is on the twin's Home instead, on the **Model evaluation** row, which reads `NN% model fit` — or "Very weak numeric fit — open for detail" when the fit is poor enough that a number would flatter it. The sentence interpreting that score sits in the row's help text. That average covers the variables the model can be scored on: observed variables with at least one cause in the graph. Two kinds are left out. Variables with no causes are not scored at all. Columns computed exactly from a formula are scored but not averaged, because the model does not learn them — a total that is the sum of two other columns would score near-perfectly and say nothing about the model. A derived column the model only *approximated* is not a formula, so it stays in. One number rarely settles anything. The variable you care most about may score very differently from the average, so read the per-variable table below before judging a model. *** ## Best predicted variables Two bar charts show up to five scored variables each: * **Best Predicted Categories** — scored by accuracy: the percentage of correct predictions. 100% means every prediction was right. * **Best Predicted Numeric Variables** — scored by R²: how much of each variable's variation the model can explain. 1.0 is perfect; 0 is no better than guessing the mean; negative means worse than that. Where no R² is available — typically a forecasting twin — the same chart uses MAPE, the average percentage the forecast is off by, where lower is better, and the section's description says so. *** ## Per-variable metrics The table at the bottom of the panel shows every variable and its performance. Click a category or boolean row to expand a per-class accuracy and F1 breakdown. On a forecasting twin, clicking a numeric row expands its MAPE curve across forecast horizons instead; numeric rows on a static twin do not expand. | Variable type | Metrics shown | | -------------------------------- | -------------------------------------------------------------------------------------------------------------------- | | Boolean / Category (static twin) | Accuracy, Precision, Recall, F1, Weighted Accuracy, AUC | | Boolean / Category (time series) | Accuracy, Precision, Recall, F1, Weighted Accuracy | | Numeric | MSE, MAE, R², Log Likelihood | | Numeric (time series) | MSE, MAE, R², Log Likelihood, plus a MAPE summary (first → last forecast horizon; expand the row for the full curve) | Reading the classification metrics: * **Accuracy** — overall correctness. * **Precision** — when the model says "yes", how often is it right? * **Recall** — of all true "yes" cases, how many did the model find? * **F1** — the harmonic mean of precision and recall. * **AUC** — discrimination ability. 1.0 is perfect; 0.5 is chance. For rare-event variables, precision and recall are usually more informative than accuracy alone. *** ## Comparing versions The **Version Comparison** dropdown selects one or more versions to chart side by side. Useful for confirming a configuration change improved things — and for spotting versions that improved some metrics while quietly degrading others. *** ## When a variable scores poorly A low metric is information, not a verdict. Common causes: * **Missing causes.** The variable's true drivers aren't in the model. * **Data quality.** Noise, errors, or too many missing values. * **Wrong model type.** A time-dependent variable in a static twin. * **Too little data.** Not enough examples to learn the pattern. If the missing driver is already a column in the Dataset, open the twin's Config tab, click **Modify model**, include the field, and retrain — see [Configuration for Existing Twin](/more-details/digital-twin/configuration.md). If it isn't in the Dataset yet, add it there first (see [Build 360 Table](/user-guide/data-views.md)), then rebuild. If the variable is genuinely hard to predict, that's a constraint to remember for any simulation that touches it. *** ## Other Working with a Digital Twin pages * [Exploring the Causal Model](/more-details/digital-twin/exploring-causal-model.md) — graph layouts and variable details. * [Inspecting Causal Relationships](/more-details/digital-twin/causal-relationships.md) — individual edges and their statistics. * [Configuration for Existing Twin](/more-details/digital-twin/configuration.md) — model settings, included variables, constraints. * [Version History](/more-details/digital-twin/version-history.md) — multiple versions of the same twin. See [Digital Twin overview](/more-details/digital-twin.md) — general overview.