4.05 - Tank DPS comparison

[Galgarion]

Hey, since we're talking about tank damage here, maybe I can pick someone's brain in an effort to improve. I'm just 64, but I'm trying to pay attention to my rotation so I can be less of a mindless button masher.

So, my dps rotation is basically me trying to squeeze in as many fell cleaves during berserk as I can. The most I can manage is four since I don't have unleash yet (plus my gear is basically lower tier last expansion). My rotation when my gauge is full is : eye combo-->heavy swing, maim-->berserk--> Fell cleave x2-->Storm's path to finish the earlier combo--->path combo-->fell cleave-->infuriate-->fell cleave.

So I can get the four cleaves in, but my question is: 1. Is that good, and 2: Does that mean I'll be able to do eight fell cleaves when I get unleashed?

Sincerely,

Bottom Feeder trying to git gud.

[DWolfwood]

snip

4 Fell Cleaves is good at your level. You won't be able to do 8 because each Fell Cleave you do is going to hinder your Beast Gauge accumulation, so it's a bit of diminishing returns basically.

As of now, I think we're looking at 6 Fell Cleaves. It was 6 in 4.00 and 4.01, but it was more of a meme really and nearly impossible to ever achieve. However with the new changes, I think 6 might be more realistic now. That said, I do know some of it is just going to amount to doing more Onslaughts during your window.

[Galgarion]

4 Fell Cleaves is good at your level. You won't be able to do 8 because each Fell Cleave you do is going to hinder your Beast Gauge accumulation, so it's a bit of diminishing returns basically.

As of now, I think we're looking at 6 Fell Cleaves. It was 6 in 4.00 and 4.01, but it was more of a meme really and nearly impossible to ever achieve. However with the new changes, I think 6 might be more realistic now. That said, I do know some of it is just going to amount to doing more Onslaughts during your window.

Okay, cool. Four is good, huzzah! But why go with onslaught over upheavel? Isn't Upheavel supposed to be our big new toy?

[DWolfwood]

Okay, cool. Four is good, huzzah! But why go with onslaught over upheavel? Isn't Upheavel supposed to be our big new toy?

Upheaval is the strongest yeah, but since it has a 30s recast, you can only fit one inside your window, whereas you can fit 2 Onslaughts in it. Basically, you'll be doing all 3 in your window.

[Quor]

Here's an example:

Let's say you started out with 2400MP and 30 Bloodgauge to speed up what I'm trying to illustrate without having to go through 5+ combos.

Here's your rotation without using TBN:

1. Hard Slash (150 pot)
2. Syphon Strike (+1200MP) (250 pot)
3. Souleater (+10 Bloodgauge) (300 pot)
4. Hard Slash (150 pot)
5. Syphon Strike (+1200MP) (250 pot)
6. Dark Arts + Souleater (+10 Bloodgauge) (440 pot)
7. Dark Arts + Bloodspiller (540 pot)

Remaining MP: 0
Remaining Bloodgauge: 0
Total Potency: 2080

We're now at 0MP and 0 Bloodgauge because we use it all. Note, that we did this in 7 GCDs because that's important and fundamentally how we get "DPS" to begin with.

Now lets look at if we use that 2400MP on TBN:

1. Hard Slash (150 pot)
2. TBN (-2400MP + 50 Bloodgauge) + Syphon Strike (+1200MP) (250 pot)
3. Souleater (+10 Bloodgauge) (300 pot)
4. Hard Slash (150 pot)
5. Syphon Strike (+1200MP) (250 pot)
6. Dark Arts + Bloodspiller (475 pot)
7. Bloodspiller (400 pot)

Remaining MP: 0
Remaining Bloodgauge: 40
Total Potency: 1975

Wait, so, does TBN mana cost double at 70? Cause, at least according to what it shows me on my 60 DRK (using mana costs appropriate for the level) TBN and DA cost the same, while SS restores the exact same amount as the cost of both DA and TBN. Or does SS just lose half of it's MP restore at 70?

