4th job Single Target DPS

Why does it matter? Are we trying to make it fair here? I thought the point was to see who does the most single target DPS, so someone's got to come on top.

Plus, we're not even being realistic here. The majority of people don't reach level 200, so might as well have a maxed out category.

the point is definitely to make it fair/realistic with respect to funding, we know what weaksauce and prosauce dps's look like. but the next question is obviously, if i have 2b mesos and get some decent gear what is my dps? and whats the difference between the classes at some set amount of money?

its hard to answer because of market fluctuations and such. realism is much harder to model, but far more useful.

the point is definitely to make it fair/realistic with respect to funding, we know what weaksauce and prosauce dps's look like. but the next question is obviously, if i have 2b mesos and get some decent gear what is my dps? and whats the difference between the classes at some set amount of money?

its hard to answer because of market fluctuations and such. realism is much harder to model, but far more useful.

Exactly, it's next to impossible to model such a thing, so it's much better just to take these numbers with a grain of salt and use them as a reference rather than truth.

Look into JoeTang's damage calculator. Something like that, except maybe less elaborate.

Actually, reread what you wrote and it seems like this is something we already added [in the new version]. Under the stats listed in the physical base attack sheet, you can input your own numbers if desired. It's limited to inputs for total attack, main stat, and secondary stat(s), rather than individual equip stats, but it's there if you need it.

Edit: Been working on this all night. L> feedback

http://rapidshare.com/files/226642493/4thdps1.4.xls

Too tired to edit first post and stuff.

Hmm..with 4,000 30% scrolls working on the weapon, a 1H Sword Paladin does over twice as much DPS as a Marksman.


er...good new feature though.

Can't wait to see what majorly reduce snipe/big bang/heaven's hammer cooldowns do to MM, AM/Bishop, and Pally damages, respectively.

If I may, I'd like to provide the necessary counterpoints. You can read or not read them as you choose, but if you read them I ask that you read them fairly, and not confuse mere scrutiny with trying to shape all commentary from others to fit with your own predispositions. I'm trying to avoid a hostile tone as much as I can, and for my part I'd say other than the equip setups this is a very good comparison. But trust me, as a person who majored in this kind of thing in college, and as the author of at least seven guides for this game and more for at least six other games, I cannot in good faith say that this study's setup is accurate as it stands. It's convenient, maybe, but not accurate.

I'll counter that the way you want to do it is more biased than the way I've done it. NL equips are only expensive because they are such a popular class, not because their items are inherently expensive. What's the point of comparing classes based on funding?

Well, I would argue from an academic standpoint that, since you're the writer of a study that people may actually use as a credible reference (and many do, not without good reason I might add), you have a responsibility to follow the basic standards used for all studies involving statistics (this thread being one). And in all credible studies of this sort, one of those responsibilities is that you have to ensure the data are analyzed in a controlled environment in which outside factors have no influence over the findings. In this particular comparison, it happens that by using unequal funding for the equips, you're failing to provide that controlled environment and allowing funding to become a factor that influences the outcome. Building "How does a Night Lord with 2.5 billion worth of equips compare to a Hero with about 0.8 billion (or less in some cases) worth of equips?" violates the "controlled environment" aspect of the study and therefore discredits its results.

If the environment is set up such that in practical uses, any nonrandomized outside factor can skew the results in any way (like the above case with funding can), then the study's results would be laughed out of any academic environment. At least, that's what I learned while I was majoring in applied math & statistics (which is pretty much the foundation for data analysis). I'm not sure whether they taught it otherwise at whichever school you learned data analysis from, or whether my professors may have been wrong about that.

Who is that information useful to? I highly doubt anyone sits around figuring out which class they should make based on a certain amount of funding.

Well isn't that the point of this comparison? Comparisons like this are made not to fuel anti- or pro-(insert class name here) sentiments (if that were one's cause, they probably have no right being the one making the comparison), but to work out for players what they could do if they made a certain job with the resources they actually have. That way, they don't *need* to do it themselves. At least, that's what I've written most of my guides for. I'm not sure whether that's your priority as well.

