Serious errors in randomisation are more likely when a complex allocation algorithm is used to try to balance covariates. So an (unnecessary) attempt to balance covariates actually results in systematic imbalance! An example of harm caused by the myth that RCTs require balance to be valid.