bugs bugs bugs (5.4)

[3wd]

I added a couple PvP-related issues that I had forgotten.

Can anyone comment on the Sleeping Gas bug? Sleeping Gas is known to be able to chain-sleep pets, in spite of Resilience. I'm wondering if this is because Resilience is not correctly proccing from Sleep abilities? Or, there is another oddity with Resilience that could be responsible. The Resilience buff pops up when the CC debuff wears off, not when the CC is applied. Therefore, if another CC hits before the old CC debuff wears off, then Resilience is not in place and the new CC will be applied. I'm not sure if this bug is responsible for the chain-sleeping behavior of Sleeping Gas, or if there is a distinct bug with Sleeping Gas.

Sleep has been always classified as a different category of CC skill, its mechanic is different than the other type of stun or shape shifting skills.

It is a known fact that Critters will not immue to sleep but able to immue to all other types of CC. (I remember this was mentioned many times before and maybe even fixed through a hotfix lately)

Therefore, I am not surprised that sleep has been excluded from the resilience buff.

[Poofah]

I agree the implementation seems different for Sleep, at least in that it can be broken by damage whereas stun/poly can't. But Resilience is supposed to work against all 3. Here's the [url=http://www.wowhead.com/petability=924]Resilient tooltip on wowhead[/url], it specifically says 'immune to stun, sleep, and polymorph'. And here's the [url=http://www.wowhead.com/petability=236]Critter passive tooltip[/url], it also specifically says 'Critters are immune to stun, root, and sleep effects.' Maybe Sleep is intended to be different (breaks on damage but not avoidable via critter/resilience?). But the tooltips need an update if that's the case.

[Ligre]

Actually, Critters ARE immune to the Sleep effect from Sleeping Gas. They are NOT immune, however, to the Sleep effect from Most Dust. This is yet another annoying bug that needs fixing.

[Kpb321]

I added a couple PvP-related issues that I had forgotten.

Can anyone comment on the Sleeping Gas bug? Sleeping Gas is known to be able to chain-sleep pets, in spite of Resilience. I'm wondering if this is because Resilience is not correctly proccing from Sleep abilities? Or, there is another oddity with Resilience that could be responsible. The Resilience buff pops up when the CC debuff wears off, not when the CC is applied. Therefore, if another CC hits before the old CC debuff wears off, then Resilience is not in place and the new CC will be applied. I'm not sure if this bug is responsible for the chain-sleeping behavior of Sleeping Gas, or if there is a distinct bug with Sleeping Gas.

If I'm not mistaken the Resilience buff only gets applied when sleep actually wears off naturally. If you breaking it early by dealing damage you get no resilience and no protection from CC. Try it out with a strider. Soothe something and pass or heal the next round so it gets the asleep debuff and then attack it. Resilience won't appear. I'd imagine it's the same thing with sleeping gas so if you are just spamming the ability your opponent will probably never get resilience and you could sleep them multiple times in a row you just need to get that 25% chance multiple times in a row. If you sleep and then heal with an infinite whelpling you should get resilience.

[Jerebear]

The resilience bug probably happens with stuns too, but harder to test since they all have cooldowns. I would imagine you would have to get stunned and kill the opposing pet with Dot Damage and then hope the next pet stuns you. Maybe a coordinated duel over vent? Frog kiss can chain as well. I see it all the time facing frogs in Jade Forest.

[Poofah]

Actually, Critters ARE immune to the Sleep effect from Sleeping Gas. They are NOT immune, however, to the Sleep effect from Most Dust. This is yet another annoying bug that needs fixing.

