Summarized %/s List

Using Hypers, 0% Defense, Fastest (2), Adjusted Mastery, no outside bonuses.
[fleft]

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47

[/fleft]

#	%/s	100% Boss	200% Boss	hits/s	m/s	Class
1	20592	31956	43369	42.39	42.84	Kaiser Transfigured Command [GMS]
2	19903	29590	39185	20.19	954.68	Angelic Buster (Soul Seeker) [KMS]
3	17001	26032	35061	16.54	793.68	Battle Mage
4	15114	23653	31987	23.4	1108.52	Wind Archer (Sky Song)
5	15037	25004	34984	44.99	2232.93	Kaiser Untransfigured [KMS]
6	14947	25519	35990	22.83	1180.87	Wild Hunter
7	14540	24474	34439	14.67	719.15	Marksman
8	13461	19903	26276	15.38	441.14	Fire-Poison Archmage (Mist Eruption)
9	12956	20229	27482	15.34	805.25	Demon Avenger
10	12839	19045	25233	3.83	170.31	Angelic Buster (Trinity) [KMS]
11	12532	18551	24662	10.83	568.18	Hero
12	12133	18687	25194	9.42	487.74	Evan
13	11716	16039	20358	9.23	366.9	Shadower (Assassinate Only)
14	11711	18362	24595	13.23	136.99	Angelic Buster (Soul Seeker) [GMS]
15	11082	16494	21902	16.78	745.91	Shadower (Edge Carnival + Assassinate)
16	10808	16696	22577	3.83	160.37	Angelic Buster (Trinity) [GMS]
17	10789	19263	27750	24.48	967.68	Blade Master (No CD Final Cut)
18	10573	16752	22931	29.44	1075.37	Thunder Breaker (Ascend + Thunder)
19	10538	15044	19551	11.28	656.0	Demon Slayer
20	10314	16325	22337	21.32	777.26	Thunder Breaker (Thunderclap)
21	10067	14499	19030	22.06	978.39	Xenon (Blade Dancing)
22	9852	14631	19363	9.7	481.76	Mikhail
23	9704	14934	20274	26.37	1559.28	Mechanic
24	9596	15069	20557	26.09	1364.54	Cannon Shooter
25	9340	13244	17195	13.9	14.6	Hayato
26	9338	16516	23715	24.75	979.54	Blade Master
27	9320	14005	18676	12.6	546.26	Ice-Lightning Archmage
28	9308	13258	17192	17.92	567.47	Fire-Poison Archmage (No Mist Eruption)
29	9230	14244	19258	19.2	1199.83	Dawn Warrior
30	8997	13785	18574	21.35	778.1	Thunder Breaker (No Typhoon Spam)
31	8746	12487	16211	14.56	545.05	Xenon (Fuzzy Masquerade - Sniper)
32	8338	15063	21765	10.77	564.84	Paladin
33	8151	11845	15018	20.95	733.92	Corsair
34	8006	11899	15697	11.96	689.62	Buccaneer
35	7786	12515	17243	6.82	357.86	Aran
36	7613	11652	15563	22.98	1160.14	Mercedes
37	7438	12498	17556	14.51	761.64	Luminous [KMS]
38	7418	11504	15583	8.38	466.43	Bishop
39	7225	11811	16381	18.94	779.1	Night Lord
40	7133	10052	12963	20.88	18.86	Kanna (Shisen Gyouha)
41	6658	11056	15504	14.68	807.44	Phantom (Ultimate Drive)
42	6416	11064	15579	13.42	737.93	Phantom (Rapid Fire)
43	6193	9152	12078	18.86	16.86	Kanna (Haja Rengekifu)
44	6051	10207	14298	22.05	1098.28	Bow Master
45	6009	9484	13032	9.17	7.95	Jett
46	5504	10066	14615	9.62	504.76	Dark Knight
47	4770	8923	13015	13.75	14.32	Luminous [GMS]

There is one concern I would like to point out; ATK% are currently accounted for in the %/s since they linearly affect damage, unlike other stats. There is, however, a flat cap of 2m~2m on a damage range, so at that point, ATK% would not make you any stronger unlike Total Damage% or Boss Damage%. The classes affected by this are Battle Mages and Wild Hunters.

Summons are currently accounted for using the max range. More research into which summons can break cap, and are affected by total damage/boss damage will need to be done to correct this.

Since I'm so brilliant and deleted the thread accidentally and remade it by hand, please let me know if there's any errors compared to the original threads, since I retyped each number instead of copy-pasting as that was significantly faster.