I think you're talking about literally it breaking your combo the way certain abilities like Unmend (iirc), old Sword/Shield Oath, etc., which is something else entirely.

Ok, that's good then.

That said, it's actually the same thing for Fell Cleave. Every time a Warrior uses Fell Cleave, they're losing out on Beast Gauge accumulation. Furthermore, we can apply the same math and actually find that the real potency increase of Fell Cleave is around half (approx. 250ish, this isn't the thread for that though). When you know this, Onslaught is actually a great skill in that it's roughly the same potency per Beast Gauge meter as Fell Cleave since Onslaught is not a GCD.

See, this is what I have a problem with. I get the underlying premise; anything that stops the generation of resource has a "cost" that can be converted into potency based on the potential resource gain that has been "lost." I just disagree with how it's presented, because it's being treated as a net-loss, without sufficient attention to the benefits gained by utilizing said resource. After all, everyone can agree that Fell Cleave has increased WAR dps since it was introduced; if it didn't no one would use it. But we have, at this point, probably millions of parses showing that WAR's who use Fell Cleave do more dps than WAR's that don't. So yes, there is a "cost" in terms of the potential resource gain being "lost" but it seems like there isn't adequate consideration for the benefits. Resources exist to be consumed, and anything that results in "wasted" resource gain is worse than if you had never used that resource.

I dunno if this will help express my point better, but based on what I've seen so far, you gain about 10 blood per SE combo. Thus, 5 SE combos allows for a single BS use, or what's called a "natural" BS. In that sense, you could consider one broken TBN to be, on a blood-generation level, equivalent to 5 SE combos. Yes, you pay the MP cost towards TBN instead of DA, but it's not a gain of 400 vs 440, because you can choose to delay your BS as long as you're not at 100 blood. This means you still get the 300 potency from the non-DA SE, then on top of that you get a further 400 potency from BS.

In short, I'm seeing it like it's presented as an either-or option, when I'm wondering "why not both?" Yes, you lose 140 potency from using TBN instead of DA, but if the mana costs are equal and you're at 40 blood or below, you can still finish that SE combo (gaining the 10 blood) and then follow it up with a BS or two.

Yes, I get that, over thousands or hundreds of thousands of rotations, things even out, but no practical fight in this game exists over thousands of combos. There are burst windows to consider, whether that's an aoe Balance, a TA on the mob, Brotherhood from a MNK or just a burn phase where you need to put out maximum damage. You're not attacking some monolithic entity for thousands of unending rotations. Putting out a few higher potency skills in a shorter period of time may mean the difference between a win or a loss, and those higher potency skills add up over time.

Maybe an anecdote from ancient gaming history will help; I used to be a pretty avid WC3 player back in the day when it first came out. I did a lot of multiplayer, and Orcs were my favorite race. I often started with Blademaster, and I was somewhat surprised to see that, on the official forums, a lot of people were poo-pooing Blademaster because they saw him as a much weaker option compared to the Farseer or Tauren Chieftan. The major point being made was that, statistically, unless you got a lot of lucky +agility drops (for more dmg, armor and attack speed on your BM) then the Chieftan or Farseer would be a "better" hero for damage, either due to the presence of aoe or because of the Farseers dogs adding a pair of extra units to your army.

The problem was, these people were basing their judgments off of that theoretical "wall" I keep bringing up. They were running loads of calculations assuming that you would be dealing X damage over a given period of time, wherein that given period of time was obscenely large so as to "control for variables." The issue at hand is that the BM was all about those variables because he had the passive skill of Critical Strike. On top of this, Warcraft 3 wasn't about sending an army against a massive, passive HP wall; it was about eliminating your opponents army and base to secure a win. No unit in that game was a massive wall, not even Mountain Giants (although they could come close). Each one was a discrete entity, and thus, for each one killed, you weakened your enemy.