Hitherto I've only discussed the theoretical implications related to this guide, but since here you're addressing a more practical concern, I'll do the same. In-game, a person has no real use for "How much DPS can I get out of a Night Lord with 2.5 billion mesos worth of equips, as compared to a Hero with less than 1 billion or an axe Hero with less than 0.3 billion?" People instead have a use for "I've got 1 billion to fund a character, and I have a plan that will make 1 billion more by level 160. What can I get out of each job with that much in spending?"

Although, I should really comment that even my method of giving each job equal spending is not enough to compare with reality, it's just the best approximation that we can in fact control. In reality one would also have to consider the class's money-making capabilities (which would be a minimal factor at best, definitely not enough for one job to have 1-2 billion, or even 10% of that, more than another) as well as the cost of their 4th job skills (for which, like ZHelms, I don't really like including the cost of them in comparisons, as I prefer the scenario where players earn those themselves).

Comparing classes based on equal scrolling is useful, because it displays inherent differences between classes and helps you to understand how strong each class is, based on the way they're designed.

In theory, maybe this is true, but again, who does it help? The way the community is set up currently, the inherent differences are irrelevant because they're too influenced by outside factors.

Additionally, I think you should stop confusing "strong class" with "high DPS." DPS is a major factor in class comparisons but in no RPG is it ever the only one. In practice there's never a strongest class, only a class that's best suited to certain roles. If there is a "stronger class" (i.e. one job can do literally everything that another job can, plus they have at least one more advantage) then it's not a RPG you're playing. But as it happens, MS is one, or at least it was designed to be before most of the community screwed up the role-playing side to it.

Heroes do more 1v1 damage than Shadowers on neutral monsters. End of story.

That's an example that's rather extreme, Heroes being among the top of all in DPS terms (along with Bowmasters, Paladins vs. some elements, maybe one of the Pirate jobs, and... actually, that's it when soloing) with Shadowers having mediocre DPS (since that's not really what their job is built for). It would impact however the degree of difference between the figures, as well as the results of the two when close together. A better example would be if the two jobs are relatively close together, close enough for "equal funding vs. assuming the same scrolling is viable" to affect the outcome.

For instance, when I put together my 3rd job DPS comparison (which admittedly contained a few errors on my part, that I'd fix if I hadn't quit SW and Basil for good) I found that the Crusader and Hermit figures, with Warrior Elixirs, were very close together, virtually tied, when I just gave them equally scrolled equips, the Hermit having a slight advantage (which also assumed the immediate use of Panic when available, which a smart Crusader would do on most bosses). But when I threw in equalizing spending, it showed some disparity - the Crusader rose to around 26k to the Hermit's 24k, and that was with low funding. (With more funding given to both jobs, the Crusader would have risen faster since the original didn't have an attack shield or Ribgol, which would give the Crusader potential to upgrade by much more attack per meso than the Hermit.)

This is the way the game is balanced. It doesn't matter if the results don't match up to what you may see in-game.

But it does matter to get your results as close as possible to the reality of what would happen in game, so that people aren't tricked into believing these results are what they'll get. Wouldn't you agree?

They never will. There are too many variables to consider, people have extremely varying amounts of funding,

All that really says is "I'd rather do what's convenient than what's accurate." The fact is, the accuracy of your results relies in part on how much work is put into them. Ensuring that the other variables are kept neutral, such as funding, might take more work, but it also gives a more relevant outcome.

RE funding: That's why you set it at a certain amount, instead of differing amounts with each job. It simply doesn't do to compare a godly funded character from one job to a weak character from another. Would you compare FangBlade to a 100 base dex Hero with a clean HG and use that to compare the jobs? No, you'd take the two jobs with equivalent builds and with equivalent funding. As for ever-changing scroll costs, that's easily addressed: whenever there's a major shift in costs that would make a significant difference, such as 30% bow scrolls inflating by 500% or GM scrolls coming out for 1h swords, you edit part of the comparison to reflect that. Did I have a stickied comparison, or one to which people frequently referred, that's exactly what I'd do.

have different levels of success scrolling their equips, Marksmen don't use Snipe exactly every x seconds, BMs get slowed down ever so much more slightly from pausing Hurricane than other classes,

1. Do people still self-scroll their equips in GMS? What I've seen of prescrolled equip prices compared to scroll prices, at least according to Basil prices as well as FM prices I've been given for Bera, suggests that it's not worth it to selfscroll much of anything anymore. It was when I quit, but it's not now.

