rogues, my 2cp

robopirateninja said:
My complex is that a massive change was made to a class on the basis of flawed parses that ignored the abilities of the classes they were supposed to balance. I'm not trying to attack you at all, but to make an accurate and fair change to gameplay accurate and extensive logs ARE required. Yes, it will be time-consuming and annoying and it may not be complete this week or next week or even in two months. I appreciate the time you have put in already, and gearing out all of those characters to appropriately represent different tier levels must have been another headache in itself.

If you want volunteers to help with this to save you some time I'm sure there are many players who would be willing to do some parses, and with adequate instruction on what to do and when they should be accurate, and cut down the turnaround time.

I don't think I claimed anywhere that I was an expert on melee classes; I main a magician and a wizard. I don't think it's necessary that you consult me when parsing classes I don't play. I do think that without adequate control over experimental variables you will see flawed results.
Click to expand...
I really don't see how they are flawed, other than the Ranger-COTP issue.

In my opinion, stances don't play a role in base DPS. About 90% of the time I am in Aggressive. The things that Allie picked up were indeed, the base DPS of the classes. A bot sitting there with auto attack on a monk can pick up the base DPS of the char. A lack of player input shows the base DPS of a class. Only when you put in the player input do you see the DPS potential of the classes, and thats what was shown when the players(me included) sent in parses from raids.

I'm pretty sure the numbers posted in the thread were the raw DPS numbers, while at their disposal was a huge amount of raid parses from players showing the same results.
 
Brimztone said:
I'm pretty sure the numbers posted in the thread were the raw DPS numbers, while at their disposal was a huge amount of raid parses from players showing the same results.
Click to expand...

Just curious, do you parse yourself only or multiple people in your raid? If you parse multiple people do you have any logs you could post comparing full on raid dps?
 
Widan said:
Just curious, do you parse yourself only or multiple people in your raid? If you parse multiple people do you have any logs you could post comparing full on raid dps?
Click to expand...
I parse the whole raid every night.

The only problem, is I don't pick up caster DPS properly. I can however, post numbers of Brandar and I on nameds pre and post nerf. I just have to dig up the logs from my Game folder and type the numbers out in the post.

EDIT:It may be a while before I post them since I have a few other things to do today.
 
The real problem with using player parsing is that it introduces a LOT of variables, which is undesirable for a study such as this. Different players distibute their AAs in different ways, favor different gear, use their styles differently. As Brimz said, different raids will also do things differently.

When I get a parse from a raid, I don't know who was semi-afk for what fights, I don't know who was boxing (and therefore wouldnt' have been playing their character to the best of their ability), I don't know who had giantkin and who had savagery and what the bard was playing and who got an ultimate blast and who was undergeared and who had picked up a weapon far above their tier from raiding with a different force once-upon-a-time.

That doesn't mean these parses are useless. They can serve as a General guide to ensure that my results reflect the general populace to a first approximation. But I didn't feel I could rely on them to give me the extremely accurate numbers that I was interested in, here. Accurate results demanded that I have very comparable characters.

As I said before, the only flaw I made was using cotp on the ranger. How much did this inflate base ranger dps? 2%? 10%? Individual rangers may estimate it, but I still have the characters and I can redo even just three parses to have a general idea of what that might mean applied across the board to all the ranger parses I did. Rest assured that it is on my list of things to do and will get done.

If anybody is angry that rangers might have gotten a 5% overnerf than they deserved due to this flaw and that I'm not working fast enough to correct it, I'm sorry. But I would still maintain that it's better than leaving them at the 20% or so overpowered that they were.
 
Brimztone said:
The Ranger stance however, does not drain stamina(AFAIK.). It also is (half)useless with AAs like Ambidexterity.
Click to expand...

It does, and it is more or less made obsolete by AAs (ambidex and the double attack w/main hand one). Even at 65 with druid regen and EoT it only lasts about 20-30 seconds and with those AAs it gives you one extra attack per "round" (offhand double).
 
