Back from my ban finally. Someone got a little salty about my honesty I guess.Quote:
Originally Posted by Shurrikhan
More to the point, no, the whole of FFlogs is poorly representative. How many people upload on FFlogs vs. the actual number who play the game? A fraction of a fraction. The people who are going to upload on FFlogs will do so either purely or partly because of self-interest; to one degree or another, they want to flex their e-peen. There might be some altruism there in a general "let me contribute data for a better understanding of the game," and certainly some personal betterment, but ultimately at least part of it comes down to someone bragging.
And you know what? That's fine. The problem comes when people begin to use it as justification for whatever thing it is they want changed or implemented. Even worse when people begin to treat FFlogs like it's some kind of job balance bible, as if the only thing that matters is what the pretty graphs on FFlogs say. We're fast approaching that point, where people use FFlogs as a tool to push their own agenda, an agenda which almost always involves some sort of "moar dps" goal and to hell with anything else. I only hope SE's wisdom of avoiding any kind of outside element like this continues to hold, as I've seen too many good games die when the "paint by numbers" crowd gets sway.
-
11-06-2019, 07:59 PMQuor
-
11-06-2019, 08:37 PMAkiudosee, funnily enough outside of comparing my own e-peen to others, not necessarely flexing but honestly seeing how i fare the "agenda" people have generally is way less "moar dps" and more "smoother, faster runs", its just that "moar dps" is the answer to how to reach that. having more damage on your side makes fights an incredible lot easier, and that is not a shortcoming of final fantasy, its just how it is, adds die faster, mechanics can be skipped (even if its just not dealing with the very very last mechanic of a fight) and hiccups can be better compensated cause you got a buffer.Quote:
Originally Posted by Quor
also yes, fflogs represents a fraction of a fraction, you are absolutely correct in that assesment, however, given that its not necessarely you yourself that has to upload a run chances are if you do any kind of raid you will be on fflogs, what people here are advocating for, no matter what their opinion is on particular classes or numbers or whatever is for RAID balance, and while i do believe you allways should keep the whole game in your mind while actually balancing it one has to accept that raids are the content were balance actually matters the most, pvp gets balanced differently alltogether, dungeons simply aren't tight enough on any kind of demands as to make it important how good or bad a class is, neither is leveling, and things like heaven on high in fact actually are basically whats the least affected by dps in the sense that whats important there are completly different things than how much damage you deal so it makes sense to talk about numbers if you want to talk raid balance, because thats where they matter.
and again, if you raid in any kind of static, 95% of the time they will upload to fflogs, if you do raids through partyfinder ? you cant even decide if it gets uploaded as you have no control over all the random people with you, the person that has cleared every boss more than once and is NOT on fflogs is in fact the outlier. -
11-06-2019, 08:48 PMShurrikhanUnless there is a specific reason for those who volunteer their parses or happen even just to be in the party of someone who has to be consistently distinct in numerical fashion from those who do not, the sheer size of the sample is going to mean more than its relative percentage when evaluating consistency, likely ease of play, and likely performance. FFlogs has 72574 from Eden Savage alone in just the last week. That is not an insignificant sample size.Quote:
Originally Posted by Quor
And veteran players on averaged very good but imperfect run are exactly what should be balancing jobs around, not the lowest denominator.
