Feature Importance SHAP via Random Forest Regression

The SHAP results from the attached workflow and data seem erratic and I cannot discern any specific scoring pattern of which features have the highest or lowest impact on the dependent variable.

Would someone either: 1) explain how I can use the results to rank independent variables impact on the dependent variable, or 2) explain what I did wrong and how to correct the workflow.

Thank you

Feature Importance.knwf (194.8 KB)

Sold Prices by ListingID.xlsx (60.5 KB)

temp for shap testing.xlsx (1016.9 KB)