That's just bizarre. The intuitive way to implement Resilient would be to have it apply as soon as a 'sleep', 'stun', or 'polymorph' debuff fades. As far as I know, there is no distinction between the stun debuff applied by e.g. Crystal Prison and Geyser: there is only one tooltip for the 'stunned' debuff applied by these abilities, it's [url=http://www.wowhead.com/petability=927]here[/url] (and it's also the same as the stuns applied by the interrupt moves, e.g. Horn Attack and Kick). Similarly, the 'asleep' debuff applied by Sleeping Gas and Moth Dust [url=http://www.wowhead.com/petability=498]is the same[/url]. So clearly the game is applying Resilient differently, based on the ability that applies the stun/sleep/poly. Ie every stun/sleep/poly ability needs to have its own tag that tells the game to apply Resilient after the CC fades. That creates all kinds of opportunity for individual abilities to slip through the cracks.

If anyone is really interested, there's actually 3 kinds of 'stunned' (most stuns fall into one category, but there's a separate category for self-stuns (e.g. Ghostly Bite), and Amber Prison has its own category), and 2 kinds of 'asleep' (1 turn sleeps, and 2 turn sleeps such as Food Coma)

http://www.wowhead.com/search?q=stunned#petabilities
http://www.wowhead.com/search?q=asleep#petabilities

If I'm not mistaken the Resilience buff only gets applied when sleep actually wears off naturally. If you breaking it early by dealing damage you get no resilience and no protection from CC... I'd imagine it's the same thing with sleeping gas so if you are just spamming the ability your opponent will probably never get resilience and you could sleep them multiple times in a row

That's interesting. That's almost certainly why we see Sleeping Gas chain more than Frog Kiss: if you spam Sleeping Gas, then each hit removes the old 'asleep' debuff before Resilient can be applied. So it's not necessary to get two procs in a row. Whereas Polymorph is not broken by damage. So Frog Kiss will not remove Polymorphed when it deals damage, so then the Polymorph debuff will always still be present at the end of the turn, and it will apply Resilient if you proc once and then fail to proc on the following turn. So the only way to chain Frog Kiss is to get two procs back-to-back.

The resilience bug probably happens with stuns too, but harder to test since they all have cooldowns.

Yes, it definitely happens with stuns: this is easiest to test with a delayed ability like Geyser or Elementium Bolt, since you can swap in a pet with e.g. Crystal Prison before the Bolt lands. And similarly for Frog Kiss, it definitely can chain, but getting chains is less likely than with Sleeping Gas, so we tend to see it less. I'm 95% convinced now that the Sleep bug and Resilient bug are really the same thing.

[Adezero]

Elemental pets should now correctly be taking less damage while the Sandstorm effect is active.

http://blue.mmo-champion.com/topic/2960 ... bruary-24/

[Awq]

9) the Haunt bug. I only put this here for the sake of completion, as I've seen it posted many times, but have never been able to recreate it myself. Reportedly, a Haunting Val'kyr will occasionally fail to rez (even though the Haunt debuff was successfully applied to the enemy pet). If you have information on how to recreate this, please share.

I use (and go against) Val'kyr all the time, and I have positively never seen her fail to revive from a Haunt when she is supposed to. The only times she fails to revive are when she is supposed to fail to revive (eg: when losing in a game of Haunt roulette). I strongly suspect that reports of this alleged bug are simply Val'kyr owners unaware that certain conditions can cause the revive to fail.

I expected it to happen against Stone Rush, Haunt, Explode, Armageddon. It happened in other situations as well!
I am unaware what the exact conditions are!

My Val'Kyr would frequently fail to resurrect when I used Val'Kyr, Scalded Basilisk Hatchling, Ruby Droplet.
It happened once every 10-15 matches.

I would use Haunt (and sometimes Curse of Doom) on an enemy pet. After that I would swap Basilisk Hatchling. I would let Haunt (and Curse of doom) kill the target, while I controlled the enemy pet.
If I used Crystal Prison the first round I would generally(?) be fine. If I used my stun the second round (if a pet had prowl/howl, or when I wanted to activate the beast racial), the chance that my Val'Kyr would not resurrect were increased. It came to the point where I did not want to save my stun even if it would give me a free round, because of a lower chance of resurrection. This might have been superstition on my part, but I did change my strategy to deal with it.