2. Regarding BM: Not necessarily. That all depends on the individual. Warriors for instance, it can be argued, get slowed down more than them if they don't have Power Stance and get knocked back, or if they get Dispelled. Night Lord DPS gets decreased significantly more than the others by each point of defense on the enemy. And so on.

DrKs can't always Berserk at every boss, some bosses require some classes to constantly adjust themselves or move to avoid attacks while other classes stand and spam, and so on.

I think people are smart enough to realize that at those bosses, DPS itself is not necessarily the sole factor in how fast or how efficiently a job can solo them. That's why DPS isn't used as the sole measure of a "good" or "bad" job (of which there never really is such a thing to any true RPGer). DPS, rather, is used partially as a theoretical measure of a character's boss-soloing potential (when combined with their tactics) and partially as a measure of a character's averaged kill rate in training (again when paired with the suitability of the character to the map and possible "overkill" damage).

What you think is "balancing" the classes is merely introducing yet another variable into the system, making the results even more diluted.

Actually, you're wrong here. The unequal funding is itself a variable, because you're setting it up so that the characters' funding affects the outcome.

What's wrong with using Khanjars and Ilbis? I consider shields and ammo to be a second part of the weapon. The level 110 weapon set for Corsairs is a Dragonfire Revolver and Eternal bullets. The level 110 weapon set for NLs is a Dragon Sleve and Ilbis. Nothing wrong there. Balanced Furies would be like archers using attack arrows, but Ilbis is perfectly fine.

Those are just fine now that you've fixed the attack shield issue, but in your original comparison you tried simultaneously to justify using those bits of ammo and khanjars at the same time as you tried to justify not using an attack shield for warriors. In fact, as I recall it you tried to claim that warriors should just have a normal speed 2h and a 1h without attack shield that cost significantly less than the 2h, and claimed it was because the attack shields "unbalanced" the classes. That's pretty much what I'd call planned crippling of warriors. As it stands your setup for the Hero and Paladin is still probably biased against them - any viable Hero or Paladin comparison would definitely use at the very least a NRW, a Ribgol or a Red Katana + att shield instead of the much weaker (in DPS terms) Dragon weapons (with the sole exception being "with SI"). Optimally they would have a Lunchbox, but that was an event item that most players wouldn't have access to.

I skimmed through most of it. It looks like we just have two different goals when it comes to DPS comparisons. I'm trying to show how strong classes are based on the way the game is set up. I am completely avoiding any ties to the nature of the market in GMS or any other version, or the relative populations of classes. I only intend to compare classes based on the numbers the designers inputted into the game, to compare how they balanced things.

Those are just fine now that you've fixed the attack shield issue, but in your original comparison you tried simultaneously to justify using those bits of ammo and khanjars at the same time as you tried to justify not using an attack shield for warriors. In fact, as I recall it you tried to claim that warriors should just have a normal speed 2h and a 1h without attack shield that cost significantly less than the 2h, and claimed it was because the attack shields "unbalanced" the classes. That's pretty much what I'd call planned crippling of warriors. As it stands your setup for the Hero and Paladin is still probably biased against them - any viable Hero or Paladin comparison would definitely use at the very least a NRW, a Ribgol or a Red Katana + att shield instead of the much weaker (in DPS terms) Dragon weapons (with the sole exception being "with SI"). Optimally they would have a Lunchbox, but that was an event item that most players wouldn't have access to.

Uh, no. The Crossheider has been there from the beginning. We didn't actually have Crossheiders when I first made the comparison, but we do now, so I'm using it. The phys attack table has always shown the 1h weapon users to have more weapon attack than the 2h weapon users.

I believe you're referring to a post where I said I wouldn't use Maple Shields, because they were an event item. That's something I'm still not sure if it's fair to include.

