Dual Weilding has **NO** advantage! Funcom's MR changes ruined dual weilding!

[sbermunk]

Originally posted by Lucid Flow


How exactly does it factor damage then? If it were a straight damage reduction, I would see it. If it were an Attack Rating reduction, I would also see that.

Are you saying that there is a lower % to roll a higher number in your damage range? I don't believe it. Game mechanics don't even work that way because there would be nothing to base the % off of. In a game like Daoc it's based off of character level vs skill level. In AO there is no such calculation.

So you explain to me *how* the damage is reduced. As of now, Mikebond and Flows tests are the most accurate because as of this date, they correspond with everything I have done with myself and other chars as an example.

I'd just like to insert a comment here... Saying "I don't have a theory that explains a result, therefore the result is invalid" makes no sense. Theories must explain experimental results, you can't choose to simply ignore results that don't fit. A lot of bad science is propagated in the real world this way...

My impression is that Tetra's tests have been conducted carefully, and that with that large a sample size, it seems hard to argue that the results aren't statistically significant...

[__________Lucid Flow]

Originally posted by Xurbax
My impression is that Tetra's tests have been conducted carefully, and that with that large a sample size, it seems hard to argue that the results aren't statistically significant...

They have not been conducted carefully. Actaully, seeing as he killed various Hardeneds at various level, the AC factor was never counted. Sure, over 419 shots (which is what he listed) then yeah, maybe. I'll show you there is a clear and definate discrepency in his numbers.

Firstly, statistically speaking, given a random sample of a certain number of shots, the averages should be the same. If there was a clear difference in damage offhand, then it would show though the entire number sample. But it doesnt. Watch.

Total Average:

Dual Weild: 338.91
Single Weild: 350.34

Conclusive you say? Not really. When broken down, you see the latter part of the sample seems to randomly do a lot more damage then the former.

First 100 Shots:

Dual Weild: 350.63
Single Weild: 339.11

Last 100 Shots:

Dual Weild 313.69
Single Weild: 360.76

First 200 Shots:

Dual Weild: 351.485
Single Weild: 350.34

Last 200 Shots:

Dual Weild: 327.55
Single Weild: 355.44

First 300 Shots:

Dual Weild: 348.13
Single Weild: 343.43

Last 300 Shots:

Dual Weild: 335.94
Single Weild: 350.65

Average of the Averages:

Dual Weild: 338.04
Single Weild: 348.54


Conclusion?

Well it's simple. Given any number of tests, you could find one to be higher then the other, or another to be another. In many occasions, the dual weild damage is higher the the single weild. In this number set, the single weild is higher. But in other number sets, dual weild will come out the same.

A difference 5-15 points in any given number set of random numbers is NOT conclusive evidence. Since it is random numbers in a given range, you have the luck of the draw to give you numbers.

For example, if he did 200 shots, it would have been nearly the same average. Heck, if he did 300 shots, Dual Weild damage per hit would be higher! But he chose to do 419 shots and the last 100 shots is where he rolled many high numbers in the single weild category. The numbers are skewed as the last 100 shots he seemed to get mugh higher rolls then in the first 300 in the single weild category. In the First 100, 200 and 300 shots, Dual Weild is actually slightly higher. Near the end he got lucky.

Is 300 shots not good enough? These numbers are random. The damage is the same.

[sbermunk]

Originally posted by Lucid Flow

They have not been conducted carefully. Actaully, seeing as he killed various Hardeneds at various level, the AC factor was never counted. Sure, over 419 shots (which is what he listed) then yeah, maybe. I'll show you there is a clear and definate discrepency in his numbers.

The mob AC point has some validity. Over a large enough sample of mobs tested, the AC factor should balance out, but I don't recall how many mobs were killed in the tests...

Firstly, statistically speaking, given a random sample of a certain number of shots, the averages should be the same. If there was a clear difference in damage offhand, then it would show though the entire number sample. But it doesnt. Watch.

The averages of a smaller subset of the samples won't necessarily be the same as the overall average. What we are really trying to determine is what the value converges on over time (i.e. an infinite number of samples).