It might be because of how Feign Death interacts with Haunt, or it might have been because my Basilisk Hatchling/Val'Kyr killed a stunned target. I have not used this team in quite a while! I used it at the start of 5.4. The bug might have been fixed in a patch/hotfix.

This bug has not happened for Val'Kyr teams that focused on avoidance. It vaguely remember that it also happened for Val'Kyr + Tiny Twister combination (Wild Winds, Bash, Cyclone).

[Poofah]

Elemental pets should now correctly be taking less damage while the Sandstorm effect is active.

http://blue.mmo-champion.com/topic/2960 ... bruary-24/

This didn't work for me yesterday (2.24), however this morning Sandstorm properly shielded Wastewalker Shu's pet, and also my Living Sandling. Can anyone else confirm/deny that elementals are getting the Sandstorm shield ingame?

It vaguely remember that it also happened for Val'Kyr + Tiny Twister combination (Wild Winds, Bash, Cyclone).

This one could be explained by Cyclone killing the Haunted pet before Haunt could properly expire. It seems that end-of-turn effects that go off before an expiring Haunt can sometimes cause the pet to die without Haunt properly rezzing the Valk (posts are on page 1, toward the bottom).

It might be because of how Feign Death interacts with Haunt, or it might have been because my Basilisk Hatchling/Val'Kyr killed a stunned target. I have not used this team in quite a while! I used it at the start of 5.4. The bug might have been fixed in a patch/hotfix.

It's unlikely this was fixed, afaik Haunt has not been changed since 5.4 launched.

There are many examples where a stun does not cause Haunt to bug, so that seems less likely than the Feign Death possibility. Pet swaps are handled in a peculiar way -- they create their own turn. If you choose to swap, the opponent gets an action during this 'swap turn'; if you're forced to swap due to active pet death, then there's basically an 'empty' turn where the combat log shows 'XYZ is now the active pet', and then neither pet gets an action. During this 'empty' turn, debuff durations do not tick down. This creates bizarre issues with things like Minefield--if an UD pet swaps into a Minefield and dies, it gets 3 UD rez turns instead of 2, probably because the UD buff can't tick down during the swap turn.

So what might be happening is that on the turn after you Feign Death, the Haunt debuff can't tick down. When the Haunt debuff expires, that should trigger the Valk rez. But if the normal expiration gets skipped by the swap turn, then maybe the trigger can't go off.

If anybody happens to run into this bug (Haunt fails around the same time as Feign Death or Fade), it would be very helpful to look at the combat log, and note exactly what turn 'XYZ becomes the active pet' happens, and what turn Haunt actually expires.

[Awq]

It vaguely remember that it also happened for Val'Kyr + Tiny Twister combination (Wild Winds, Bash, Cyclone).

This one could be explained by Cyclone killing the Haunted pet before Haunt could properly expire. It seems that end-of-turn effects that go off before an expiring Haunt can sometimes cause the pet to die without Haunt properly rezzing the Valk (posts are on page 1, toward the bottom).

Yes. Cyclone caused the bug! I am certain of this. It might have been the combination of a stun + multi attack as well as cyclone, but I have not played this team enough to give additional information. I might run a Val'Kyr + Tiny Twister (Wild Winds, Bash, Call Lightning) to see if the bug still happens. Testing Haunt with the Basilisk will be my priority.

It's unlikely this was fixed, afaik Haunt has not been changed since 5.4 launched.

There are many examples where a stun does not cause Haunt to bug, so that seems less likely than the Feign Death possibility. Pet swaps are handled in a peculiar way -- they create their own turn. If you choose to swap, the opponent gets an action during this 'swap turn'; if you're forced to swap due to active pet death, then there's basically an 'empty' turn where the combat log shows 'XYZ is now the active pet', and then neither pet gets an action. During this 'empty' turn, debuff durations do not tick down. This creates bizarre issues with things like Minefield--if an UD pet swaps into a Minefield and dies, it gets 3 UD rez turns instead of 2, probably because the UD buff can't tick down during the swap turn.

