There is the diceware method to make passwords random and strong to crack. But is it advisable to use this method to make random names for email addresses? That is the concern I have, when a person makes a random email address, they inevitably introduce things with which they can be identified, breaking the previously stated randomness, what ways are there to avoid this?
Do you need your opsec to be resistant to state-level actors (oppressive regime, censorship, illegal activities)? Well then you need to make sure you don't introduce anything that will deanonomize you.
Are you trying to be resistant to mass data collection efforts used for profit? Being on the pseudonymity spectrum is a good step.
Dealing with the latter is like dealing with a bully. Make it not worth their time. They just want to put you in bucket X so they can estimate the most likely way to influence you for reason Y. Pseudonymity is about having multiple aliases that get put into different buckets so their privacy invasive efforts are less effective.
That is a good idea just so that you don't have to think about any potential privacy issues. Your email could be {firstword}{secondword}{4 numbers} and so long as the words and numbers are randomly generated, you can avoid accidentally including personal references or biases.
Your username does not need to be high-entropy, though. It will be semi-public. So it's not about strength against dictionary attack or similar, it is just about leaving the selection process up to a random process that isn't witnessed by a third party. You can write scripts that will generate these kinds of things using Python and the faker library.
like why do you need email addresses? use temp mail for garbage registrations and two-three stationary emails which you remember for stuff which is linked anyway
I sort of do this because I own my domain. I generally pick an annual keyword email filters can lock on, followed by an identifier with whom I'm contacting.
It's easy to trace if addressed get breached, especially unreported breaches, and add to a burn list if they get spammed.
Also, if I have no intention of responding I give fake info or if I need that rare password reset link I know when to look in the spam.
Yeah, using my domain is it's self a bit trackable, but enough friends and family use it I figure poisoned data is sweet justice.
Fun fact, but for some reason old fake accounts have boomed in popularity; like data brokers with bad information bounce verifications off each other, linked it to some poor sap in another state, and snowballed into an actual profile. I'm going to use that identity as an alt profile for something someday.