My Thoughts on some Coherent Changes to Pariah

Amranu

This isn't a rant, but rather a review of some of the problems of the kit of Pariah that we have been dancing around in order to find a good place for it (which have been nicely highlighted by yesterday's change to Virulence).

I plan to lay out what I think the current problems are with Pariah's kit now that this change has been implemented and discuss some of the problems that these changes are trying to solve; I'll also discuss some changes to the core mechanics of the class that will hopefully address criticisms of the class both from people that use it and those that have to fight it. Hopefully this leads to some nice discussion and we get the class fixed up, sooner rather than later.

Let's first take a look at yesterday's change and the problem it was attempting to address:

* VIRULENCE has been reworked somewhat. This was always intended as the lock threat but its proving too punishing a fork as it stands. It now requires the target be exposed to be used, and always requires and consumes balance/equilibrium (basically its a standalone cast now rather than something you can combo with a plague/trace if you hold your expose in reserve).

The main problem with virulence in its previous form was that it was available to be used too quickly for too large a gain. In particular, because of the ability to hold expose one could sting/trace/virulence in one balance at 3 plagues. Virulence was intended to be used at 4 plagues to replace sting and complete lock while tracing a memorium affliction. However, with expose in its current form, even if virulence was made to require balance one could still sting/virulence to lock at 3 plagues.

General Problems: Now, I've at this point ranted to half of Achaea about why I think this change on its own is terrible, but I think this nicely sums up the main issue this change has highlighted: because latency only requires 3 plagues, virulence will never be viable compared to latency if it cannot be paired with sting, and a lack of virulence drastically hurts the meaningful choices of strategy available to the Pariah. Latency is rather simplistic on its own, and it was the combination of scourge + virulence that made plague and memorium aff choices meaningful, and with virulence in its current state these choices matter significantly less because realistically you are always pushing for latency over anything else.

Solution part 1: One thing we can infer from the problem above is that in order to allow meaningful choices for Pariahs again we need to make sure virulence is a viable alternative to latency. Since not allowing sting with virulence is necessary to avoid virulence being too strong, we need to nerf latency from its current potential to allow virulence viability again. Solution: increase latency requirements to 4 plagues. This will bring virulence back into viability by having them compete in a similar way: If you can lock, you can use virulence, otherwise latency and hope for the best.

Problems with Virulence currently: Virulence in its current state is just bad, while the above change is a step to it becoming viable, it is still not very useful in its current state as it cannot be paired with a trace. This means that for instance that the use of peace, or single afflictions such as dizziness from virulence is impossible. The class should be modified to allow virulence to be used in this capacity again. However, with expose in its current state this can't be done.

Solution part 2: If we modify expose so that it can no longer be held from one trace to the next, sting will no longer be able to be paired with virulence and we will once again require 4 plagues to be stuck for the large affliction burst even if we once again let it be used balancelessly. However, the lack of ability to hold expose has some obvious problems: both latency and scourge in its current form uses held expose for the kill. Specifically, scorpion/scourge or scorpion/latency.

I suggest we make expose last for a very small amount of time, say 0.3 seconds. This will allow one to in line a higher sorcery with a trace but not allow it to be held. After this is done, the following changes should be made:

1. Virulence should be made once again available without balance with expose.
2. Trigger expose for an epitaph started with scorpion to account for latency/scourge.

Problems with shield and shield spam: The kit in its previous form was too quick, and there wasn't enough time to cure defensively given Pariah's speed. With the above changes made, affliction priorities will make significant difference due to the need to stick 4 plagues for either virulence or latency. This will mean that there is room to weaken shield as a defensive measure, given that the time to kill for Pariah will have been increased significantly and curing priorities will be made more relevant in defence.

Solution part 3: In order to account for this, allow fissure once again to trigger expose, so long as the epitaph chain is of length 2-3 or more (this used to be 1 or more, however this causes a problem during latency whereby the use of shield became irrelevant. Only making it generate weakness with a chain at least 2-3 long should prevent this).

Room for tweaks: After these changes have been made, the Pariah's time to kill will have slowed down significantly and the use of shield as a defensive measure will also have been toned down. Due to the length of the time to kill being longer, tweak ensorcell's balance time, length, and potentially whether or not ensorcell stops an epitaph chain to effectively tune the how strong the class with the above changes is. It may need to be buffed or nerfed depending on how strong allowing expose with fissure is, but more effective counterplay will now exist against virulence and latency with curing prios, and so I believe that buff to fissure will be viable and should help resolve some of the problems Pariah have had with it while not making it so their opponents don't have defensive measures.

