Question: How does ridge regression work?

My answer: OLS models are BLUE – best linear unbiased estimateors.

But sometimes forcing unbiasedness causes other problems. In particular, if the independent variables are fairly collinear, then the variances of the parameter estimates will be huge and small differences in the input data can make huge differences in the parameter estimates.

Ridge regression allows some bias in order to lower the variance.

Ridge should be considered with other shrinkage methods such as lasso and elastic net

I can’t say that that is wrong, but in my experience, ridge’s main advantages are when there is colinearity.