Weapon Ratios

[Synthlol]

Since anyone can remember, the standard for judging weapon quality (in terms of swing damage output) has been the ratio of its damage to its delay. Of course, the main hand damage bonus encourages faster weapons in the Primary slot, but for the offhand weapon, the advice is invariably to "use the best ratio" to maximize dps.

I have to ask: Has it ever been proven that (barring the MH damage bonus) equal ratio weapons produce the same damage output in the long run?

It doesn't seem to be self-evident that damage and delay carry equal weight. How do we know that in the offhand, a 10/20 and a 15/30 weapon average to the same dps? I'm no expert on the formulas that control swing damage.

[Swish]

How do we know that in the offhand, a 10/20 and a 15/30 weapon average to the same dps?

You can never prove it fully, but assuming both weapons have the same weapon skill (eg 1h slash) the law of averages should keep things even.

Rogues will always go for the highest damage weapon in the main hand and the best ratio in the off hand... inevitably we have a server of rogues with epics in main hand and edge of the nightwalker or CSS etc in their offhand if they aren't VP raiders.

Warriors have more choice outside of epix and things like weapon procs come into it.

[Supaskillz]

You could prove it if you know the code driving the dmg formulas, but I dont think anyone is showing you those.

People parse and it seems to be that the best weapon setups are those with the best ratio in off hand. I think you are safe assuming this is true.

Rogues are the weird example where there is a trade off with high dmg and low delay for the primary. Low delay takes advantage of primary dmg bonus better while high dmg leads to better backstabs. Swish over simplifies by saying highest dmg weapon in primary. In the extreme example: thornstinger is going to way out dmg efreeti war spear even though efreeti war spear has much higher dmg.

[sox7d]

Since anyone can remember, the standard for judging weapon quality (in terms of swing damage output) has been the ratio of its damage to its delay. Of course, the main hand damage bonus encourages faster weapons in the Primary slot, but for the offhand weapon, the advice is invariably to "use the best ratio" to maximize dps.

I have to ask: Has it ever been proven that (barring the MH damage bonus) equal ratio weapons produce the same damage output in the long run?

It doesn't seem to be self-evident that damage and delay carry equal weight. How do we know that in the offhand, a 10/20 and a 15/30 weapon average to the same dps? I'm no expert on the formulas that control swing damage.

Been saying this forever. Thank you. Who's to say 1 dmg = 1 delay when it comes to raw ratio? I'd love to see a mythbusters series for P99 with all the bro science floating around... (hybrid penalty... AC mechanics...)

[Swish]

If people's enthusiasm to know matched people's enthusiasm to go ahead and parse and test...we'd have answers ;)

Personally I like to relax and play, if my offhand does marginally less per hour than some other rogue's off hand with the same ratio I'm not going to lose any sleep over it.

[Danth]

Faster weapons will give marginally better service against weak monsters since if you kill something in a few hits, a big final hit will mostly be wasted overkill. Slower weapons are of course preferred for jousting and are less of a liability if fighting something that has a damage shield. As such there are some situations where an individual player will favor faster or slower weapons in spite of ratios.

Danth

[Synthlol]

Rogues will always go for the highest damage weapon in the main hand and the best ratio in the off hand... inevitably we have a server of rogues with epics in main hand and edge of the nightwalker or CSS etc in their offhand if they aren't VP raiders.

Warriors have more choice outside of epix and things like weapon procs come into it.

Rogues are the weird example where there is a trade off with high dmg and low delay for the primary. Low delay takes advantage of primary dmg bonus better while high dmg leads to better backstabs. Swish over simplifies by saying highest dmg weapon in primary. In the extreme example: thornstinger is going to way out dmg efreeti war spear even though efreeti war spear has much higher dmg.

Faster weapons will give marginally better service against weak monsters since if you kill something in a few hits, a big final hit will mostly be wasted overkill. Slower weapons are of course preferred for jousting and are less of a liability if fighting something that has a damage shield. As such there are some situations where an individual player will favor faster or slower weapons in spite of ratios.

Danth

These nuances I understand.

I'm questioning the validity of certain assumptions which seem to be commonplace. It may be nit-picky, but I'm curious and don't know how to find an answer. Say a monk has three weapons to choose from for his offhand:

1. Blackjack (http://wiki.project1999.com/Blackjack)

2. Knobbed Warclub (http://wiki.project1999.com/Knobbed_Warclub)

3. Efreeti Mace (http://wiki.project1999.com/Efreeti_Mace)

He's wearing silk armor and doesn't care about the minor weight differences. Which weapon is best? Or, are they really equivalent?

Maybe the better question is: does weapon damage scale linearly at the same rate that delay increases time between attacks?

[Messianic]

Only way to know is parsing data. Unless someone does the work and posts it, we're all just shooting the poop. Which is ok.

[Jimjam]

How do fractions round when the damage a hit does it work out? I imagine some ratios would produce more beneficial round errors than others!

