Screwed again: Proposed THF job adjustments.

[noodles355]

What the math doesn't take into account is that you can't reliably SA once mobs start riding spinnars.

You can. it's not hard. Tell the other melee(s) to stand opposite you. Once you're hate capped, pretty much every action will spin the mob. Just hit SA as soon as you see the other hate-capped melee land a hit. You can land SA very reliably once you realise that the mob doesn't spin randomly, but spins on last actions. Sure, you wont land it all the time, the War could land an attack round, then get critted in the face, losing enmity and forcing the mob to spin back whilst you're waiting on JA delay, but it works the majority of the time.

[Motenten]

Presumed constant elements:

Target: Bhukis
Area: Outside Abyssea
Buffs: Haste
Food: Pizza +1 (acc is a real issue here for both jobs)
Merits: full weapon (GA or dagger), 5% crit, 5 str (forgot to change to 8; don't feel like redoing)
Weapons: Empyreal for both
Gear: tricky to be completely fair, but I'll try to set up solid sets for each
War gets /sam
Thf gets /nin


Variables:
War with/without Berserk/Agressor, with/without Hasso, plus overall average
Thf with various combinations of SA and TA use and stacking


Complications on war:
Percentage distribution of combinations of Berserk/Aggressor/Blood Rage

For Berserk/Aggressor:
3 minute fight: 100%
5 minute fight: 60%
7 minute fight: 71%
10 minute fight: 50%
Very long term average: 43.2%

In the abstract, the 43.2% number should be used. However one can almost always expect that timers will be reset by the time of the next fight, so will mostly be using numbers that are more biased towards matching up. I'm going to use the 50% value.

Likewise odds of neither buff being up can be as high as 28% (at about 4 minutes), but drop down to 11.5% at about 13 minutes. I'm going to use 15% here.

That leaves 35% of the time with one or the other buff up. Giving the advantage to merited Berserk here, I'll put those at 20% Berserk only and 15% Agressor only.

I'll be highly optimistic and place Blood Rage in periods with both Berserk and Aggressor active.

So we have:

Berserk+Aggressor+Blood Rage = 10%
Berserk+Aggressor = 40%
Berserk = 20%
Aggressor = 15%
Neither = 15%

And the same with and without Hasso (depending on the conditions of the fight).

            Hasso      Seigan
BAR       113.001      93.756
BA        101.179      83.956
B         101.053      82.815
A          75.866      62.912
N          75.768      62.023

With overall weighted averages:

Bison Steak (first pass on numbers)
Total      92.823      72.949
Pizza +1
Total      94.727      78.261

Now for thf:

Will consider various combinations of being able to land SA and TA, with or without Rudra's (ODD adjusted appropriately).

No SA/TA or Rudra's:         51.876
No SA, Solo TA, solo Evis:   54.658
Solo SA/TA, solo Evis:       58.585
Solo SA, TA Rudra:           68.822
Solo TA, SA Rudra:           69.240
SA & TA Rudra:               78.499
No SA (miss), TA Rudra:      63.101

So best case for thf matches the war's damage if war can't use Hasso, but short term (when all of war's JAs line up) war will do a lot better. If war can use Hasso, they win by a fair margin (~20%) both short and long term.

Most likely the thf is going to lose a few points from missing SA+Rudra (get ODD, don't get stacked ws) some of the time (3% DPS loss from missing 1 in 5).


Also, I'll be uploading new versions of the war and thf spreadsheets shortly (general cleanup on war, a couple minor fixes for thf), so you can see what gear I happened to pick for the comparison (didn't spend a lot of time trying to optimize the gear).

[SpankWustler]

Target: Bhukis
Area: Outside Abyssea

Even though the hypothetical healer for this hypothetical situation is only hypothetical, I can hear him crying and screaming and pounding his hypothetical Cure V macro. He's pounding that hypothetical Cure V macro as if it's the reason for every hypothetical problem in his hypothetical life.

The result is interesting. As somebody who is usually on a support job or a weird melee job, I did not expect Thiefy Thiefington to come that close to Warry Warrington going all-out. It makes me wonder how Thief's damage would fare if it ever gains Dual Wield III naturally, and could sub Warrior when damage taken wasn't a concern.

[Insaniac]

You can. it's not hard. Tell the other melee(s) to stand opposite you. Once you're hate capped, pretty much every action will spin the mob. Just hit SA as soon as you see the other hate-capped melee land a hit. You can land SA very reliably once you realise that the mob doesn't spin randomly, but spins on last actions. Sure, you wont land it all the time, the War could land an attack round, then get critted in the face, losing enmity and forcing the mob to spin back whilst you're waiting on JA delay, but it works the majority of the time.

