Shadow Priest Trinket Discussion -> Emerald Nightmare

[bruissen]

FWIW. Got the HC swarming plaguehide, and for ursoc it had a 67.88% uptime and ~8.05% of my damage.

[Djriff]

FWIW. Got the HC swarming plaguehide, and for ursoc it had a 67.88% uptime and ~8.05% of my damage.

Logs for that would be great if you would.

The problem with some trinkets right now is two-fold.

At some point for twisting wind specifically the proc rate was changed from 1.5 to something, without logs of it, it's hard to figure out what it is now as it's still datamining at 1.5 as the hotfixed data hasn't been pushed yet.

Additionally the same goes for the difference in tooltips for in game and on wowhead/wowdb/amr. The in game damage is much lower than any of those 3 sites is showing, however without knowing if the proc rate was nerfed or buffed it's hard to make a determination on trinkets.

I know @Swol is pouring over logs as am I to try and figure out what is what, it's just taking some time since it's a lot of stuff to go through. So until we know more, i'd avoid Twisting Winds, or if you get it, do a bunch of bosses with it, target dummies for 20m so we can get proc rates.

For target dummies in particular i know this will sound weird, but don't use Voidform, and take off all of your gear other than the trinket for 5m. Then do the exact same test with all your gear and post logs + haste % here. It'll go a long way to helping determine proc rates for trinkets.

If anyone can help out, it'll speed things up immensely.

[bruissen]

Logs for that would be great if you would.
.

Won't let me post links untill I've posted enough >.< I'll pm you on the priest discord if you're on it

[Djriff]

Won't let me post links untill I've posted enough >.< I'll pm you on the priest discord if you're on it

I am same name as here.

[Reposed]

I have an Unstable Horrorslime (850ilvl) and the proc tends to be about 4.5% of my overall damage.

[GT4]

In game tooltips are wrong, this is a known issue.

Also I can't look at anything right now as I'm at work.


That's why I said sims<reality. Once we have a decent amount of live logs with these trinkets we can evaluate them further. That's how theorycrafting works.

There's a difference between sims based on speculation and sims based on actual testing/knowledge (the latter being the standard). You're not communicating that difference when suggesting "X is best according to sims".

As for the tornado trinket: I get 3 different numbers for the same ilevel version (in-game tooltip, wowhead, AMR tooltip). In addition to that, from what I've seen (which is admittedly not a lot), the proc might not even reliably hit smaller non-moving single targets of which there are quite a few important ones in the current raid tier. As long as neither are properly addressed (I asked the AMR guy for a source of the trinket damage number but got no reply), you can't just go around and say it's BiS according to sims without giving that information as well.

This lack of proper communication is why so many people are now of the opinion that "sims don't matter" (in general) when in reality they absolutely matter if you take into account context and don't trust them blindly.

[Djriff]

There's a difference between sims based on speculation and sims based on actual testing/knowledge (the latter being the standard). You're not communicating that difference when suggesting "X is best according to sims".

AMR sims are not based on speculation. The data they are using is directly from the game that's being data mined, the game is giving us bad data, not AMR's fault. Now that it's been identified, it's being fixed. It's also known the ingame tooltips for a lot of things are wrong.

As for the tornado trinket: I get 3 different numbers for the same ilevel version (in-game tooltip, wowhead, AMR tooltip). In addition to that, from what I've seen (which is admittedly not a lot), the proc might not even reliably hit smaller non-moving single targets of which there are quite a few important ones in the current raid tier. As long as neither are properly addressed (I asked the AMR guy for a source of the trinket damage number but got no reply), you can't just go around and say it's BiS according to sims without giving that information as well.

Even if the trinket has 50% uptime, it's crazy good according to the numbers AMR originally had from the data-mined stuff. With in game numbers? Not so much.
Additionally the proc rate was changed via hotfix but is not reflecting in the data-mined stuff for some reason. That's why i asked for live logs so i can see how close they are and what's changed.

This lack of proper communication is why so many people are now of the opinion that "sims don't matter" (in general) when in reality they absolutely matter if you take into account context and don't trust them blindly.

What lack of proper communication? People that use sims for TC'ing know not to trust them blindly.

[GT4]

AMR sims are not based on speculation. The data they are using is directly from the game that's being data mined, the game is giving us bad data, not AMR's fault. Now that it's been identified, it's being fixed. It's also known the ingame tooltips for a lot of things are wrong.


Even if the trinket has 50% uptime, it's crazy good according to the numbers AMR originally had from the data-mined stuff. With in game numbers? Not so much.
Additionally the proc rate was changed via hotfix but is not reflecting in the data-mined stuff for some reason. That's why i asked for live logs so i can see how close they are and what's changed.


What lack of proper communication? People that use sims for TC'ing know not to trust them blindly.

Data mining has never been a reliable source of data for item procs. You shouldn't treat it as one.

[Djriff]

Data mining has never been a reliable source of data for item procs. You shouldn't treat it as one.

I can honestly say i didn't know that.

[Deleted]

Data mining has never been a reliable source of data for item procs. You shouldn't treat it as one.

True, but a lot of the time it's all we have to go on until live logs become abundant. Like I said before, going into a raid entirely blind is worse than having a rough guess. Data mining is just that. Also, huge errors with data mining like twisting, etc, are the exceptions not the usual. Most other trinkets are behaving as expected.

[tiptopmemer]

I have an Unstable Horrorslime (850ilvl) and the proc tends to be about 4.5% of my overall damage.

