Probably. A lot of things have changed since then, and I didn’t think of considering it until afterwards, when I added my edit.
Simple: the purge can only be triggered during the purge window. Get your purge meter over the line before the window and log off.
I think you’re conflating the logic for the purge meter and for the purge trigger a bit more than warranted. There’s one meter per clan, but the purge triggering is a server-wide thing. I base that latter claim on the fact that in all these years, I’ve never once heard a clan say “help, we’re getting purged” and another clan go “shіt, we’re getting purged too”, which would indicate that there’s only one purge happening at a time.
So when the purge window starts and the server decides that it would be a nice moment to have one happen, it then has to choose where and whom to purge. The exact details of these algorithm are not known, which is why we’re all guessing here how it works
My theory is that it works along the lines of querying the game database for all eligible purge candidate locations and then selects a random one. Maybe there’s some prioritization and weighting going on, who knows. I certainly don’t know
A small correction: what I pointed out is that the value 0 has a special meaning. The setting is players per clan, yes, but the value 0 means you need at least one player on the server, otherwise the server won’t even bother felling a tree in the forest when there’s no one to hear it fall
At least, that’s what my test showed back then.
The most likely explanations are that I either screwed up my tests back then somehow, or that the logic changed
However, there are other details we don’t know that might be at play here. For example, I don’t know whether the trigger algorithm prioritizes online candidates.
Another thing to consider is how many purges happened during your tests. The purge window on official PVE-C servers is only 4 hours long, and each purge will have a 10 minute “preparation” period before spawning the first wave and a 1 minute regroup period between the waves. Plus, there’s a 5 minute delay between two different purges. If you could beat each wave instantaneously, that would still come out to a theoretical maximum of 12 purges (6 wave purges, to be precise) per window.
Like I said, it’s more likely that the logic either changed or I screwed something up, but now I’m warming to this topic. I might even decide to test it again