Anyway, those are my thoughts on some changes that should hopefully resolve a lot of the problems people have with the class, both fighting it and using it. Cheers.

Gallida

Just refund every person with Pariah all of the lessons, arties, and crypt rooms in full.

Maybe you can rerelease the class when it's finished.

Gallida

Alternatively, you can take the same "delivering more than two afflictions in a single balance is baaaaaad" to its logical conclusion and gut the other abilities in other classes that can do the same.

Occultist Firelord, Alchemist Inundate are the main offenders. Make them require balance/EQ/tertiary balance to perform and forbid other afflicting abilities.

Iaxus

I think Amranu's on the right track, especially now that he's had time to discuss it and form a coherent plan.

It's unfortunate Pariah was released in the state it was in, and I do think it could've gone under more testing, but once the cat is out of the bag there's no putting it back in. To that effect, Pariah needs/needed nerfs to get it under control, and even in it's current state Latency is still a very strong path (though I do acknowledge that it makes the class one dimensional).

My biggest concern with these changes is doing them all at once. I think the best approach to this would be to implement some of the above in small batches and then watch how the class plays out for a time, and adjust as needed. For instance, nerf Latency to bring it up to 4 plagues (in line with Virulence) and watch for a time to see if it's still feasible, then decide from there if it needs further nerfs, or if the buffs should start from there.

Also to preface any attacks on me, I'm strongly in the corner of "make Pariah fun", but that means it has to be fun to fight against as well. The oppressiveness from fighting the class wasn't fun before, and now it isn't fun to play the class. There needs to be a middle ground.

Amranu

My biggest concern with these changes is doing them all at once. I think the best approach to this would be to implement some of the above in small batches and then watch how the class plays out for a time, and adjust as needed. For instance, nerf Latency to bring it up to 4 plagues (in line with Virulence) and watch for a time to see if it's still feasible, then decide from there if it needs further nerfs, or if the buffs should start from there.

I very strongly disagree with this. Each of these changes are required as a direct consequence of the other changes. For instance, you cannot move expose without adding it to scorpion, you cannot nerf latency without buffing fissure and adjusting ensorcell. And you cannot keep virulence in the state it is in without forcing latency to be the only viable kill path, which requires changing expose.

The class is hard to balance because each piece of the kit influences other parts. Ramifications of changes to Pariah need to be taken account and adjusted for at the same time. The virulence change applied yesterday was bad because the consequences were not considered beforehand and other changes were not made to the kit to account for the problems.

Gallida

I don't think Pariah has ever needed the nerfs after the initial Extract bleeding reduction, everything afterwards has been a kneejerk reaction based off not understanding the class or how to defend against it.

Alchemist can deliver six afflictions at once with inundate/truewrack, all but one of which are used in an lock. They can do this while being able to throw paralysis on every balance to prevent you from curing humours, and has room hinder to keep you in the room.

What makes this ability somehow balanced when sting/trace/virulence isn't?

Iaxus

Doing it slowly over time allows for seeing each change so that the class doesn't just come back as a completely unstoppable powerhouse, though. I do think you thought this through a ton, but there are bound to be things you didn't foresee that arise with each of these changes, and doing them all at once makes it much harder to pinpoint and fix anything that was overtuned.

Amranu

The only thing that can realistically be left out are changes to ensorcell, hence when I said it should be tuned after the fact. The rest of the changes is returning to release Pariah but slowing it down significantly rather than making shield give complete immunity to it, with minor nerfs to fissure rather than major ones. Could be tested on the test server first, of course.

Mizik

How well would the class function without paralysis? My biggest problem with it is it is zero fun to fight. Too punishing if you don't do anything but shield. The Pariah can't be hindered while the target just watches him fucking go from behind shield or lose.

Reminds me of Magi in retardation.

Amranu

It needs paralysis to buffer for pyramides or other plagues, otherwise it would be virtually impossible for it to get to 4 plagues.

Gallida

We've had two separate flavors of "Crazy aff output but no paralysis" in Occie 2.0 and Air Lord, and those aren't exactly paragons of design.

What about allowing for multiple burrowed swarms in a target and changing Virulence to be a function of burrowed swarms rather than active plague afflictions, where Virulence removes the burrowed swarms and adds the Virulence afflictions rather than the plagues?