Yeah, its AoE based, it'll be quite powerful for mythic+
@everyone
PPM is a dumb system its not reliable and not self correcting, many PPM trinkets can proc back to back, ive seen 2 hits in a row (consistently) from the likes of Caged Horror, Voidsight, Oakheart's and Devilsaurs.

Sims, nor WoWhead data is reliable, do your own testing with logs.

[GT4]

True, but a lot of the time it's all we have to go on until live logs become abundant. Like I said before, going into a raid entirely blind is worse than having a rough guess. Data mining is just that. Also, huge errors with data mining like twisting, etc, are the exceptions not the usual. Most other trinkets are behaving as expected.

If there's huge discrepancies between tooltips (e.g. in-game vs. wowhead vs. datamined) or the mechanics simply aren't clear (xavius trinket) and some trinkets seem way stronger than others in situations they shouldn't be in (AoE proc supposedly doing 10k more single target DPS than best single target proc), they should absolutely be expected not to behave "as expected".

If you're in a position to release (partial) "BiS lists" which I've seen around here with those trinkets being BiS even though they can expected not to be simulated properly and you don't at least put that as a disclaimer (and I've seen this disclaimer missing in multiple cases now), you're in my opinion doing the community a disservice altogether. If you tell your average player "x trinket sims the best", you can't expect them to do the research on how reliable that data is. At that point, they could just as well sim it themselves.

[Ilir]

the trinkets are a mess.

I'm using devilsaur (or dinosaur electrocution thingy) and it is always doing ~6% of my damage, 835 version

[Deleted]

If there's huge discrepancies between tooltips (e.g. in-game vs. wowhead vs. datamined) or the mechanics simply aren't clear (xavius trinket) and some trinkets seem way stronger than others in situations they shouldn't be in (AoE proc supposedly doing 10k more single target DPS than best single target proc), they should absolutely be expected not to behave "as expected".

If you're in a position to release (partial) "BiS lists" which I've seen around here with those trinkets being BiS even though they can expected not to be simulated properly and you don't at least put that as a disclaimer (and I've seen this disclaimer missing in multiple cases now), you're in my opinion doing the community a disservice altogether. If you tell your average player "x trinket sims the best", you can't expect them to do the research on how reliable that data is. At that point, they could just as well sim it themselves.

Ok, so what is the alternative? To pass on every trinket that drops? We've now cleared normal and heroic this week and several trinkets dropped, but live logs are still scarce to say the least. I'm very much interested to know what you would do in this situation that exists right now.

[Djriff]

the trinkets are a mess.

I'm using devilsaur (or dinosaur electrocution thingy) and it is always doing ~6% of my damage, 835 version

Devilsaur is doing ridiculous damage from what I've seen.

Ok, so what is the alternative? To pass on every trinket that drops? We've now cleared normal and heroic this week and several trinkets dropped, but live logs are still scarce to say the least. I'm very much interested to know what you would do in this situation that exists right now.

No, but if you have a Darkmoon Deck: Hellfire and Chrono Shard you won't be replacing it until Mythic EN anyways. Good news is that we'll have some logs with these new trinkets and be able to see how they are stacking against each other. Obviously twisting wind is an anomaly and until I see people using it I won't be able to make a decision on it.

Live data = refined sims = more accurate results.

For example @GT4 linked logs with their 120s+ StM and looking at their rotation @Swol was able to improve the rotation of the simulator.

[Zoopercat]

@Djriff @GT4 Yeah, datamined data for proc items can be iffy. We actually check proc-items in game as much as possible (by testing them from flaskataur on the beta to finding ppl online with the item). We don't blindly trust mined data. That doesn't mean an item might not need updating, but we try to leave comments on wiki for the items that need final info and testing.

[Fleckens]

While not a emerald nightmare trinket, any idea how good the infernal writ(http://www.wowhead.com/item=137485/i...254&bonus=1826) from court of stars is? got an 850 of it and not sure if it's better or worse than naraxus spiked toungue at 840.

[Sxq]

So I got the Xavius trinket, Bough of Corruption and did a bit of testing

- it's very similar to Prophecy of Fear in that it only procs on direct damage, meaning that it does not proc on mind flay or mind sear or dot ticks or void torrent
- it does seem to proc on shadowy apparitions
- it does proc on void eruption and shadow crash for aoe
- there's a lot of synergy with mind spike because it procs on both the initial spike and the aoe detonation

- proc rate is kinda low overall
- initial proc gives 1 stack to target, every 2 seconds it will either add 1 stack to a nearby target or if there are no other targets it will add the stack onto the initial target
- in pure single target it will stack up to 10 (20/2=10)
- i've found that once your target is above ~~5 stacks, it's actually better to spam the initial tick of shadow wordain as a filler instead of mind flaying
- damage varies pretty wildly due to low proc rate and the fact that if it procs on a low hp add it can be completely wasted
- procing late in void form or stm can be very strong because of high haste and void bolt spam

Not sure how good it is for shadow yet. I'll try to get some raid logs tonight with it if people are interested. Wouldn't be surprised if it's once again an arcane mage trinket because I'm assuming they can cheese it like the did with PoF in WoD.

[killertree]

860 Swarming Plaguehive did 6.4% of my damage on the first heroic boss.

i.imgur.com/v4FWpr4.jpg

I've seen it as high as 8%

[Deleted]

I tested Twisting Wind 850 the last 4 days and it seems really bad! in Mythics it was like 1-2% in the raid the best i've seen was 3%