I think I mentioned this in another thread, but the goal is not to avoid giving a penalty to projects, but the goal is to come up with fair way to relatively rank every single project against every single other one. By factoring the amount of ratings a project has will avoid any “unfair” moments where a project with low number of ratings with less statistical significance ranks higher than a project with a similar score that has higher number of ratings with higher statistical significance.
That said though, in the future we may expose this option for hosts but given the complexity of understanding the dynamics of the ranking, by having this as an option, I’m concerned hosts may choose something that they may think is better but actually reduces the quality of the relative ranking.
Are you hosting a jam where you think that the median has degraded the ranking?