I think this whole endeavor is pointless if we can't be told why Pariah needs to be consistently kneecapped while other classes get uncurable and unescapable damage and momentum.

Atalkez

I’m not sure why you’re comparing an aff bomb from a slow TTK class against one that until recently had the highest or close to while also being the hardest to mitigate. You’re claiming people don’t understand the class while making an absolutely off base comparison.

The issue that pariah has had from the moment it was released is that you just never can touch it. You’re on defensive from the third round of the fight, and you can’t cure you way through it nor can you hinder it in any capacity. The only thing that works is stupidity and that’s a crapshoot on if they hit secrets or lose an aff. If they don’t you lose. That’s not ideal either. It was shredding other momentum classes with almost no effort for the most part.

I’ve suggested changes to other parts of the kit that have still been left alone, so this wasn’t the direction I thought needed to be taken. I see the issues that are mentioned and recognize them. That said, the thought process that because you could shield four times and be okay only to be in danger if you dared to attack more than twice is not realistic and fixing that is a good thing.

Gallida

I'm using the alchemist comparison for two reasons:

1. You cannot outcure an alchemist, Formulation or Sublimation be damned. Your options are to be faster or leave the room, neither of which are sure things given alchemist's hinder and incredible defensive potential. If the issue with Pariah is uncurable and unhinderable affliction momentum, Alchemist has that already because single wracks ignore clumsiness.

2. Alchemist has an affliction bomb one of their finishers. If the issue with Pariah is "Virulence drops too many afflictions in one balance, nerf pls", Inundate does the same thing, except that it also comes with a complete class that continues to dodge the nerf bat.

There are other classes that do what Pariah does already (crazy momentum, room hinder, forks requiring priority swaps), and none of them are seen as a problem. If "not fun to play against" is a requirement, that's opening a huge can of worms.

The biggest sin Pariah committed as a class is the ability to kill top combatants with no monetary investment, which is only allowed for factional classes.

Atalkez

You can mitigate alchemist just fine, with a few minor changes to how you approach the fight. It takes far longer to get a phlegmatic bomb than virulence. It took 3 affs to get virulence, it takes a minimum of 8 to get a full phlegmatic off. That’s a huge difference.

3 is a very jaded view. Every class that’s been released in the last four years fit that mold. Elementals. DW, Psion all can without a dime spent on the game. Pariah was far too fast and with very little actual counter play and that is why it’s consistently been addressed. Nothing in the game worked like Pariah up until this point, with only Occultist being a direct threat momentum va momentum with no breaks in the fight - and nobody is saying that Occultist doesn’t need toned back.

Gallida

tl;dr for archaeon: "let's lower the entry for combat and make it more accessible. shit, whales are dying. no, not like this."

DW: 350 credit tax in the form of the dagger. It's less egregious than a diadem for eq-based classes, but it's still a tax.

Elemental Lords: Earth is the only one that I'd say is a playable class. Air is an effective gimmick that falls off a cliff, Water is a coinflip, and Fire only kills people who get overconfident.

Psion: Blast is the only feasible kill route because the others rely on your opponent curing suboptimally. Blast also starts to fall short when your opponent is tanky enough to fully prioritize sipping mana.

It's funny that you bring up Occultist, because Occultist needs a literally discontinued artie which will run you more than a L3 weapon for any class, just to counter the spaghetti code that causes primebonded ents to take themselves for walkies. Oh, and a diadem. On the other hand, it's the flagship class for a faction, so it should be slightly overbearing to encourage people to actually join the faction.

The only universal three skill classes that work out of the box without a credit investment are Shaman and Jester, because those are gated behind scripting and combat knowledge rather than a wallet.

You lose to those? They earned it, it took time and effort to work all the moving pieces together. Everyone Amranu kills with Pariah, he could kill as Shaman and nobody would have blinked.

The fact that Pariah offers a way for a faction-neutral class to kill the entire game with a copy of AK and two trans skills is untenable. It's an outlier that needs to be punished.

Saonji

"Damn this class is ridiculous; really needs toning back."

Plays Priest.

Shecks

Nothing new from me Amranu, just going to echo what I've said to you before, here:

I think the change to Fissure was dealing with a very real problem but wasnt met with the best possible solution. The problem is just that Pariah just gives too many affs all at once to the point where prios and swaps don't really matter and the only counter is to turtle, often until Encorcell drops so you have a chance to heal up out of room before coming back to fight for another 5-10 seconds before you're in kill territory again (and shield spam again).

