Interesting findings regarding rare respawns (Molten Front/Skoll issues)

Anything related to Hunter pets.
Forum rules
Treat others with respect. Report, don't respond. Read the complete forum rules.
User avatar
Vephriel
Illustrious Master Hunter
Illustrious Master Hunter
Posts: 16357
Joined: Wed Jan 06, 2010 8:07 pm
Realm: Wyrmrest Accord US
Gender: Female
Location: Canada

Interesting findings regarding rare respawns (Molten Front/Skoll issues)

Unread post by Vephriel »

Hey all, I wanted to share this research someone posted in the Secret Finding community. A user who goes by the name Kormex discovered this spawning behaviour in regards to mounts like TLPD and the Frightened Kodo, but I believe it can affect some of our hunter tames. Namely Skoll and the Molten Front rares which is why I wanted to post it here, because I remember we were concerned about the Molten Front recently. I was given permission to share this. :D

Again this is essentially a hypothesis right now, but there's a lot of backing to it and it seems something that's worth keeping an eye on.

"TL;DR (Why did you write an entire essay about this Kormex, I don't wanna read 10 pages of background info):
If certain zones are empty for 15 minutes, they reset, which also resets all rares and spawn timers. This is why spawns in various zones are funky on dead servers."

After a lot of observation and testing, why some rares are not up on low/medium pop servers as often as they should has been solved. It appears that some zones will "reset" if completely empty for 15 minutes. This means that every NPC in them will disappear, and the next time someone enters the zone, they will load back in all at once, and all rare timers will reset to 0. It's basically as if server maintenance happened, except ONLY for that zone, and it will stay down until someone enters it.

This doesn't typically happen on the big CRZs, as the zones will never be empty for 15 minutes straight. It also happens less often on high population servers, as well as higher traffic zones. If anyone enters less than 15 minutes from the previous last person leaving, the reset will not occur. This means it's still possible for lower population servers to go unaffected if there's enough traffic at any given time.

The current zones I have tested and observed this happening in are:
Storm Peaks, New Darkshore, every Draneor zone except SMV and Tanaan (did not test Tanaan), and Azuna.
I have heard from others that this also happens in Molten Front and Timeless isle, but did not test either myself.

Now, for evidence! First of all, there's simple observations that many people have made about the affected zones. We've known for a long time that Storm Peaks spawns were not functioning correctly, and that something appeared to be weird with Voidtalon outside of SMV on RP servers. However, we didn't know exactly why, or what exactly was happening.
I first got suspicious after messing around with an observation that Allie had about Frightened Kodo. She said that if she couldn't find kodo doing laps, she would take a break and then come back, and he would be more likely to be up after that. I tried this myself, and had great results - 45 minutes of doing laps yielded nothing, but after a break I found him on the first server I tried. We also noticed that we would often get weirdly long loading screens in Darkshore sometimes.

It turns out, these two observations were connected, and a long loading screen is actually the que that kodo will be up. A regular load time for me was about 5 seconds, while the long load was 40-65 seconds. As you know, kodo is up right after server resets happen, which is what lead me to believe that the long load was actually the server having to take extra time to load up the zone from nothing, and kodo was up due to it being a mini-reset. After further testing, I found that the minimum time of me being offline for this to happen was 15 minutes.

I then wondered if this might also be the explanation for Storm Peak's weird spawns. My biggest initial piece of evidence was how rare it was for me to get spawns in under 2 hours on empty servers. With the spawn being 2-8 hours, the zone resetting could explain this.
To test my theory, I found a Skoll. I checked /who on both alliance and horde, and there was no one present but me. I then logged out, waited 15 minutes, and logged back in. Skoll was nowhere to be seen. I checked /who on both factions again, and there was still no one there.
I then decided to repeat the experiment to test out any anomalies, but using gathering nodes instead of Skoll so I wouldn't need to camp another one every time. Gathering nodes are up at reset, but they get re-randomized when that happens, and there's a cap on how many can be up at a time. So, I found a clump of 3 gathering nodes, noted what they were, and logged out for 15 minutes again. When I got back in, 2 of the nodes were gone, and the 3rd was a different material.

I also timed my loading screens, and my regular load time was once again 5 seconds. My long reset loading time was 15 seconds, much less noticeable than the Darkshore time, but still significantly longer. My loading bar also seems to get stuck in the same spot for the extra time, regardless of what zone I test.

An additional observation I've had in the past is that several times, when flying into a reset zone instead of logging in, every NPC in the zone will be absent for 10-30 seconds, before suddenly all appearing at once. I also will not be able to cast anything, as if I'm lagging, until the NPCs load in. I suspect this is what is normally causing and being masked by the long load screens.

I was able to replicate the gathering node/load screen test on Draenor: in Frostfire Ridge, Gorgrond, Talador, Nagrand, and Spires of Arak. I was NOT able to make it happen in Shadowmoon Valley, even when using the PTR at a dead time of night. This lines up with previous observations that SMV spawns were much more consistent on RP servers than other zones. Whether or not the zone reset also restarts the 48 hour waiting time for VT is unclear.
I also have had the test show resets happening in Azuna. This also lines up with observations that crystals are often easier to find up on higher population RP servers, and are sometimes weirdly hard to find on lower population RP servers, especially with warmode on. I would like to test it with a set of crystals up, to confirm they despawn if they were previously up, but have not had a chance to.


Now, what does this all end up meaning? For the most part, it just means that suspicions we've had for a long time are confirmed, not just coincidence. Things like needing to camp for TLPD to spawn, and SMV being better for VT on RP servers. It also means that high pop RP servers / the smaller CRZ groups are ideal for Draenor Rares and Long Forgotten Hippogryph, rather than low/medium pop RP servers. For Frightened kodo, it can actually be exploited for easy kodo finds by setting up trials ahead of time, then logging out for 15 minutes in between each check.

Oh I should also mention it doesn't happen in every zone, I've only confirmed it for the ones I listed.
User avatar
Wain
The Insane
The Insane
Posts: 13515
Joined: Tue Jul 27, 2010 1:54 am
Gender: Male

Re: Interesting findings regarding rare respawns (Molten Front/Skoll issues)

Unread post by Wain »

Huh, interesting! Also, a pretty flawed system if it really is that way :/

Thanks for posting it :)
Shaman avatar by Spiritbinder.
User avatar
Talihawk
Petopia Artist
Posts: 642
Joined: Sun Oct 16, 2016 2:55 pm
Realm: Dalaran

Re: Interesting findings regarding rare respawns (Molten Front/Skoll issues)

Unread post by Talihawk »

Maybe it's to conserve server resources?
User avatar
Vephriel
Illustrious Master Hunter
Illustrious Master Hunter
Posts: 16357
Joined: Wed Jan 06, 2010 8:07 pm
Realm: Wyrmrest Accord US
Gender: Female
Location: Canada

Re: Interesting findings regarding rare respawns (Molten Front/Skoll issues)

Unread post by Vephriel »

Just wanted to add that tonight Kormex proved this theory 100% using a TLPD corpse, so this definitely seems to be how these areas are working right now.
Post Reply