Change Log

• Accidentally deleted entire thread with no backup [18/01/13]
  
• Updated GMS Kaiser to include Majesty of Kaiser resetting Final Trance [19/01/13]
  
• Updated Cannon Shooter's Buckshot to be 40% damage, not -40% (60%) damage. [04/02/13]
  
• Added Thunder Breaker [04/02/13]
  
• Added Wind Archer [28/02/13]
  
• Fixed Thunder Breaker's Ascend Thunder combo
  
• Updated Mercedes & Phantom to most recent KMS version [24/03/13]
  
• Fixed Shadower's Venom damage and Wind Archer's Spiral Vortex[27/03/13]
  
• Removed Wind Archer Spiral Vortex as it is redundant [02/04/13]
  
• Updated Kaiser, Angelic Buster, and Luminous for GMS [05/04/13]
  
• Updated Hayato with max defense and fixed the DoT [05/04/13]
  
• Added Dawn Warrior [29/04/13]
  
• Updated Blade Masters Final Cut to be multiplicative [29/04/13]
  
• Updated the majority of the table for the values in my most recent tables. Let me know if you think anything is off and I will investigate. [23/05/13]
  
• Updated Aran and Battle Mage; the last table seemed to use an SE value for Arans [17/06/13]
  

Alas for Phantom. ;_; Then again my former main (Bowmaster) hardly fared better.

Surprised how high demonic avenger is

The Xenon numbers you posted in the Thief thread has over 17k %/s with hypers. It looks like the number here doesn't include Hyper skills.

The Xenon numbers you posted in the Thief thread has over 17k %/s with hypers. It looks like the number here doesn't include Hyper skills.

My bad. Thank you for noticing.

Thanks a lot!

I'm surprised by Cannon Shooter, didn't expected them to be good if extremely funded(If I'm getting it right?).


Overswing is needed to use Beyonder in GMS.

As a point of clarification, these numbers are fully-self-buffed with all %damage buffs included but NOT including +WATK/MATK buffs, correct? For classes like Angelic Buster and Xenon, since they have very high %damage modifiers, their numbers are a little misleading because they aren't as affected by %boss as other classes, and classes like DB are under-represented because of their high +WATK modifiers (Katara, Thorns, Channel Karma, but DOES include Final Cut as it is a % bonus).

I understand the way you calculate these %s, you are supposed to put in your range and multiply by the % to get more or less your real DPS. I'm not entirely sure how you include buffs that increase range because they're %-based.

Pretty interesting how all five classes are actually represented by at least one job at the top. Figures that warriors are overrepresented, but oh well.

I love how the bottom 4 (5 if you eliminate Phantom Rf completely, since it's pretty much useless) consists of three hero classes. Heroes indeed.

I actually have a question regarding whether or not Arans require the use of Overswing to use Beyonder. It was previously not necessary, and then I saw evidence that it was necessary. I have not checked since Tempest.

I can confirm it is necessary to use OS as of GMS Ver 125.1.

I actually have a question regarding whether or not Arans require the use of Overswing to use Beyonder. It was previously not necessary, and then I saw evidence that it was necessary. I have not checked since Tempest.

Overswing + Final Blow is the only way to activate Beyonder.

As a point of clarification, these numbers are fully-self-buffed with all %damage buffs included but NOT including +WATK/MATK buffs, correct? For classes like Angelic Buster and Xenon, since they have very high %damage modifiers, their numbers are a little misleading because they aren't as affected by %boss as other classes, and classes like DB are under-represented because of their high +WATK modifiers (Katara, Thorns, Channel Karma, but DOES include Final Cut as it is a % bonus).

I understand the way you calculate these %s, you are supposed to put in your range and multiply by the % to get more or less your real DPS. I'm not entirely sure how you include buffs that increase range because they're %-based.

Pretty interesting how all five classes are actually represented by at least one job at the top. Figures that warriors are overrepresented, but oh well.

I love how the bottom 4 (5 if you eliminate Phantom Rf completely, since it's pretty much useless) consists of three hero classes. Heroes indeed.

I believe that is indeed correct, and most unfortunately it cannot be changed easily. It might take half an hour or so per class to construct a linear correlation using all the +WA/MA buffs internal to a class, but I believe Joe's not doing so because each class's weapon has different WA ranges (e.g. claw vs sword for the same level), and WA/MA is extremely easily swayed. For the same reason, I cannot use Joe's tables when I'm doing my own DPS calculations for my Phantom and permutating the various skills for Impeccable Memory (FTR, Dark Harmony's 40WA > Vipercision's 15%WA for me). I don't doubt Joe's calculations for across-the-board calculations, but when I do my own numbers I don't assume anything at all (esp. crit proc rates for different skills as a binomial distribution - does make life hell for me when I'm doing single-hit vs multi-hit skills on my Phantom; don't know if Joe's taking that into account). This further explains why Joe doesn't take into account +WA/MA - choices are changed at different breakpoints.

Did I say anything terribly wrong here, [MENTION=535]JoeTang[/MENTION];?

Hadriel

>paladin stronger than phantom


excuse me while i die laughing on the floor

You should at least make cannoneers at fast 3/4 instead of faster 2.