The solution to this should have been, IMO, to reduce the stacking delayed affs and perhaps reduce the speed of the class in general such that shield isn't absolutely necessary, instead of just making shield a minor reset for the Pariah. Another alternative would be to make clumsiness affect Pariah so that momentum classes have a fighting chance (noting the Pariah gets to slow THEM down with clumsiness, in almost all cases).

I also think slowing the class down gives room for other buffs to the class that currently can never be considered.

It should be mentioned that Lethargy was basically the only aff that actually affects Pariah (and Occie / Apostate), and it just got significantly nerfed, which is ostensibly a buff to these classes and a major nerf to classses that rely on Lethargy to counter ultra fast aff stacking. 2H is a great example as it is kindof a joke to take 2H into a fight with Pariah now as you can't slow prep and your momentum takes at least twice as long as Pariah needs to kill you. Lethargy wasn't enough to fix this problem but at least it helped - and now that's been axed. I think this applies to other classes vs Pariah too.

As for the Virulence change, you know I don't sympathize because as far as I can tell it was a response to full AI automated locking which lead to results probably way beyond what was intended. If it were me I wouldn't have done this virulence change, I would remove all the 3p cure lines*for plagues so that auto-Pariahs don't have virtually perfect aff-tracking at all times so at least there's some RNG / risk to the ultra fast lock path (noting Virulence has a cooldown). This would nerf auto-Virulence without hitting manual players as hard and without actually nerfing Virulence itself. (I remember seeing these 3p lines on day 1 of release and asking "wtf why would they do this this is so OP for scripters?", and I still feel exactly the same way now.)

* For those of you who don't know, most of the plagues have 3p lines when they're are cured which allow the Pariah to not only know what you cured but also what you DIDN'T cure, making it the only aff class in the game that doesn't have to deal with uncertainty going into kill executions. They removed Expert Diagnose from the game (pretty much) for this reason then gave back a much better version of it exclusively to Pariah for essentially no reason.

Saonji

I think this thread fucking sucks because it's making me agree with Shecks.

Also we can't forget that Pariah still often has periods where it hits twice- with double affs- between herb eats. When's that getting fixed, btw?

Gallida

Removing the third-party messages would increase the scripting complexity and reduce the reliability--Amranu remains unaffected and the other two people using the class get to suffer. It's the perfect suggestion if the problem with the class is that it kills good people.

Shecks

@Gallida as someone who has written a top tier aff tracker including for Pariah and probably 40-50 iterations of offense automation on around 15 classes, this is simply not true. I'm not going to break down the reasons for it but I'm pretty sure Amranu would agree too, as another person who sees Achaea as a personal programming challenge.

Amranu's tracker doesn't use probability btw, it uses either binary or tri-state on each aff, which would make perfect automation under my suggestion essentially impossible for him without rewriting his entire tracker from the ground up. You need probability-based tracking, perfect retroactive math to calculate those probabilities, and also an offense system that leverages the probabilites perfectly just to match what human intuition does without a shred of code. Afaik I'm the only person who does this (emphasis on the afaik as there may very well be others) - but after I fully automated Pariah and ran it through sims to see what it could do, my reaction was to delete the class and take Paladin instead. Removing the 3p lines would make it far more difficult even with my tracker and offense system.

The only viable way to stop automation is to make aff tracking insanely difficult to do accurately - which is again exactly why Expert Diagnose was axed with extreme prejudice. The exact same logic applies here. Should note I've spent enough time on my aff tracker to bash to dragon like 10-20 times, and ofc you have to deeply learn a class before you can automate it, just to do what a skilled manual player could do just by practicing. That's not OP until you distribute it to other people (something I haven't done since 2015).

Saonji

@Shecks FYI Amranu's does use probability, since he uses AK (albeit a very heavily modified one). Every single person who uses AK (which is quite a lot of people now) does probability-based tracking. Since that's... Well... What AK does.

Granted, a lot of people might as well be using have/have not, given how many people just do [ if score == 100 then ]. But yeh.

Gallida

Even if you take out the third-party messages, at least three out of the five plagues have visible symptoms (random epilepsy procs from mycalium, bleeding on ginseng eat from flushings, not focusing for sandfever, not sure if rebbies has one offhand), and the one that doesn't is hardcoded to be cured after paralysis--and you're not applying slickness without virulence so that's not a concern. If your tracking can determine the probability of asthma, you can determine the probability of plague affs. Harder but not insurmountable.

