Having a lab system, 30day eval is always free, to turn on all the alerts and see how they work is a great way to learn.
These are the ones I use or have built and rely on the most - looking forward to others input...
Node reboots
hardware health is critical (lists items in email)
interface, node, application down
I have added list of components down in the application down alert so I don't get an alert for every component
volume utilization >90% (limited to fixed disk to ignore ram)
admin other then your list (names of your admin accounts) logs in- great audit tool
snmp lastsync > dateadd(minutes,-10,GetDate()) to find stale snmp nodes
volume or interface where responding = N to find orphan objects
nodes table joined to custompollerassignnent c where nodes.machinetype Like 'cisco%' AND c.custompollerassignmentID IS NULL (maybe also poller name if you apply multiple custom pollers) to verify that your nodes have their custom pollers are applied.
Nodes not in groups if you organize that way- Some Things Solarwinds Admins Can't Live Without
