Page rank intuition

[Andrew Spruce]

Hi,

My first post so be gentle.

I have a directed graph of projects and i am trying to establish a good proxy for the highest demand project.

I would expect this to be the node with the largest number of unique antecedent nodes (identified by recursively collecting the results of DG.pred[‘node’] and removing dups) but when i stack rank nodes based on the number of unique antecedent nodes and compare with the page rank they are quite different.

What might account for this? 

Would the popularity of an immediately antecedent node boost page rank in a way the raw antecedent count would not? How does the dampening affect the page ranking in this case?

Thanks in advance