Conclusive you say? Not really. When broken down, you see the latter part of the sample seems to randomly do a lot more damage then the former.

I don't think I used the word "conclusive"... I would instead say "persuasive".

Well it's simple. Given any number of tests, you could find one to be higher then the other, or another to be another. In many occasions, the dual weild damage is higher the the single weild. In this number set, the single weild is higher. But in other number sets, dual weild will come out the same.

A difference 5-15 points in any given number set of random numbers is NOT conclusive evidence. Since it is random numbers in a given range, you have the luck of the draw to give you numbers.

For example, if he did 200 shots, it would have been nearly the same average. Heck, if he did 300 shots, Dual Weild damage per hit would be higher! But he chose to do 419 shots and the last 100 shots is where he rolled many high numbers in the single weild category. The numbers are skewed as the last 100 shots he seemed to get mugh higher rolls then in the first 300 in the single weild category. In the First 100, 200 and 300 shots, Dual Weild is actually slightly higher. Near the end he got lucky.

Is 300 shots not good enough? These numbers are random. The damage is the same.

Again, there will always be variations in smaller sample sets.
That is why the more samples, the better, if we are trying to find out which (dual/single) is higher over a long period of time. If I flip a coin many times, I may have runs of 'heads' or 'tails', but if I keep a running tally, the average number of each will converge on 50%/50%. (Ok, really I should say 'probably will converge'.) That is what we are trying to determine - the long-term average.

On that basis, your arguments above don't make much sense.
You are saying "I can pick any subset of the samples to prove that either is better than the other, so therefore dual-wield is better". I can't come up with any other way to interpret what you wrote above...

You could argue that maybe the sample size needs to be bigger considering the small difference between the two averages - but then the point has clearly been made that dual-wield doesn't seem to carry enough of an advantage over single-wield to be worth it, so arguing any further seems rather pointless...

[___________LostLogic]

IMHO, dual wielding pistols of equal QL should result in 200% damage dealt. I mean, whats the point of wasting all those darn IP otherwise?

Am I the only one who thinks this way?

[____________Asteroth]

Just make sure ur off hand weapon is faster than ur main weapon.

ELSE

U are in for some surprises.

[__________Lucid Flow]

Originally posted by Xurbax
The averages of a smaller subset of the samples won't necessarily be the same as the overall average. What we are really trying to determine is what the value converges on over time (i.e. an infinite number of samples).

I don't think I used the word "conclusive"... I would instead say "persuasive".

Again, there will always be variations in smaller sample sets.
That is why the more samples, the better, if we are trying to find out which (dual/single) is higher over a long period of time. If I flip a coin many times, I may have runs of 'heads' or 'tails', but if I keep a running tally, the average number of each will converge on 50%/50%. (Ok, really I should say 'probably will converge'.) That is what we are trying to determine - the long-term average.

On that basis, your arguments above don't make much sense.
You are saying "I can pick any subset of the samples to prove that either is better than the other, so therefore dual-wield is better". I can't come up with any other way to interpret what you wrote above...

You could argue that maybe the sample size needs to be bigger considering the small difference between the two averages - but then the point has clearly been made that dual-wield doesn't seem to carry enough of an advantage over single-wield to be worth it, so arguing any further seems rather pointless...

Well, I probably didn't state it clearly enough. What I am saying is that if dual weild damage is actually lower then single weild, more often then not the dual weild damage in any sample set would be lower then the single weild. But the fact remains that even if Tetra stopped at 350 shots, he would find that the damages are the same. But he chose to continue forward for 69 more shots to get a higher average on the single weild side.

Is that coincidence? Who knows. Maybe he wasnt aware of it, maybe he just wanted to get a higher average on the single weild side.

The damages ARE the same. Any number of tests can prove that to you.

But you want to know what the kicker here is?

The highest number in BOTH damage sets is in the Dual Weild Category. These are the highest numbers in both damage sets.

