They don't necessarily require audit data so defined that it records all Navigation / Interaction within the game for all cases at all times. They already store plenty of data on each character by default; such as quests completed and completion time stamps. Basically anything to do with Achievements has to be stored in an easily accessible way.
Use these statistics to narrow down the reports to the most likely HB culprits, then start to run audits on those accounts. Take a sample of that Audit to actually analyze, we don't need to see what someone did for an 8 hour time period, maybe only 2 random 30 minute sections. Data cut down by that much would / could still highlight which Accounts have been using HB reliably while still being manageable. I also assume they spent more than a week pulling together all these flagged accounts into the Ban Bin to be hit with the Hammer. I would assume they have been doing this for many months.
I'm not saying what I'm suggesting is 100% the way Blizzard Banned via Wave but it's the type of method I think they used; maybe even hope because if not, and Honorbuddy is detected, we may see 6 Month bans on the regular and the slow death of HB and it's community.
Thanks for the well thought out response Roboto, you are correct, it would be a pain in the ass, but given enough time it is feasible and imo just as possible as putting malware on all our PCs or fooling the Bosslands team for the first time in 5 years or etc...
your theory is possible, however unlikely as there have been way more effective ways to detect honorbuddy.
just a simple example: prior banwave the process itself was visible on the system - now it is no longer (try for yourself, error 0x5)
everyone knew this was the fact but since the lang-day-ago court cases of the EFF against blizzard, scanning ram and processes was not allowed in area, not belonging to their own processes _by law_ - but still technicly possible. (yaya i know, the gm automated reponses and emails still state this as a fact...)
second, the ban wave way way too precise because it hit (almost?) everyone who was using honorbuddy in whatever way in a specific timespan.
this also includes manual botting, such as raid only with custom movement featuring alternate CRs and custom botbases
and how do we explain the banwave, happening to another "quite popular" bot, cmpeting with honorbuddy, a few days after? they litterally know who was using what bot.
since (almost) all professional wow bots work in the exact way, it would be way to hard to differ these two bots
on a side note: i lost a mule account in the banwave which was used for only 5-10 mintues of basic questing (did something like lvl 4-10) - the chances of this account matching reliable against any pattern should be near zero
third, IF they got this pattern, it wouldnt have changed, post banwave, the bot didnt change that much apart from the way as it injects into wow - i think there's an apoc quote in the old banwave thread about it - so if it was patterns and player reports, why did bans drop to almost zero? I honstly dont believe it's the lack of botters
I never saw that, link please? Last i heard Bosslands did not know or were not going to reveal any discoveries they have made.
dispite his wording he is right; in fact, I remember a post something alike this as well, but it was never confirmed by the core devs (at least as far as i know)