... Well, I still think that if your goal is what you say (and DPS by itself doesn't really show how "strong" a class is, especially not single-target DPS and especially not with SE or SI) then the title should probably be changed to something with the word "theoretical" or "intended" in it. Otherwise, people might take this as showing the jobs' DPS accurately within the game as it currently stands, with the practical implications therein.

Personally, I've playtested all the jobs (except Pirates) on those "alternate" servers and am using one to make a video showing all 13 in action (due in late May for 3rd jobs and mid June for 4th jobs) and I've found that even with the equal scroll costs they have (usually either 1 meso or set to some x amount of mesos depending on the scroll's percent chance of working), the strongest jobs in any given situation aren't necessarily the ones with the most DPS.

If you'd like to have a further discussion, I'd be happy to talk on AIM sometime. It can be found on my now-defunct Basil profile's ID. I could use someone to explain how to play and build good Pirates anyway.

I believe you're referring to a post where I said I wouldn't use Maple Shields, because they were an event item. That's something I'm still not sure if it's fair to include.

its fair to include maple shields once you up the funding level enough. when 6 att costs more than 500-700m on a cape/shoe/weapon/glove, itll be better to switch up that 14 att warrior shield to a 20 att maple. or something like that. i think the difference is about 100m per attack on shields from 15-20 att. just a question of funding, once again. i guess thats partially senyains point.

... Well, I still think that if your goal is what you say (and DPS by itself doesn't really show how "strong" a class is, especially not single-target DPS and especially not with SE or SI) then the title should probably be changed to something with the word "theoretical" or "intended" in it. Otherwise, people might take this as showing the jobs' DPS accurately within the game as it currently stands, with the practical implications therein.

Personally, I've playtested all the jobs (except Pirates) on those "alternate" servers and am using one to make a video showing all 13 in action (due in late May for 3rd jobs and mid June for 4th jobs) and I've found that even with the equal scroll costs they have (usually either 1 meso or set to some x amount of mesos depending on the scroll's percent chance of working), the strongest jobs in any given situation aren't necessarily the ones with the most DPS.

If you'd like to have a further discussion, I'd be happy to talk on AIM sometime. It can be found on my now-defunct Basil profile's ID. I could use someone to explain how to play and build good Pirates anyway.

It was never meant to be practical. This is DPS. It's not my fault if people take this data literally to mean "which class is the strongest." I even put disclaimers up saying the exact opposite. I feel that my spreadsheet is a pretty good guideline to help people understand how classes are balanced, and that's all that it's good for.

^ Yeah, that's it!

... ok, fine, I simply hadn't noticed that the Crossheider actually had attack points on it. Although, it's still very far from any setup I'd call adequately representing Heroes or Paladins. Even a weak Ribgol/NRW (around 115) or decent RK setup (around 106 + 15 att shield) would outperform the ones on here, except with SI. If this is going to be cost-neutral it might as well be all the way cost-neutral, not only when being cost-neutral favors Thief jobs.

By the way, modular, I never got to thank you about those Combo tables. They came in handy for several new ideas and guides I put up on Basil before I quit this kind of thing. I heard the 21 Combo, 30 Coma build is starting to gain some attention in GMS. I tested and compared them myself (albeit on a server where leveling from 80 to 90 only involves maybe 2 hours of play) and found the new build far, far preferable to the old.

So yeah, updated for latest skill changes.

Let me know if I need to make any changes to Buccaneers due to reduced delays or w/e.

I have a request but i'm not sure if it deserves its own special thread or not.

Is it possibly to do a DPS calculation along the lines of

"What damage range does a class using a 10 minute apple + 5 minute cider out DPS an NL using a 15 minute apple"

I'm not sure if it'll be a constant value i.e.

- The damage range must be +1000 higher than the NL
- The damage range must be x1.2 higher than the NL

And this would compare all classes (Except mages I guess) vs. NLs.

Reason: Me and another person are getting tired of always being shafted against apples during boss runs. It's really annoying when a 12x/13x/14x NL gets an apple over a 16x corsair or a 18x DrK. We know we out DPS them during the 10min apple duration, and we're assuming the damage we put out during the 10minutes makes up for the 5minutes of alchemist NLs have. I'm just seeing if that really is the case and at what point (Damage range) it is true.

