The default option for GrowthExperiments' A/B testing system is named 'null' (as in "null hypothesis"), which is a poor choice as it can be mistaken to the special null value most programming languages have. This leads to all kinds of confusing situations, e.g. ge.utils.setUserVariant(null) and ge.utils.setUserVariant('null') do similar but not identical things.
Changing the name will be harder once the A/B test starts so let's change it before then, e.g. to 'control' (as in "control group").