Shot 288 was 603 damage in the DUAL WEILD Category
Shot 311 was 601 damage in the SINGLE WEILD Category

[Tetra]

Originally posted by Lucid Flow

Is that coincidence? Who knows. Maybe he wasnt aware of it, maybe he just wanted to get a higher average on the single weild side.

I had planned on going to 1k shots but people were showing up wanting to kill the bullies and since i was at 300some i went till i had 400+whatever it took to finish the last mob then did the same and chopped off the last couple hits with single wield.


Considering you still havent done any tests and have proclaimed yourself that your weapons always do minimum damage I hardly see how you have any right to participate in this thread as an informed or even knowledgeable poster.

[__________Lucid Flow]

Originally posted by Tetra
Considering you still havent done any tests and have proclaimed yourself that your weapons always do minimum damage I hardly see how you have any right to participate in this thread as an informed or even knowledgeable poster.

Participate? Heh, thats a funny statment by you but that is ok. You can't explain the discrepencies in your own numbers and that is ok too. I decided it would be better to use your own evidence against you to prove the point rather then create my own.

[______________Smoore]

Numbers aside, the really important test is, in fact the damage over time test. Basically, ok, let's say I'll end up doing the same average damage single or dual wielding over the course of 400 shots. The question becomes, when I fight that Real Mean or whatever, how much faster will I pop damage into him dual wielding.

Hypothetical number here, as an example:

Single wielding, I take 20 shots to down him at 2 seconds each, doing a total of 10k damage.

Dual Wielding, I take 20 shots for the same damage and drop him.

How much faster did I drop him dual wielding. Did I drop him faster? If I can take away 5 of his hits on my by dual wielding and popping in damage faster, is it worth it?

There's the real question. Damage over shots is basically unimportant. Performance in actual combat, measured by 'How fast can I drop X mob' is what matters. The tests that started this thread were in regards to this, I believe, and those are what should be built on, not only number of shots, but damage over time.

Time single for 40 seconds, same for dual. Do it 4-5 times for a decent sample. Figure out how much damage was done in those 40 second periods with single and with dual, and there's the important data.

Not that I dual pistols anyway, but that's my view of it, if I was thinking about doing so, that's the info that would matter to me. I don't care if my average damage is 7 lower if the mob drops faster anyway.

[Ashuras]

To those who did tests with Riders.....

U forgot one thing.. dual wielding riders has a significant advantage over single... becoz of Rider's (slower) speed.

There is almost no point trying to dual wield 2 fast weapons.

The BBI faithful is almost as good as riders... only problem being their crazy requirements and lack of BURSTs.

[__________Lucid Flow]

Originally posted by Ashuras
To those who did tests with Riders.....

U forgot one thing.. dual wielding riders has a significant advantage over single... becoz of Rider's (slower) speed.

There is almost no point trying to dual wield 2 fast weapons.

The BBI faithful is almost as good as riders... only problem being their crazy requirements and lack of BURSTs.

The other note on that is you can dual weild the same qlvl Rider. That will give you a 1.5x max damage output over dual weilding one.

Pistols are generally faster (with a faster animation speed as well) but generally have lower damage and extremely higher multi reqs.

What happens with most pistol is you can equip a q200 in your main hand and a q120-130 in your offhand. You essentially slow down your bigger hitter to make room for a pea shooter in your offhand. This slow down added to the smaller damage for the sacrifice makes it debatable on whether to dual weild or not if they have similar speeds.

I think the highest multi-req should be 1000 on pistol and 1400 on Shotguns. Average pistol multi Req should be 800. 1h Shotguns have a very good damage output, but pistols are pea shooters. At least give us the 1.5x damage output advantage.

[Tetra]

Originally posted by Smoore
Numbers aside, the really important test is, in fact the damage over time test. Basically, ok, let's say I'll end up doing the same average damage single or dual wielding over the course of 400 shots. The question becomes, when I fight that Real Mean or whatever, how much faster will I pop damage into him dual wielding.

The reason it's important is that after ql100 or so the multiranged req's on pistols spirals out of control with a practically exponential advancement, right now fully implanted with maxed multiranged at the 125 titlecap I can buff into a set of bbi faithfuls that's ql180ish primary and ql120ish offhand...