It'll also be a fun side project for you :x

-text-

Easy enough. Compare your damage with a 73.3 att pot and the NL's damage with a 100 att pot. If your damage is higher, you win.

It's pretty unlikely that a NL can make up 40 levels of AP with the additional attack.

I was trying to edit the stats in the 'Base Dmg-Phys' sheet in the latest version (1.5) and all those values are locked now.

I wanted to see what the DPS would come out to for a dexless NL since the defaults for NL are set to VERY high dex, of which, my friend pointed out, very few high levels in maple global would actually have.

I was trying to edit the stats in the 'Base Dmg-Phys' sheet in the latest version (1.5) and all those values are locked now.

I wanted to see what the DPS would come out to for a dexless NL since the defaults for NL are set to VERY high dex, of which, my friend pointed out, very few high levels in maple global would actually have.

Few Sleve-using NLs would have 150 dex before MW? A well-scolled Sleve > dexless claws.

Edit: Btw, I did the calculations with MW 30 under the assumption that it would come out soon. Also, I decided not to use MW to equip weapons. That's why the NL has like 16x buffed dex. The NL in the calculation uses a level 100 overall scrolled for LUK and LUK earrings instead of dex. I realize that NLs would be more likely to be using dex equips and there's always the "but lower secondary stats = more damage with MW!" thing, but it's really a small difference and I wanted to be consistent with my calculations.

I could re-add the stat-overriding function, though. I'll get around to it.

Although, I should really comment that even my method of giving each job equal spending is not enough to compare with reality, it's just the best approximation that we can in fact control. In reality one would also have to consider the class's money-making capabilities (which would be a minimal factor at best, definitely not enough for one job to have 1-2 billion, or even 10% of that, more than another) as well as the cost of their 4th job skills (for which, like ZHelms, I don't really like including the cost of them in comparisons, as I prefer the scenario where players earn those themselves).

It's not a minimal factor like you make it out to be... I'm probably comparing best vs. near worst 3rd jobs (WK and Hermit) but I found there to be a massive difference in money-making. As in the difference between 500k a level and 8-10mil a level. As most merchants know, it takes money to make money - an extra 20 million every couple levels is a very healthy stimulus.

And with the difference in income, there's a difference in equips. I'll use relatively current prices for the gear I own because it's not fair to compare purchase price.

Level 117, White Knight vs. Level 117, Hermit
Weapon: 10mil Morningstar, 300mil Casters + 95mil Throwing Stars
Overall: 40mil Battle Empress, 60mil Avenger
Helmet: Zakum Helmet, Zakum Helmet +1 70% dex (15mil)
Glove: 35mil Imperial, 160mil Stormcasters
Cape: 10mil Icarus 2 + 8mil Icarus 2 (str & dex), 55mil Pink Adventurer + 80mil Pink Adventurer (luk & dex)
Shoe: 15mil Dark War Greave, 60mil Dark Moss
Earring: none, 95mil Gold Drop
Face: none, 100mil Rat Nose

Total: 118mil vs. 1020mil


Of course, because I'm sensible, my Paladin uses a lot of this gear now... so I can't really unlink their costs anymore. It's not really the ideal spending pattern, but I was working within the market in both cases (availability, especially on BWs and female overalls, is poor - I upgrade what I can when I can).

Oops... I should apologize for Stereo's semi-necro-reply. I thought all those posts (Dusk/Senyain) were new because I saw May 6 and thought it was June 6, and hadn't seen them before. And mentioned on IRC that I was going to read them.

Request: Can you do the calculations for barr + DU instead of barr + DS? The delay reductions lets you use barrage right after DU and DU is way faster than DS, so i want to know which combo is stronger.

What about Barrage to Energy Orb? Though Energy Orb wouldn't be always on.

Request: Can you do the calculations for barr + DU instead of barr + DS? The delay reductions lets you use barrage right after DU and DU is way faster than DS, so i want to know which combo is stronger.

Barrage + DS is stronger. Don't know if you recall, but Barrage+DU+FF was in the original sheet. The delay reductions don't make it much stronger since FF is almost as strong as DU.

Barrage + Orb would require me to know how fast Orb is o-o