Llanoldar Lluindar said:
It does, and it is more or less made obsolete by AAs (ambidex and the double attack w/main hand one). Even at 65 with druid regen and EoT it only lasts about 20-30 seconds and with those AAs it gives you one extra attack per "round" (offhand double).
Click to expand...
Oooh. Sorry about that mistake.
 
Allielyn said:
The mob was too high for cripples, yes, so currently I'd guess rangers still have a nice advantage there. It just won't show on raid trash/bosses.

Characters were buffed with most major long term buffs, but without styles, short term atk clickies, or short term buffs such as savagery and giantkin. The goal was not to parse MAX dps per class, but comparable dps per class.

Also, thanks to everyone who sent me their own parses from raids, etc. It helped work as a check that what I was doing was comparable to the real-life (lol?) application.
Click to expand...

i understand that your gathering base melee dps... but shouldnt savagery be counted as beastlord dps at the least? beastlords being one of the worst melee dps having it on themselves would atleast give a slight idea of what it could add. my reasoning behind this is because rogs get nothing to boost the attacks of others.. they have styles just like everyone else. this is also true for monks, who i believe are balanced a bit with feigned death skill.... but i would imagine such spells are savagery would play a factor in beastlords dps scale.. i dont want to see nerfs(or as some call them "fixes) going all over... i just dont want to be a mediocre dpser compared to others when playing my rog. also kinda curious on the only 20-30 dps gained from 250 aas + a tier or 2+ of raid gear=/
 
No, because I wasn't parsing SOLO dps (self buffs only).

In fact, on most raids, the beastlord doesn't savagery himself - he sticks it on the rogue or the monk. In this case, you could call it Beastlord dps, but only in terms of beastlord utility dps. CoTP follows these same rules and should have been used for either all the parses or none. The only buff differences between all the parses should have been self-ONLY buffs, which, truth be told, don't add a huge amount of DPS except in the case of the beastlord's Pet.

In regards to Beastlord dps, yes I agree that the increase in beastlord melee dps between the 250 and 400 AA sets was mediocre at best. We talked about it in staff chat, and some tweaks have been made. I have not yet been able to reparse those changes.
 
Allielyn said:
No, because I wasn't parsing SOLO dps (self buffs only).

In fact, on most raids, the beastlord doesn't savagery himself - he sticks it on the rogue or the monk. In this case, you could call it Beastlord dps, but only in terms of beastlord utility dps. CoTP follows these same rules and should have been used for either all the parses or none. The only buff differences between all the parses should have been self-ONLY buffs, which, truth be told, don't add a huge amount of DPS except in the case of the beastlord's Pet.

In regards to Beastlord dps, yes I agree that the increase in beastlord melee dps between the 250 and 400 AA sets was mediocre at best. We talked about it in staff chat, and some tweaks have been made. I have not yet been able to reparse those changes.
Click to expand...

i suggested using the savagery on the beastlord because they have one of the lowest dps's. generally a beastlord will keep 1 target savaged.. at times 2.. i figure if u just keep it to 1 on the parsing, and on the blord... thatd factor it in at a LEAST amount. i am not sure if utility dps is counted... but if its not i dont think kicks and backstabs should be factored in.. because for monks and rogues that is their utility dps... not to mention there are a few mobs where backstabbing isnt an option. same goes with certain other things.. but i dont know where savagery cant play a factor in dps. or cunning.

also i really woulda like to hear something about rogues possibley having a stance to avoid rampage/ww or an innate higher chance to avoid or something. currently rogues very little assistance on these mobs, where monks can pull them, rangers can bow them, beastlords and savage rangers and cunning casters and whatnot...

also, if your needing someone to help with parsing id be glad to try and find a time to help with them.
 
If we were calcualating vicarious dps, bards would be close to 500!