If the weapons do a percent or two lower average damage when dual wielding them and your wielding a set of weapons where the ofhand is only 2/3'rds the primary hand that's a big deal really since your doing below 1x damage *ANYWAYS* because your only shooting 120% faster while your offhand is close to 60-70% at best the ql of your offhand... well.. just have to connect the dots and think about how much crap your damage would be if you replaced your offhand rider exec with one that was only 2/3rds the ql of your primary.

[engelstein]

Originally posted by Tetra



The reason it's important is that after ql100 or so the multiranged req's on pistols spirals out of control with a practically exponential advancement, right now fully implanted with maxed multiranged at the 125 titlecap I can buff into a set of bbi faithfuls that's ql180ish primary and ql120ish offhand...

The delta differences at higher QL are worrying. But hell, with the Faithful you have all of 5 data points to go from. Not a great place to start making a claim on exponential req progressions.

[Tetra]

Originally posted by Engelstein
The delta differences at higher QL are worrying. But hell, with the Faithful you have all of 5 data points to go from. Not a great place to start making a claim on exponential req progressions.

Your right, i think the actual term I was looking for is logarithmic but math was never my strong point so i could be wrong there... but hey lets have the numbers speak for themselves

QL64 is 302pistol/268multiranged
QL100 is 471/412
QL113 is 532/503
QL135 is 635/656
QL150 is 706/760
QL161 is 757/837
QL186 is 875/1011
QL199 is 936/1102
QL200 is 941/1109

Now I'll also take this time to point out that faded pistol and multiranged clusters both take the same slot...

now anyone can go in game and look at these weapons... but really I think most will agree that my numbers which are correct are not something I simply made up.... stating something that you can look at an unchanging number on an item is not a claim... it's a fact.

[_______________Chill]

If the current dual wield requirements are to stay the only way to make it effective at high level is to have 2 separate and independant attack bars for each weapon to allow them to shoot at the same time.
Attack with the right hand as you will without dual wielding (1.0/1.0 for example) then add a 2nd attack bar for the off hand that is moving at its own speed without interfering with the other. In this case you have an extra attack from a lower QL weapon. With 2 weapons of the same QL we would do double damage then.

[engelstein]

I don't think it's logarithmic at all. Between items in the database, the progression is linear. So at QL 99 the Faithful takes a jump from 4 MR/QL to 7 MR/QL. More of a piecewise function than logarithmic.

I am assuming that AO always uses linear QL interpolation, which is certainly not necessarily the case.

[Jayde]

From my experience, stat requirements are 100% linear. You can map between any of the opening/closing database items with complete accuracy.

Of course, sometimes the different "steps" change this progession...which is why you need to map between the opener/closer. (I.E. QL 1 then QL 99. Not QL 1 to the QL 100 or QL 200 database entry.)

-Jayde

[Tetra]

Engelstein, do you just argue for the sake of argument or are you actually trying to make a point where you somehow make the multiranged advancement rate of on pistols to be reasonable in some way?

In fact what is your point? I had said that multiranged req's skyrocket out of control and all you have done is try to claim that they are perfectly normal and acceptable.

[Thyrra]

Originally posted by Tetra
In fact what is your point? I had said that multiranged req's skyrocket out of control and all you have done is try to claim that they are perfectly normal and acceptable.

i didn't read him as saying that the current multi reqs are fine. i did read him as saying that they likely have a steady rate of increase rather than an accelerating rate.

advents, if nobody else, should be able to wield two q200 pistols. advents and enfs should be able to wield 2 q200 melee weapons. this assumes that the current state of weapon effectiveness continues.

[Mercatura]

Linear or not, MR is far too high to begin with.

Theres some multimelee weapons that aren't insane, which is nice. Theres plenty that are though.

Frankly, for all the IP and pain in the @ss of MR/MM in general, two weapons should be: twice as fast, seperate reloads where applicable, and seperate specials. God forbid it doesn't suck