I think you’ve taken a position without confirming whether you are correct, and you’re now trying to justify it because you’re ego won’t admit that maybe you were wrong.
No, there’s nothing wrong with PANIC, because:
Rorquals are dying at record high levels
Backup means an opportunity for a bigger fight
The only reason to complain is because you lack the ability to escalate the attack when the defence escalates.
That isn’t the fault of the PANIC module and clearly there are plenty of attackers that can deal with it, especially against the large alliances, who are losing more rorquals than the small alliances. The data shows that.
This is the core premise of your opening argument.
Hey Scipio, can you link a source for that map? Super curious if there’s code to generate it or something similar, or just to see the post it came from!
Awesome post btw!
The majority of those losses weren’t during a deployment, they were either killed before reinforcements could get to them, when the users made some kind of error, or at a time when no reinforcements were available.
The most infamous example of the latter was when PL killed an entire 40-man multibox Rorqual fleet with no response because they timed their attack to when the standing fleet was under manned and there were no FCs online to respond.
It doesn’t, do your math again. Those 63 Rorqual represent about 2.9% of total losses globally and of those 63 losses some percentage would have happened anyways, considering that Goons have lost Rorquals in Delve just about every week since November of last year.
Oh, I almost forgot, that 10% figure was only for losses in Delve, but Goons and their satelite corps make up a substantial portion of the Rorqual losses in Querious as well, and those almost completely cancel out the 63 Rorquals lost that one week of deployment.
large alliance response times are damned near instant if they want it that way. They will often wait on purpose just to let you think youre getting a rorqual kill when in fact they are waiting for you to commit and then the cyno drops and you pray 1/2 your group gets away. So shorter length doesnt hurt them but it might hurt smaller groups that may need more organizing time.
Panic vs Cyno wont effect much either, cyno alt is just too easy any serious miner will make one and be able to afford paying its PLEX in one day. It will sit on grid, stealthed up, and lights a cyno on behalf of the miner. If you make all cynos inhibit all panic modules the aggessor brings and lights a cyno and the rorqual dies pretty much automatically, end of rorquals again and CCP just fixed that problem.
Punishing large groups of miners just because there are large groups of them is so counter to what EVE stands for I can hardly believe this was seriously discussed (then again this idea was being discussed on reddit and there is plenty of stupid there so i guess i shouldnt have been so surprised).
you mentioned that you believe that if CCP kept track of where rorquals are dying they would make some kind of change. Guess what they know exactly where they are dying, whom is killing them, when it is happening, etc. etc. and ARE making decisions based in part on that information.
If you happened to have noticed that large alliances lose fewer rorquals, well maybe the answer to your problem of rorquals dying has presented itself and you just dont like the answer or to state it plainly enough for a reddit poster to follow: get big or keep dying.
be prepared to answer what you think EVE stands for, because I’m pretty sure lots of people are gonna say miners are pvp fodder first and resource acquisition second.
be preppared because someone other than me is gonna pick apart whatever else you said and either defeat it with logic or just call it stupid because you suggested reddit sucks.