But yeah...calculating parses is just too difficult and too lengthy to get "accurate numbers". Players would have to be playing as efficiently as possible for long, long periods of time to justify any numbers. I think the way Allielyn went about it in the first place was probably fine for getting a rough estimate for the base dps on melee classes. However, I do not believe a nerf to a class (especially that big) should have been implemented based on this small (yes, it took hours but it is still a small sample size) study. There really needs to be testing done for many months to determine if a balance is required. Logically, that is the only way to do it. Realistically, this takes way too much time and doesn't seem like it would get done accurately anyways.

The CotP mistake was a mistake that needs to be looked at, but once the new parses come back we can see how much variation it caused. Using disciplines without all spells/gear/AA/etc would just scew the results again. It may not be perfect, but I do think parsing a Steel raid with each of the "best" class doing their maximum dps on an entire raid (trash and nameds alike) and then doing this for a few months would be most accurate (even a week straight would help a lot for smoothing out the averages). Not to say that I think you should be wasting your time doing this, but I do believe a lot of experiments need to be done before any big change (see: Ranger nerf) gets put into the game.

Like Allielyn said, there are just too many factors to be looked at at the same time, so accurate parses are nigh impossible and even semi-accurate ones are difficult to obtain.

If you two-box me Tyrone, I will take your side of the argument instead...
 
Spiritplx said:
If we were calcualating vicarious dps, bards would be close to 500!

But yeah...calculating parses is just too difficult and too lengthy to get "accurate numbers". Players would have to be playing as efficiently as possible for long, long periods of time to justify any numbers. I think the way Allielyn went about it in the first place was probably fine for getting a rough estimate for the base dps on melee classes. However, I do not believe a nerf to a class (especially that big) should have been implemented based on this small (yes, it took hours but it is still a small sample size) study. There really needs to be testing done for many months to determine if a balance is required. Logically, that is the only way to do it. Realistically, this takes way too much time and doesn't seem like it would get done accurately anyways.

The CotP mistake was a mistake that needs to be looked at, but once the new parses come back we can see how much variation it caused. Using disciplines without all spells/gear/AA/etc would just scew the results again. It may not be perfect, but I do think parsing a Steel raid with each of the "best" class doing their maximum dps on an entire raid (trash and nameds alike) and then doing this for a few months would be most accurate (even a week straight would help a lot for smoothing out the averages). Not to say that I think you should be wasting your time doing this, but I do believe a lot of experiments need to be done before any big change (see: Ranger nerf) gets put into the game.

Like Allielyn said, there are just too many factors to be looked at at the same time, so accurate parses are nigh impossible and even semi-accurate ones are difficult to obtain.

If you two-box me Tyrone, I will take your side of the argument instead...
Click to expand...

looking over the numbers briefly again.. the beastlord does nearly the same dps as a ranger now... and their utility is through the roof compared to the ranger imo=/ not to mention they easily catch up to rog in dps factoring in just 2 spells=/

basically the way i see it is there was 2 ways to go.. rogs go up.. rangers and beastlords go down.. and seeing as first part of step 2 went in.. i feel beastlords will likely be watched here now=/
 
Eh, kicks and backstabs aren't "utility" dps . . . any way you look at it.

The sample sizes were not months long, but the important thing was that they had a relatively high degree of precision involved. I included the standard deviation, even. It's easy to tell what classes have "spiky" dps compared to others. It is entirely possible to get accurate parses across the board, so long as you can control exactly what level, how many AAs, how those AAs are distributed, exactly what gear, buffs, timers for hitting skills like BS and archery and flying kick, etc. Which I did. With a LOT of GM and player input at every stage of the level.

To give you an example of how much data this ends up being compiled between (and as a result, how accurate the final results are), one 10 minute fight on the 400 AA Beastlord set included over 2000 hits between the beastlord and the pet (and an additional 1200 misses).

