4th job Single Target DPS

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.