[Genome] Chain Vs Net tables

[Fungazid]

Hello UCSC team, 

Thank you a lot for your help so far.

As far as I understand chain tables represents all
possible conserved regions of 2 given species. I saw
that chain table tend to link specific target regions
to many query regions (one to many).  For example:
specific human target region is linked to 10 different
regions in mouse genome. In order to reduce the
enormous complexity of the chain representation, I
thought of using net table. 
1) But what are the pro and cons of each table (net Vs
chain)?
2) Does net table better represents orthologous
regions than chain?  
3) About the levels (‘level’ column in MYSQL net*
table): Are there overlaps between levels (for example
level-1 position is chr1:100-200 and level-2 position
is chr1:150-250)? 
3) Why there are much less rows with level=1 than rows
with level=2 (‘level’ column in MYSQL net* table)?


Best wishes,
thank you, 
Avi