The three people still trying to use Pariah might switch to Scourge as a route because it's easier to track or more reliable in the face of imperfect tracking, but that's already the case for, say, Depthswalker mutilate vs dictate/lock. You're just raising the scripting burden and thus the skill level necessary for the class--reiterating my overall point that the class appears to need nerfs because anyone can use it to kill good people.

Saonji

Rebbies does have one when it fires iirc (could be misremembering). But I mean... You're never really going to see it fire, since it fires on salve applies. And you're only doing anorexia with virulence, so you'll see whether or not they have it anyway when the virulence line fires.

Amranu

I don't think the Alchemist argument is particularly amazing. Pariah could pull off virulence in 7 attacks, Alchemist can pull off the aff bomb in 9, but Alchemist is also hindering every attack, and can lower the number of humours needed by sticking asthma or impatience. Importantly, if the Alchemist doesn't need weariness, ttk with phlegmatic is fairly comparable with virulence prior to the change, with the caveat that defensive prios are more straightforward against Alchemist than Pariah.

Regardless, virulence is now significantly less reliable and in nearly all cases now slower than phlegmatic while also not providing hinder every attack.

In regard to the 3rd person lines for plagues, I could maybe see an argument for removing flushings line, but the rebbies line is fairly necessary - eating a single kelp and not knowing which was cured means you have to repeat 4 attacks to stick all of them again, since every kelp aff in the kit is necessary for lock against fitness. As stated below I think the increase to 4 plagues stuck already deals with this problem fairly well.

Amranu

Another thing I'd like to point out is I don't think anyone really appreciates how much more difficult it is to stick 4 plagues for one balance rather than 3. I'd be happy to show you all, sufficed to say with decent prios ttk for virulence right now is quite long, if not entirely avoidable with defensive prios sans shield

Gallida

Virulence is clearly intended to be the kill route for styling on people, like Psion Flurry, Waterlord Drown, or DWB Damnation. Every class needs at least one and Pariah having three viable routes is imbalanced.

Shecks

@Amranu I am not at all saying you're wrong here, on that last point. I would argue a different point altogether. Pariah shouldn't be entitled to a Virulence lock in all cases, but it should be able to achieve one of its killpaths. As it stands, you have to fork about halfway through stacking into either a virulence path or one of your other 3+ killpaths (There are dozens of routes but simplifying down to the actual killing attack there are at most 4 or 5 ways to kill). I don't think you should be arguing that Virulence should be achievably quickly, but instead you should be arguing that the threat of Virulence should make it possible to punish people who don't recognize what path you're on and punish their mistake appropriately. If they are prioing vs Virulence, virulence should be able to be stopped 100% of the time with good prio swaps and maybe a shield or two - but that should leave them open to a fork to another killpath within 5-10 seconds. As Gallida said, Virulence exists (as far as I can tell) as both a threat used to fork (like Darkshade) or a style kill, not a primary killpath that should always be possible, and definitely shouldn't be fast/easy.

I can't say definitively that this is currently working as I described above, but I do think that this is what you should be arguing for instead of, paraphrased, "I should be able to virulence lock anyone who doesn't spam shield oh and also please revert the change that allows me to maintain full momentum while stripping shield" (sorry, but that's basically what I interpret your argument as).

Amranu

Virulence needs to be viable for scourge to be viable, which is something I discussed previously (though I don't think it's been discussed here explicitly). I want the change reverted in the meantime so I can play the class I've put months of work into without becoming incredibly angry at the state of it.

Saonji

I like how Pariah has multiple (reliable) routes that all fork closely with one another, while also having one of the lowest "buy-ins" in the game. Hell they can't even use SoA like the other "low buy-in" classes.

Meanwhile Occultist only has one actually worthwhile, and remotely doable (versus a halfway decent person, with not stock curing), route. Even if it had more than 1, they don't currently interact with one another at all (while also requiring entirely different primebonds to even do properly anyway lol). Also a much higher buy-in. About the exact same average TTK as well.

Really highlights how great the game's balance is.

Amranu

Occie doesn't give nearly as much of a shit about shield though. Tradeoffs.

Saonji

Depends how early on you shield, but yeah. Also I've already argued Gremlin should be sped up slightly in line with other razes, but not scale with affs. Baseline raze speed is kinda slow, def gets a little too fast with affs though. (end result is having a raze that's roughly equivalent with other razes, if not marginally slower)

In any case, I'd take a class with actual options as opposed to being the textbook definition of one-trick pony (even if that 'trick' is good).