If I set up my own guard node and always use it when connecting through the Tor network, would I be fully protected from confirmation attacks?
Also, how would my anonymity be affected with such a setup? Would this setup be pointless if an attacker seized/shutdown my Tor guard node?
I'm assuming you mean traffic confirmation attacks where the first and last relays use traffic patterns to link the incoming and outgoing traffic?
The bridge replaces your guard relay, so the circuit looks like:
Tor client (you) <--> Bridge (you) <--> Entry relay <--> Exit relay <--> Destination
If the entry relay and exit relay are cooperating maliciously, then they can easily link the bridge's IP address to the websites its circuits connect to. If the malicious party knows that you run the bridge or that your personal IP address is the only one to connect to that bridge (for example by asking the company running your bridge server), then they can link that traffic to you. Now traffic confirmation attacks aren't needed at all since there are only 2 hops in the network, and they can communicate directly.