So what might be happening is that on the turn after you Feign Death, the Haunt debuff can't tick down. When the Haunt debuff expires, that should trigger the Valk rez. But if the normal expiration gets skipped by the swap turn, then maybe the trigger can't go off.

If anybody happens to run into this bug (Haunt fails around the same time as Feign Death or Fade), it would be very helpful to look at the combat log, and note exactly what turn 'XYZ becomes the active pet' happens, and what turn Haunt actually expires.

I will test this team again! I will first only use Feign Death in different ways, without a stun. I will use it during different rounds to see the results. I will test it with different values of health as well.
After that I will test stuns in different scenarios.
I hope to be able to create a controlled setting where I can replicate the bug.

I personally feel that it probably was the stun or possibly even Trash in combination with a stun. I can not remember it correctly, but I can remember that the Val'Kyr would 'die' even when I did not use Feign Death at all.
If I used Feign Death when Haunt had 1 round left, it would swap in the Val'Kyr. It might bug with 2 or more rounds left, but this does not match my experiences at all (I might remember it incorrectly).
It might be Feign Death, but I have a feeling that if it is Feign Death that it will be quite complicated. I guess that the bug gets influenced by the health of both team mates if Feign Death bugs it.

Please ignore this post for now, until I can replicate the scenario. I will not like having to play my team incorrectly, ah well!

[Maor]

Consume Corpse seems to be able to miss despite 100% accuracy.

Debuffs were an enemy Cyclone and the DoT from Make it Rain, the weather was Cleansing Rain. Nothing that should affect hit chance.

The target, Spawn of G'nathus, had died to Cyclone damage on the backline at the end of the precious turn, which I imagine is probably what caused it to fail.

[Poofah]

If I used Feign Death when Haunt had 1 round left, it would swap in the Val'Kyr. It might bug with 2 or more rounds left, but this does not match my experiences at all (I might remember it incorrectly).

That makes a lot of sense. If you Feign with 1 round left, then at the end of the Feign Death turn Haunt will expire and the Valk will rez. And then on the next turn (which is the 'swap' turn), your Valk will be alive and able to swap in (if it's the highest health backrow pet).

If Feign Death is the issue, I would expect the bug to happen when you Feign while Haunt has 2 rounds left. Then at the end of your Feign turn it will tick down to 1. And so then it will be set to expire at the end of the 'swap' turn. But debuffs can't expire on the swap turn. And possibly this messes up the rez portion of Haunt(?), but this is just a guess.

If the stun is causing it, that seems like it should be easier to replicate (hopefully).

Consume Corpse seems to be able to miss despite 100% accuracy.

It can. It seems to be coded so that if you have a valid corpse, then it's 100% accuracy, and if there's no valid corpse, then it's 0% accuracy (and it displays 'miss' when you use it). Once a corpse is 'used', the game marks it as invalid somehow. In this case, I suspect that the game didn't recognize your Spawn of Gnathus as a valid corpse, but I don't know what caused that to happen (maybe dying on the backrow had something to do with it?).

Originally you could miss Wish (for example), but I haven't heard any reports of heals missing since the big accuracy changes in 5.3. So my guess is this is specific to Consume Corpse.

[3wd]

There are multiple reports indicated that Demolish had been nerfed to the bottom, its hidden 65%-70% hit chance definitely got changed and base on a still pretty small sample, it's believe it could even dip under 50% hit chance.

This skill now probably changed from one of the most overpowered skills to one of the worst.

[Awq]

Update:
First I've tested Val'Kyr + Scalded Basilisk Hatchling on wild aquatics (and some other pets, to see if the Aquatic racial did not influence the outcome). I have used Feign Death on every turn, with different health values on my Val'Kyr + other pet. It did not bug. I would generally pass with my Basilisk Hatchling, which might have influenced the results. Trash can be too unreliable to create a safe setting for testing (I was lazy, and did not want to repeat fight). It created a safe setting to rule out Feign Death as the only culprit.