The nature of the Blademaster's crit ability was such that it could not be accurately depicted in the huge theoretical calculations that were being done by theorycrafters on the forums. Sure, in the long run, the value of that crit is negligible, but in a game with dozens of discrete units, each with their own comparably (to the Wall) small HP values, being able to quickly and suddenly kill one of them is a huge boon. A pair of crits back-to-back may be enough to two-shot the enemy Archmage. Robbed of their hero, the Human army would be at a huge disadvantage.

In FF14, this is most often represented in fights with adds, but there are occasional moments of discrete DPS on bosses themselves where you can't fairly apply the "theoretical wall" approach, simply because even those bosses aren't meant to be fought for more than 10-15 minutes at a stretch. Cruise Chaser with his Plasma Shield is a solid example of this. The adds in A12 are another one.

I'm not sure if this will adequately convey what I'm trying to say, but if it does great. If not, that's ok too. All I can say for sure is that I look at this math and I get the same feeling that I've gotten quite a few times when people theorycraft something out to the Nth degree; it's great on paper, but it doesn't apply to real world situations as well as people think it would. I've learned to trust that feeling over the years.

[SyzzleSpark]

Wait, so, does TBN mana cost double at 70? Cause, at least according to what it shows me on my 60 DRK (using mana costs appropriate for the level) TBN and DA cost the same, while SS restores the exact same amount as the cost of both DA and TBN. Or does SS just lose half of it's MP restore at 70?

Sounds like someone spends too much time in Grit.

Syphon only restores a DA's worth in Grit, out of Grit it is half. This entire discussion is about proc'ing extra Bloodspillers with TBN out of Grit. In Grit it is a gain and has always been because of Bloodspiller's Grit potency.

Theorycrafting on DRK in 4.0 has been a depressing affair. There's definitely losses and gains to be mathed out, but they're so tiny as to be insignificant, hence the feeling that the job is "braindead" and doesn't reward optimization.

[Quor]

Sounds like someone spends too much time in Grit.

Syphon only restores a DA's worth in Grit, out of Grit it is half. This entire discussion is about proc'ing extra Bloodspillers with TBN out of Grit. In Grit it is a gain and has always been because of Bloodspiller's Grit potency.

Ah, good to know.

And the rest of my post?

[SyzzleSpark]

Ah, good to know.

And the rest of my post?

Yes, you pay the MP cost towards TBN instead of DA, but it's not a gain of 400 vs 440, because you can choose to delay your BS as long as you're not at 100 blood. This means you still get the 300 potency from the non-DA SE, then on top of that you get a further 400 potency from BS.

Leaving aside for the moment the fact that the rest of your post is a novel about a job that you have not yet gotten to 70, you're disregarding math in favor of... well basically nonsense. People are telling you that over a given number of GCDs using TBN+BS is a potential loss against not using it over the same number of GCDs, and you're basically saying "well, yeah, over that given number of GCDs, but after that you still get the potency yadda yadda..."

Saying "Well, you can just delay using a GCD for this and then use that afterwards" increases the data set. If you use Bloodspiller instead of Souleater, and then you come in and say "well, we can use Souleater afterwards!" then you've increased the data set. We're talking about 4 GCDs now instead of 3. And it doesn't matter because you still spent 2400 mana and gained *nothing*, because you spent 140 potency (2400 mana) to gain roughly the same amount against your PPGCD, on average, spread across any number of GCDs in an encounter. Spending that 2400 mana on a Dark Arts is a direct potency increase, because it is off the GCD and doesn't delay your resource generation, over your base Souleater combo.

You can't increase the data set for one set of statistics and leave it the same for the other. If the given number is 3, 4, 5, or 100 GCDs, it doesn't matter, it still at best only breaks even.

[Galgarion]

Upheaval is the strongest yeah, but since it has a 30s recast, you can only fit one inside your window, whereas you can fit 2 Onslaughts in it. Basically, you'll be doing all 3 in your window.

Appreciated. Thanks a lot.