Yes, I'm sure the guys struggling step by step from a 20th percentile to a 45th, asking friends for advice and practicing on dummies between each, are doing so just so the world can see them stroke their massive, below-average parses. Couldn't possibly be for self-improvement. Nope, not a chance.Quote:
Originally Posted by Quor
Imagine that? When over a hundred thousand runs by, as you put it, people more than likely interested in performing well for said runs, contribute data towards their performances at various levels of effort and optimizations -- people might notice where a particular job is lacking and seek to remedy that imbalance. The horror!Quote:
Originally Posted by Quor
The only agenda to be found as a result of measures of performance and imbalance therein is improved performance and, again, balance therein. These aren't grand political crusades. They're data. And I should hope people pushing for a balance change would consult as much data as they can. Theory-crafting is one thing; seeing how well players can actually manage that would-be optimal output is another. Both are useful. Which makes me wonder, which are you suggesting -- that we center our discussions of balance solely on the data of perfect theoretical play, or that we purposely avoid data altogether? I cannot imagine a productive result from either. -
11-06-2019, 08:56 PMShurrikhanWhat? Seriously... what?Quote:
Originally Posted by FallenWings
Do people still not understand what rDPS is? I mean, it's gotta be that or some hidden warrant telling us the 6% lead AST holds over WHM and SCH should being treated with no more concern than the 0.8% lead WHM holds over SCH... Or perhaps this "ideal world" just really is just one founded on the forcible removal of BLM, SAM, and MCH, and weakening of MNK, SMN, and RDM, all just to elevate DRG, NIN, DNC, and BRD? Are we really just looking for a far fewer competitive compositions than we had in Stormblood?
Contrary to what fflogs metrics may say, every raid-buff goes both ways. Not only does someone have to provide it, but someone also has to capitalize on it. The more personal dps you have, and the more control over its burst, the more you can do with those raid buffs. To say that unless you're throwing out rDPS abilities on CD you must somehow be contributing less effort than the people who bank their resources or vary their rotations beforehand to exploit those CDs is ridiculous. Both are compositionally dependent coordinated efforts.
Yes, perhaps ideally there should always be at least one or two buffers in a given party, but the ability to exploit each is every bit as essential. In some cases, where the buffers also have absurdly strong burst, the two will overlap and form a meta comp, but short of that an ideal balance will include a balance between or down a spectrum of buffers and exploiters at opposite extremes. -
11-06-2019, 09:25 PMAkiudoyes but as raidbuffs depend on the level of the whole group, not just your own raising just your own level will in fact result in less of a "win" to the group so if even at the absolute very very top dpS (again for emphasis, talking purely dps buffs here) don't contribute more raid dps than they cost you in personal dps that means they will in fact in weaker groups mean classes with raidbuffs get punished for no pay off.Quote:
Originally Posted by Shurrikhan
Imagine 2 totally equal players playing the same classes which are perfectly equal in dps. if you now take 1000 dps off from one of these classes but in exchange give it a raidbuff you now end up in a situation where either in the very best groups the one buffing the group will have to be at least slightly higher, or the class buffing will in fact be ALLWAYS lower, that defeats the purpose of offering a dps buff, a dps buff is in fact a detriment if it doesn't even offer an advantage at the absolute top end, mind you ast being 6% above whm is bad and was definitely overdone, but the reason for that isn't that at the absolute top ast is stronger, its that its too much stronger, if the difference at the top where half as big as it is than whm would be stronger until around 80% and than very slowly get weaker than ast by a negliable amount.
saying "a class with raidbuffs at the absolute realistic top should be equal to a class with no dps buffs" is equivalent to saying "a class offering dps buffs should allways be weaker than a class offering no buffs unless played in one of the very best groups, than they are allowed equality", its like saying its fine if you earn 20% less because of your nationality if you are a basic worker as long as your pay is equal if you got to be ceo
edit: just to clarify two things, a) either way in an ideal case the dps differences between classes move in a region where it doesn't really matter what dps you use (think tank balance, who really cares which tank does the most dps?) (this is not a general statement, only concerning the "2 equal dps , one selfish one has a dps buff" situation)
and b) it really doesn't matter if blm, sam or on the other hand nin is on top by 100 dps, if whm or ast bring 5% more dps, there will allways be a meta, unless absolute perfect balance is achieved, and thats a pretty lofty goal. in that regard it isn't any better or worse if at the end of the day blm and samurai or summoner (utility buffs are a different story, strictly concerning the fact smn has a dps buff, blm is selfish here) and ninja offer the most dps or if you are better of paring them with a mch or a brd, as such it is futile to break ones mind over the possibility of a certain class ending up slightly stronger than others, selfish or not, the best we (or rather the developers) can aim for is "so close 99% of the playerbase won't give a damn) -
11-06-2019, 11:20 PMCKNovelNot exactly, they indeed "give" the most rDPS, that's the point.Quote:
Originally Posted by FallenWings
rDPS is simply your personnal DPS to which you add the DPS given to teammates and minus the DPS taken from raidbuff that aren't yours.