I did take note that I might have to recreate the same setting, to ensure that Haunt would kill the enemy pet. I however felt like I was randomly guessing, and felt that I would need more information before I could even guess appropriate tests.
I wanted to test this combination in PvP, to give me knowledge of the bug. I could only vaguely remember when it happened, so I wanted to see experience a wide range of situations to see when it would bug.
I've done around 20 matches, it bugged 3 times so far. It has bugged once while I did not use Feign Death. It is fair to say that Feign Death is not the ability that bugs it, or that if it is, that there are multiple abilities that bug Haunt.

I am inclined to say that it could be random, but other Val 'Kyr teams would then encounter the Haunt bug. If I have to guess, Trash is the ability that bugs it. I will test it further. I now have a framework to work from.

[Kring]

[ability]Ice Lance[/ability] under Blizzard hits its additional damage even when blinded from an adder. I did confirm this yesterday (2014-03-12) with both a [pet]Tundra Penguin[/pet] and a [pet]Rotten Little Helper[/pet].

[Jerebear]

Any thoughts on updating the OP since some of these may have been addressed?

[Poofah]

[ability]Ice Lance[/ability] under Blizzard hits its additional damage even when blinded from an adder. I did confirm this yesterday (2014-03-12) with both a [pet]Tundra Penguin[/pet] and a [pet]Rotten Little Helper[/pet].

This is bug #7 on the list. Many multi-hit abilities behave this way: Moth Balls, Blessed Hammer, Ice Lance (under Blizzard), Puncture Wound (when poisoned). I don't have a comprehensive list, and it's not consistent (ie, the 2nd hit of Alpha Strike/Tail Sweep does not hit when Blinded, for example).

Any thoughts on updating the OP since some of these may have been addressed?

I edited to reflect the Demolish and Elemental/Sandstorm changes. The Elemental/Sandstorm issue seems pretty much fixed, as far as I can tell. Demolish is clearly hitting less often, but I have no idea if it's actually 50%, and I have no idea what change they actually made to the 'miss streak protection' that apparently was causing it to hit too often. To my knowledge, none of the others has been addressed.

[Kring]

[ability]Ice Lance[/ability] under Blizzard hits its additional damage even when blinded from an adder. I did confirm this yesterday (2014-03-12) with both a [pet]Tundra Penguin[/pet] and a [pet]Rotten Little Helper[/pet].

This is bug #7 on the list. Many multi-hit abilities behave this way: Moth Balls, Blessed Hammer, Ice Lance (under Blizzard), Puncture Wound (when poisoned). I don't have a comprehensive list, and it's not consistent (ie, the 2nd hit of Alpha Strike/Tail Sweep does not hit when Blinded, for example).

I know, but Blizzard [url=https://twitter.com/TheCrafticus/status/436699739464007680]claimed that they fixed the blind bug in 5.4.7[/url] and I wanted to confirm that this bug still exists at the moment.

[Poofah]

That's actually referencing a different bug, it's #6 on the OP list, affectionately called the 'negative accuracy' bug. I listed them separately because I'm fairly sure that the root cause is different. In the case of certain multi-hit abilities, the 2nd/3rd/4th hits just don't check for accuracy buffs/debuffs. But the initial hit does -- the initial hit of Ice Lance will never hit if the caster is Blinding Poisoned, for example, but the 2nd always will (under Blizzard). This is different from sub-100% accuracy moves such as Demolish, where the initial hit can occasionally hit through Blinding Poison.

Bug #6 might be related to the same 'miss streak protection' that caused bug #1 (Demolish 75% accuracy). Ie there might be a hidden additive buff to accuracy after a miss, and this might push the accuracy above 0% when Blinded. I don't have enough experience with Blinding Poison in 5.4.7 to say whether the negative accuracy bug is fixed. Bug #7 almost certainly is not related to this -- it's probably just a peculiarity of how the multi-hit abilities are coded. It is definitely not fixed, in any case.