Also i have a selfish request if you're up for it. How significantly does ignoring cooldowns 20% of the time affect %/s? (Cannoneers specifically but it might be cool for other classes since it's easily the best inner ability if it works)

You should at least make cannoneers at fast 3/4 instead of faster 2.

Also i have a selfish request if you're up for it. How significantly does ignoring cooldowns 20% of the time affect %/s? (Cannoneers specifically but it might be cool for other classes since it's easily the best inner ability if it works)

Practically nothing.

It'd be nice if it was easier to estimate the ranges, since percents only tell half the story - my Hero's range is ~40k with the exact same gear as my Kaiser, who has ~54k (66k transfigured), 25 levels lower. Likewise, my Aran's range is near double my Dark Knight's, at the same level with the same gear.

It'd be nice if it was easier to estimate the ranges, since percents only tell half the story - my Hero's range is ~40k with the exact same gear as my Kaiser, who has ~54k (66k transfigured), 20 levels lower. Likewise, my Aran's range is near double my Dark Knight's, at the same level with the same gear.

aran range displays snow charge though, which i think is also in JT's calculations, so it would be more like range/snowcharge for the actual thing.

The next step for such calculations, which takes a lot of effort, is to do linear correlations for all classes for all variables: (all)stat, WA/MA, %ignorePDR, %bossdamage, %dmg, %stat/all, attackspeed. The idea would be to provide an initial intercept based on all the internal buffs a class would possess, taking into account both passive and active skills (also factoring in some casting delay). The crucial bit would then be the gradient for each variable /edit: because that shows the true potential of a class. Naturally one can turn this into a multi-dimensional plot i.e. a real damage calculator/optimiser if you're given several constraints (+30%ignorePDR or +6%stat, which might be potentially better?). Only then can one truely say which class is superior 1v1, because you can start to put in conditions: how funded? amount of %var based on no. of eq pottable? Solo buffs with potions? Full party with all buffs and potions etc.?

Standard assumptions would of course be level 200, normalised WA, no potential. Linear correlation => changing one variable at a time obviously but it doesn't take a genius to realise that some variables are multiplicative while some are additive. But even then, I think it's so much effort it's almost unreasonable for Joe to do this.

What say you, @JoeTang;? What would be the next practical fair comparision of classes 1v1?

Hadriel

The next step for such calculations, which takes a lot of effort, is to do linear correlations for all classes for all variables: (all)stat, WA/MA, %ignorePDR, %bossdamage, %dmg, %stat/all, attackspeed. The idea would be to provide an initial intercept based on all the internal buffs a class would possess, taking into account both passive and active skills (also factoring in some casting delay). The crucial bit would then be the gradient for each variable. Naturally one can turn this into a multi-dimensional plot i.e. a real damage calculator/optimiser if you're given several constraints (+30%ignorePDR or +6%stat, which might be potentially better?). Only then can one truely say which class is superior 1v1, because you can start to put in conditions: how funded? amount of %var based on no. of eq pottable? Solo buffs with potions? Full party with all buffs and potions etc.?

Standard assumptions would of course be level 200, normalised WA, no potential. Linear correlation => changing one variable at a time obviously but it doesn't take a genius to realise that some variables are multiplicative while some are additive. But even then, I think it's so much effort it's almost unreasonable for Joe to do this.

What say you, @JoeTang;? What would be the next practical fair comparision of classes 1v1?

Hadriel

I don't have any particular interest in building an in-depth damage calculator. It has been tried in the past to standardize gear, and the basis of which I never found to be meaningful nor fair. It fundamentally comes down to that no two characters will in any likelihood ever have the same range other than 2m~2m. All gear except for the weapon can basically spawn with identical stats across all classes. Weapons possess a difference in weapon attack, but the sum weapon attack as it gets higher results in a smaller difference. Thus, the largest differences between classes are what inherent stat bonuses, and weapon attack bonuses they may possess. To draw parallels across all 32 classes, which may or may not have differing base weapon attacks, and differing additional stats and weapon attack, amongst other bonuses is beyond the scope of what my tables present. With the introduction of Demon Avenger and Xenon, the difference in ranges is even greater; Xenon gains 3 times as much of a bonus from gear as any other class does. Demon Avenger gains bonuses from a stat that no other class does.

Presently, the largest scale application towards difference in funding I could apply, which I also believe to be the most meaningful, is a difference in Total/Boss damage.

aran range displays snow charge though, which i think is also in JT's calculations, so it would be more like range/snowcharge for the actual thing.

But does it displays +20% Boss Dmg afrom Cleaving Blows and +5% from regained memory?
[MENTION=535]JoeTang[/MENTION]; ?

I knew that Insoya/Leafre table was horribly off. The creator honestly thought Buccs were rank 3.. really? Thanks for doing these tables [MENTION=535]JoeTang[/MENTION];

Batman maybe a awesome, handsome, rich badass, but he only has a few strong moves at the same time unlike other super heroes who have a lot of strong moves at the same time.