Some jobs eats more raidbuff than the others, tanks are a good example: Gunbreaker benefits from all buff, Warriors doesn't benefits from Crit&DH buff, PLD&DRK don't fully benefit from the physical buff.
Meaning the Gunbreaker rDPS given is the most negative since the job takes the most out of the raidbuff but gives nothing in return.
In an ideal world, all DPS jobs rDPS should be within a 500 dps range, jobs with no support should profits from all raidbuff and support not profit a lot. If all the support jobs would have higher rDPS the community would go back to Stormblood mentality where SAM and BLM suck because they don't bring anything to the party, but now with MCH. You could also have double ranged (DNC&BRD) composition without any mages since they bring little.
I think we all remember the end of HW where rDPS was everything. Trick Attack, Hypercharge, Requiem and Battle Litany. -
11-06-2019, 11:39 PMAkiudobut raid dps is still everything, this also doesn't change either way no matter if brd, nin or blm is in front, it just means right now you take the samurai instead of the nin because he in fact adds more raid dps through his superior personal dps alone.Quote:
Originally Posted by CKNovel
and jobs with no support are in fact the one that profit the most from raidbuffs, thats why the dancer will buff the samurai and not the redmage, just that in fact these dps buffs than get counted for the one adding the buff, not the one doing the damage. you say "rdps was everything, trick attack, hypercharge, requiem...." like thats somehow worse than "lets all remember the time where when you wanted the highest raid dps you stacked the classes that in fact were the most self sufficient".
the problem at the end of heavensward was that brd and mch indeed were too strong (because they basically only got buffed over and over to beat out each other) and synergized way to much with dragoon, the only reason raidbuffs are ever worth it is if at the least at the top they offer somewhat of a group dps increase, and at the end of the day it isn't any worse if in a 99% group the group does 150 dps more if the take ninja instead of samurai instead of the other way around, the one thing thats really important is that the discrepancies between both cases are close enough it doesn't matter unless you are a total "going for fastest speedrun on every single eden turn" kind of person, its just that entering the end of heavensward square completly failed at getting the "close enough" part right. -
11-07-2019, 02:02 AMWrenElessedilIt sounds like you don't understand rDPS. If a job whose only utility is damage buffs is not providing more rDPS to the party than a selfish DPS, then you should always take the selfish DPS. Selfish DPS should be highest in aDPS, while support DPS should be highest in rDPS (before accounting for non-damage support abilities). Ideally it would be, in terms of rDPS provided, badly timed support < selfish DPS < well timed support, so a selfish job provides slightly more rDPS in the worst case and a bit less in the best case.Quote:
Originally Posted by Shurrikhan
-
11-07-2019, 02:38 AMForteNightshadeThis is irrelevant. When enough data is compiled, it will maintain a statistical average regardless. You don't need one million uploads to determine the DPS output of Dragoon. A few thousand is all it takes provided each job has a comparable amount themselves—which they do. Going from 4,000 Dragoon parses to 20,000 won't impact the overall average. In order for such an anomaly to occur, it'd require an obscene degree of coordinated sabotage and the creator of FFlogs to not notice a bombardment of parses all from the same handful of statics, all with suspiciously bad parses in the same relative time frame. Basically, it'll never happen.Quote:
Originally Posted by Quor
Does that mean FFlogs is 100% accurate. Far from it. Some assumptions have to be made simply because it and/or ACT can't properly track everything. That said, FFlogs still remains a good source of information. Calling it a mistake is just welcoming blissful ignorance. -
11-07-2019, 04:18 AMDogempire
The only argument you can really make about FFLogs being inaccurate is that most people are going off the average dps of people who do savage, and that people who do savage do much higher dps than the average population.
Though there's also the fact that gear is basically normalized a couple months into the tier since most people will have BiS or at least very close to it, so it ends up evening out regardless.