Crystalline Conflict Tier List (with deep maths!)

[Myrha_Lhlalheva]

I'm fully aware of the PLD burst combo. I addressed it in the first post, and regardless you're not doing the BRD burst correctly because you're not factoring in the delay on Apex Arrow and how it lines up with the rest of the burst. BRD is picked in LP precisely for it's 30k burst and for it's ability to put down Wardens.

Unfortunately none of us are Dr. Strange but it turns out you don't need a PhD to add up Bard burst and see it's better than PLD.

Sorry, I had refreshed to the bottom of the page and didn't see your 10 posts. You can type a shorter post and edit it in order to circumvent the character limit! I was also not responsible for crunching RPR numbers, but yes, the rubric was built to consider only the most ideal possible burst--although in the case of said burst being contingent on LB usage, we averaged a "with LB" and a "without LB" number.

Not sure what you mean by "not factoring the delay on Apex Arrow", since Apex isn't even part of the ~1GCD STB being calculated. If you mean that there should be another 8400 tacked on there for Blast Arrow, that would break our rules on how may GCD's we're allowing during the window we're considering. If we just gave everyone else another free GCD as well, then...there isn't really a point.

Trying to catch up and address your other posts:
WHM STB was calculated as 40k with LB, 20k without. Seraph Strike is included.

MNK burst with LB is 72k and does easily top out the scales--but we also averaged it with its "w/o LB" value, which put it at a borderline value on the rubric between 4 and 5 points. Due to the way the rest of the rubric was graded, borderline values are rated in the lower bracket.

RDM's low mobility rating is likely because we used the rule of "half value for mobility that was only valid on one type of target". It might seem like an unfair assessment, but I don't think you can argue that 2 enemy-only dashes and 2 enemy-only backflips is just as mobile as, say, Thunderclap.

DRG mobility rating is also a clerical error, it should be 4 and not 5. This is the result of us retooling our entire rubric 3 times and some of our re-scoring slipped through the cracks. I'll see if I can update the graphic to reflect the mistakes that you all are catching--we genuinely appreciate it. That being said, I don't think the value of Elusive Jump should be understated, and High Jump is only a 10s cooldown.

WHM is edging out AST because of Cure III being AoE, and its party-wide regen from Purgation.

GNB's defensive numbers are skewed by Junction DPS dragging the average down. I wouldn't fault you at all if you consider that a flaw of the system, but I would also say that it's not too uncommon for a GNB to matched against a full team of 5 DPS. As for failing to account for burst survivability, we'd have to do some real funky stuff like weighing % mitigation more heavily than healing, and I don't think there's a good answer for doing that. The DEF stat is only trying to measure how much defensive utility is generally available to the job, not how it performs against and in particular situations. AST was also awarded an additional point just for HAVING Bole/Macro/Celestial. Maybe you could make the argument that each of those should constitute a point in and of themselves? But we didn't crunch the numbers on them because of their indeterminate nature.

DNC and RPR having high CC ratings due to unblockable+uncleansable AoE crowd control is a tough one. It's definitely a case wherein the numbers don't tell the whole story...but it's almost like there's a giant disclaimer. At the top of the post. That says you should use common sense and in-game experience to contrast against the numerical system. In any case, DNC/RPR LB are both supremely impactful when they do hit, and we're considering things based on what happens when buttons hit. If we considered counterplay and enemy mitigation for everything, then nothing would ever work and all the numbers would be lower.

[Cidel]

Look at all the new accounts created just to post in this thread...

[Myrha_Lhlalheva]

Look at all the new accounts created just to post in this thread...

engagement tweet go

[Pet_me]

Look at all the new accounts created just to post in this thread...

I post in every cc thread they all get deleted so my post count stays 0

[Aisi]

Not sure what you mean by "not factoring the delay on Apex Arrow", since Apex isn't even part of the ~1GCD STB being calculated. If you mean that there should be another 8400 tacked on there for Blast Arrow, that would break our rules on how may GCD's we're allowing during the window we're considering. If we just gave everyone else another free GCD as well, then...there isn't really a point.

How is it a free GCD when you use silence to confirm the combo and to get the Pitch Perfect proc in the first place? Does silence not count as a hard GCD?

RDM's low mobility rating is likely because we used the rule of "half value for mobility that was only valid on one type of target". It might seem like an unfair assessment, but I don't think you can argue that 2 enemy-only dashes and 2 enemy-only backflips is just as mobile as, say, Thunderclap.

The mobility is flawed if you judge it only as an enemy-only backflip because the utility will be nearly identical - backflipping to disengage from a fight. But according to your metric an untargeted dash is... quadruple the value of an enemy only dash? There is zero attempt to quantify the mobility as a tool to disengage very liberally. If you want to compare Thunderclap to RDM backflip alone, Thunderclap requires you to have a teammate outside of the fight to escape to, while Backflip has so much more freedom to disengage when you want - why does needing an enemy targeted matter in 95% of situations? You're either in melee range doing melee combo, and then you can opt to backflip whenever you would want. Or you're not in that range, and you can simply sprint and use your ranged privilege to reposition as freely as you want. Add on top of that RDM naturally splits gap closer and disengage into two separate cooldowns - you get so much freedom to choose how to engage, disengage, and keep cooldown rollings. 2 gap closers and 2 disengages is stronger than 2 any targeted dashes because you get more freedom in how to use those 4 button presses. Why is it not possible to argue that it's at least comparable to 3 charges of Thunderclap? The entire reasoning feels like it's written from the PoV of someone who doesn't play melee at a high level.

That says you should use common sense and in-game experience to contrast against the numerical system. In any case, DNC/RPR LB are both supremely impactful when they do hit, and we're considering things based on what happens when buttons hit. If we considered counterplay and enemy mitigation for everything, then nothing would ever work and all the numbers would be lower.

So why didn't you use common sense in creating or curating the metrics? If DNC LB only lands, say, 10% of the time because of real gameplay circumstances, how is that reflected in the numbers? If we disregard counterplay and enemy mitigation for everything, then everything ever would work and you get inflated numbers.

[Myrha_Lhlalheva]

How is it a free GCD when you use silence to confirm the combo and to get the Pitch Perfect proc in the first place? Does silence not count as a hard GCD?

The mobility is flawed if you judge it only as an enemy-only backflip because the utility will be nearly identical - backflipping to disengage from a fight. But according to your metric an untargeted dash is... quadruple the value of an enemy only dash? There is zero attempt to quantify the mobility as a tool to disengage very liberally.

If we disregard counterplay and enemy mitigation for everything, then everything ever would work and you get inflated numbers.

You know what? That's a good catch. That would tack an extra 8400 onto BRD burst, which would move it up one point for STB on the rubric. It'd still sit squarely in C tier, and it'd still be the lowest overall scoring job though.

The utility of a targetless dash is nowhere near identical as a targeted only dash. Consider the following: backflipping forwards. Also, Elusive Jump dispels Bind and Heavy, so you can consider its high rating to be a reflection of that bit of utility.

The reason that we erred on the side of disregarding counterplay vs accounting for counterplay is that there's no way to accurately quantify how much counterplay is occurring, but there IS an objectively "all things being equal" value in disregarding it. If we just start arbitrarily deciding how often any given thing works, then we might as well just throw out the whole system and rank stuff by eye.