Good values for R2 metric?

The R² value (also called the coefficient of determination) measures how well a regression model fits your data. It ranges from 0 to 1:

  • R² = 0: The model explains none of the variability.

  • R² = 1: The model explains all of the variability (perfect fit, but can be a sign of overfitting).

  • 0 < R² < 1: The model explains some of the variability — the closer to 1, the better the fit.

General Guidelines for R² Value Interpretation

R² Value

Interpretation

0.9 – 1.0

Excellent fit – The model explains nearly all variability.

0.7 – 0.9

Strong fit – The model captures most of the data variance.

0.5 – 0.7

Moderate fit – Acceptable in many real-world applications.

< 0.5

Weak fit – The model doesn’t explain much; improvements are likely needed.

= 1.0 exactly

Too perfect? – Often indicates overfitting, especially on training data.

 

  • In marketing or social sciences, even 0.3–0.5 may be acceptable.

  • In physics or engineering, values above 0.9 may be expected.

Why R² Alone Isn’t Enough

While R² gives a good idea of fit, you should also check:

  • Adjusted R² – accounts for the number of predictors.

  • RMSE / MAE – tells you how far off predictions are.

  • Residual plots – help detect patterns your model might be missing.

©2025 Samyak Infotech Pvt Ltd. | All trademarks, images and logos are the property of their respective owners.