I know this stuff but what isn't being and can't be quantified is the extra time you spend looking for the right moment to SA. If you ride the timer on SA you are gonna miss way more than 80% of them or you will have to turn for a second or 2. For instance if you activate SA the WAR takes a hit you get hate back and the mob turns and starts casting so either you run around to the other side or count that SA as a loss. You also can't put all the other DD on 1 side of the mob on tons of NMs. For example Bhukis.. if they are all standing on one side you are looking at getting everyone of them doomed except you just so you can land SA more reliably. You just can't do that on any mob with a cone attack. It's not reasonable to ask people to do that on an ironclad type mob with cone melee or a MNK mob because of counters and guards. TA is easy and you can ride the timer no problem but with SA there's a whole series of variables that are stacked against you getting the max potential out of it.

[Motenten]

Just a minor note regarding the issues Insaniac brings up:

Yes, that's certainly an issue. It's roughly accounted for in the spreadsheet by allowing you to specify an average amount of time beyond the recast timer for you to be able to use the abilities. The default is 5 seconds, which corresponds to being able to use the ability anywhere from 0 to 10 seconds after the recast is up, which should cover most minor fluctuations. It does not, however, account for lost melee damage from the times you may need to turn away briefly.

As for the conal issue, no, you don't want all the DDs stacked on one side of the mob. At the same time, most of the DDs don't want to be in front of the mob anyway, so usually they'll try to let the 'tank' keep hate as much as possible rather than spin the mob and have to eat that doom move, which in turn makes it easier to land SA.

[Insaniac]

Just a minor note regarding the issues Insaniac brings up:

Yes, that's certainly an issue. It's roughly accounted for in the spreadsheet by allowing you to specify an average amount of time beyond the recast timer for you to be able to use the abilities. The default is 5 seconds, which corresponds to being able to use the ability anywhere from 0 to 10 seconds after the recast is up, which should cover most minor fluctuations. It does not, however, account for lost melee damage from the times you may need to turn away briefly.

As for the conal issue, no, you don't want all the DDs stacked on one side of the mob. At the same time, most of the DDs don't want to be in front of the mob anyway, so usually they'll try to let the 'tank' keep hate as much as possible rather than spin the mob and have to eat that doom move, which in turn makes it easier to land SA.

I think your spreadsheets are an amazing work. I'm not trying to take anything away from them. I would never try to blame you for not quantifying things that can't be quantified. The only thing I'm trying to get across are that these X-factors for THF only widen the gap between the THF and WAR spreadsheet results.

[Motenten]

It makes me wonder how Thief's damage would fare if it ever gains Dual Wield III naturally, and could sub Warrior when damage taken wasn't a concern.

In that hypothetical case, using the best combo (stacked SA and TA Rudra) but no other changes, DPS is 102.990 with Berserk up and 82.315 with Berserk down; weighted average is 94.72, on par with war using Hasso.

[Insaniac]

I made some edits to the OP like I said I would weeks ago. Mostly trying to better represent the content of the thread since it has become much more than just THF rage (which there is still plenty of) The main thing was links to adjustment ideas posted in the thread. I tried to catch as many as I could but may have missed them if they were short posts. If anyone has one they would like included find it and post the link and I will add it if it's not already covered in one of the other posts.

It was funny skimming back through the pages and seeing the different directions the thread had taken. The basic cycle is:

SE posts something > We rage > we calm down and discuss > we digress a bit > SE trolls us again and the process repeats

[SpankWustler]

Thanks for taking the time to plug in numbers from what was just me day-dreaming. I wasn't expecting that.

So all it would take for Thief to stand on fairly level ground with the most damaging jobs at 99 is if their Dual Wield grew every eight levels, similar to the growth rate of Critical Attack Bonus. That doesn't sound unrealistic at all, which makes it all the more depressing that it's very unlikely to happen.

[Motenten]

Will also add one extra note to the factor of scaling damage with haste. When adding just one March (not double-march; am allowing the second to be scherzo or mambo or whatever is appropriate for a dangerous mob), war's DPS increases by 33% (Hasso) or 27% (Seigan), while thf's DPS increases by 12.5% with both SA/TA stacking, and up to 19% with only TA stacking (though the total DPS is lower than in the 12.5% increase's case).

Hasso war ends up about 42.5% ahead of best-case thf, while Seigan war ends up about 12.5% ahead. The hypothetical DW3+/war thf improves by about 20%, so the gap between fully offensive thf vs fully offensive war would still remain at about the 20% range.

So even for the comparisons where, under optimal conditions, thf can match war, as soon as a brd is added war pulls waaay ahead of thf. Thf gets too bogged down in JA timer limitations to be able to take full advantage of the extra haste (though part of that is the fault of the spreadsheet which can't easily deal with complex mixtures of stacked and held TP).

And of course as Insaniac points out, war pretty much always works at best-case or better (Retaliation), while thf usually works below best-case.

On the other hand, thf has faaar better raw survivability than war. Any time I'm not on thf, dnc or mnk, I always feel like I'm naked and have a massive target painted on my back (and front.. and head). War can manage with good support, but it's always painful to watch them try to survive anything even mildly difficult if they're not being constantly babysat.

I could certainly live with the trade-off in survivability vs offensive power if SE would make a more serious effort to work on the utility aspects.