[Quor]

snippety

Disregarding the facts that I've seen a lot of DRK's who do not have WAR at 70 try and tell me how to play WAR.....
Theory is still just theory. To paraphrase the saying, if you want to know how something "should" work, ask an engineer. If you want to know how something actually works, ask a mechanic.

I'd like to see some real world results. If I had DRK at 70 I'd do it myself, but I've been hitting the WAR tank cheevos pretty hard.

Well, pretty hard for someone who doesn't powergame anyway. Just another 124 to go! Kill me now

Take a fight, like O1 for example. It's a pretty good setup for this, because there's high uptime on the target, not a lot of mechanics that will stop GCD rolling for very long, it's solo tankable so the OT can stay DPS the entire time, and there's lots of errant aoe capable of popping TBN bubbles. Do a few, 3-4 tops, of those fights just focusing on DA+SE, with no TBN's used at all. The only BS usage would be from naturally generated blood. Then do the same for 3-4 fights, only this time, use TBN as much as possible during times when it would pop. O1 has a lot of errant aoe, so there are plenty of opportunities where tossing it on a squishy will ensure something eats the 10%. Hell, you can even go in with friends and tell them to strategically stand in certain aoe's to ensure maximum TBN popness. The point is to try and get as many BS via TBN as possible, in addition to natural BS.

Then compare the two. 3-4 fights should be enough to get an idea of where it all stands, and if there's no significant difference then we have our answer, using real world data instead of theorycrafted on-paper numbers. If there's a difference, one way or the other, that's significant as well. But I haven't seen any direct comparisons in a controlled environment like that (which makes sense, since Omega has only been out for 2 weeks and I doubt anyone thought to try a test like this yet). Obviously any fights that have "derp" moments would be tossed out of the pool and redone, but you get the general idea.

Hard data in a real situation is what would convince me. I enjoy the fun of theorycrafting as much as the next guy, but the proof, as they say, is in the pudding.

Oh, almost forgot:

You can't increase the data set for one set of statistics and leave it the same for the other. If the given number is 3, 4, 5, or 100 GCDs, it doesn't matter, it still at best only breaks even.

I accounted for that earlier in the thread.

SE combo = 700 potency over 3 GCD's.
Bloodspiller = 400 potency for 1 GCD.
TBN, if broken, allows for another BS due to the 50 blood gained. Thus, for this example, I'll assume TBN was used (but not DA, on BS or SE, that'll come later) to allow for a third BS, and that the DRK is at 100 blood gauge going in. Now the SE combo takes place over three GCD's, while BS is a single GCD, so we'll be comparing a point at which the two intersect, which will consist of 12 GCD's. This is 4 SE combos or three SE combos + 3 BS.

4 SE combos over 12 GCD's results in a net gain of 2800 potency.
3 SE combos + 3 BS over 12 GCD's results in a net gain of 3300 potency. Again, this assumes no DA's in either combo, and single use of TBN in the second combo to enable a third BS. But even without that third BS, if we just factor in a Hard Slash as the 12th GCD (instead of a BS) then that's still 3050 potency total over 12 GCD's, which is more than just running the SE combo.

Now let's look at what DA does, compared to TBN (since the argument is that TBN is always a potency loss over DA). For this example, we will use four DA's in the SE combo (all on SE) and 3 DA's + one TBN in the BS combo. That means an equal amount of mana was spent, and again we are starting at 100 blood gauge in the tank for the BS combo.

4 SE + DA combos over 12 GCD's results in a net gain of 3360 potency.
2 non-DA SE combos plus one DA+SE, and 3 DA+BS (including TBN) over 12 GCD's results a net gain of 3860 potency.

Or, for those who would prefer to see an average, you have 280 PPGCD for the pure DA+SE combo vs 321.7 PPGCD for the combo that uses 3 DA+BS and TBN.

Now that of course is an extra TBN in the BS combo's, but you can easily subtract 140 potency to account for one less DA in the BS combo's. That still leaves the BS combo ahead of the pure DA+SE combo 3720 vs 3360.