Pathway Analysis and hypergeometric distribution

Question

[Mozart]

Hello there, I am trying to understand the idea behind pathway analysis and how they try to reduce gene redundancy across multiple pathways. I have just some general idea behind gene-length bias (the higher is the gene length the more likely it is differentially expressed) but I don't really get why the hypergeometric distribution is important in this sense and how it can be useful to reduce this bias.

thanks in advance.