How does Random Forest behave under high noise-to-signal ratio?

Performance degrades but remains more stable than single decision trees.

High noise reduces split quality, but ensemble averaging reduces variance. However, if signal is extremely weak, model may still overfit noise patterns.