Guys. I am a scientist by trade. I promise I know how to control variables and collect a reasonably sized sample set. :tinfoil:
 
khador said:
looking over the numbers briefly again.. the beastlord does nearly the same dps as a ranger now... and their utility is through the roof compared to the ranger imo=/ not to mention they easily catch up to rog in dps factoring in just 2 spells=/

basically the way i see it is there was 2 ways to go.. rogs go up.. rangers and beastlords go down.. and seeing as first part of step 2 went in.. i feel beastlords will likely be watched here now=/
Click to expand...

Oh, I agree that general utility should be taken into account while balancing classes. I don't really think that it should be taken into account on just the DPS factor alone, however. That is why I said that they shouldn't nerf one class just because of baes DPS logs. As a non-ranger/rogue/beastlord, I would rather see all of these other classes have their DPS increased, as this makes my raid easier. Having any class get nerfed affects me, and I am not a huge fan. Increasing rogue DPS slightly, or somehow giving them more utility would have been the way to go in my opinion. But I am not running a server and am not making the rules. I think having hide/sneak be more powerful would be a fine option as Brimz pointed out earlier that it doesn't work in many high end zones rendering it obselete in some areas.
 
Allielyn said:
Eh, kicks and backstabs aren't "utility" dps . . . any way you look at it.

The sample sizes were not months long, but the important thing was that they had a relatively high degree of precision involved. I included the standard deviation, even. It's easy to tell what classes have "spiky" dps compared to others. It is entirely possible to get accurate parses across the board, so long as you can control exactly what level, how many AAs, how those AAs are distributed, exactly what gear, buffs, timers for hitting skills like BS and archery and flying kick, etc. Which I did. With a LOT of GM and player input at every stage of the level.

To give you an example of how much data this ends up being compiled between (and as a result, how accurate the final results are), one 10 minute fight on the 400 AA Beastlord set included over 2000 hits between the beastlord and the pet (and an additional 1200 misses).

im well aware that they arent considered utility. my point is that having them closely bunched together makes the ones WITH utility sky rocket. a rog cant always use backstab... most cases he can.. but not always. on those cases he loses 25-50% of his dps(depending on if he uses his stance or has chaotic stab) and then he has to deal with ripostes and other modifiers... i am just curious if theres future plans to add some utility to rogs or if they will just be pure melee dps that can be out done by other classes... right now rogs arent bad.. they just arent as good as others imo=/ last i knew they were suppose to be good.

Guys. I am a scientist by trade. I promise I know how to control variables and collect a reasonably sized sample set. :tinfoil:
Click to expand...

im well aware that they arent really utility for dps... but with utilities others sky rocket compare to rogs... rogs dont always get the option to backstab.. where most of the time savagery and other utility things can be done anywhere.. ive yet to find a zone u cant savage in=/ currently as i see it... rogues are just mediocre.. they arent bad.. not great.. can be outdone by other classes... and nearly useless on rampage/ww mobs...
 
khador said:
basically the way i see it is there was 2 ways to go.. rogs go up.. rangers and beastlords go down.. and seeing as first part of step 2 went in.. i feel beastlords will likely be watched here now=/
Click to expand...

Not sure if you missed it, but. . .

Allielyn said:
In regards to Beastlord dps, yes I agree that the increase in beastlord melee dps between the 250 and 400 AA sets was mediocre at best. We talked about it in staff chat, and some tweaks have been made. I have not yet been able to reparse those changes.
Click to expand...

Let me say it more clearly: Beasties might have gotten a little buff out of all this parsing.
 
Allielyn said:
Not sure if you missed it, but. . .

Let me say it more clearly: Beasties might have gotten a little buff out of all this parsing.
Click to expand...

are u saying beastlords were given an increase? or are you saying they are already planning to nerf them?
 
Allielyn said:
In what way is "buff" translated as "nerf"? =P
Click to expand...

key word was planning.. as in will be happening. not to come off rude.. but that was the main word in that part of the sentence.
 
You must log in or register to reply here.
Back
Top Bottom