Hmm... thinking of that, wasn't rounding errors for haste actually a problem at some point on live?

[Ravager]

Also gotta remember the importance of delay when it comes to things like push and damage shields. Facing a lot of casters, I'd take a Jade Mace in the off hand over CSS. If the druid in your group isn't making himself useful and dispelling the damage shields, you're looking at the only situation where I'd even consider using a Green Jade Axe

[webrunner5]

No doubt that with equal weapons in both hands the Secondary does less damage because it hits less often.

Put 2 weapons with the same ratio in both hands, but have them different types. By that I mean say a 1hb in primary and a 1hs in secondary. It takes forever to raise the Slashing skill in secondary compared to the Blunt weapon in Primary or vise versa.

There is a formula around but I don't remember it. I would bet the Secondary, being equal, does at least 30% less damage than the Primary.

[Cecily]

Your primary weapon will trigger once every time its delay runs out. This will run a double attack check. Your secondary weapon will run a dual wield check when its delay runs out. If that is successful, you attack and then check for a double. This is where baby quads come from. Obviously offhand does less because it's not a sure thing. If you put a moss covered twig in your offhand though... Expect to get a ton of 1hb skill ups quickly.

That's why I always suggest to put the "best" damage dealing weapon in primary and w/e other junk you have in offhand. Upgrade main, swap old main to offhand. Also why I hate primary only weapons.

[Synthlol]

Also gotta remember the importance of delay when it comes to things like push and damage shields.

No doubt that with equal weapons in both hands the Secondary does less damage because it hits less often.

Again, I understand these nuances, but they are irrelevant to the thread topic. I'm not asking what hand a particular weapon should be equipped in, or for the pros/cons of using faster or slower weapons (other than delay's direct effect on autoattack dmg output).

The topic I'm trying to gain insight on is whether ceteris paribus, do equal ratio weapons truly produce the same average damage output? If damage bonuses didn't exist, as in the offhand weapon when dual wielding, or below level 28 in the main hand, are a jade mace and knuckles dusters essentially the same weapon?

Please ignore interactions of mechanics like damage shields, push, riposte checks, damage bonuses, damage caps, etc. I'm just looking for proof that equal ratio weapons have the same theoretical damage output when considering only the effects of their damage (uncapped) and delay (no mainhand bonus).

[Dulu]

Wouldn't abilities like Double/Triple Attack and Riposte play into a weapons overall DPS?

I don't know the rate at which DA/TA procs, or if it's modified by speed.. but Riposte seems to be pretty clearly enhanced by having a slow weapon in the main hand.

[Ravager]

Again, I understand these nuances, but they are irrelevant to the thread topic. I'm not asking what hand a particular weapon should be equipped in, or for the pros/cons of using faster or slower weapons (other than delay's direct effect on autoattack dmg output).

The topic I'm trying to gain insight on is whether ceteris paribus, do equal ratio weapons truly produce the same average damage output? If damage bonuses didn't exist, as in the offhand weapon when dual wielding, or below level 28 in the main hand, are a jade mace and knuckles dusters essentially the same weapon?

Please ignore interactions of mechanics like damage shields, push, riposte checks, damage bonuses, damage caps, etc. I'm just looking for proof that equal ratio weapons have the same theoretical damage output when considering only the effects of their damage (uncapped) and delay (no mainhand bonus).

In that case find a torpor shaman and a cleric and do a ton of parses on a long fight, maybe on Priest of Discord and let him ch himself and don't use haste. Short of having the code, you won't get that proof any other way.

Hypothetically, a low delay weapon of the same ratio will out damage a high delay weapon in a short group fight because you'll get more successful swings in before the mob dies and a miss or missed double attack doesn't lose as much damage as in a high delay weapon, though the difference is probably minimal. In a long fight they're probably the same.

[kaev]

In that case find a torpor shaman and a cleric and do a ton of parses on a long fight, maybe on Priest of Discord and let him ch himself and don't use haste. Short of having the code, you won't get that proof any other way.

Hypothetically, a low delay weapon of the same ratio will out damage a high delay weapon in a short group fight because you'll get more successful swings in before the mob dies and a miss or missed double attack doesn't lose as much damage as in a high delay weapon, though the difference is probably minimal. In a long fight they're probably the same.

Actually, the slower higher damage weapon should, on average, do slightly more damage. This is because the delay doesn't count until after the first swing (or dual wield check), which is immediate on pressing auto-attack (assuming target is in melee range, of course.)

Compare two weapons, one 10/20, and the other 20/40:
T+0.0: both weapons swing, the 20 damage weapon will, on average, do twice the damage of the 10 damage weapon.
T+2.0: faster weapon gets second swing, damage is now equal
T+4.0: both weapons swing, slower weapon now has 2x20, faster weapon has 3x10, you do the math...

All things being equal, the shorter the fight, the more it favors the higher damage weapon, with the most extreme case being a fight that lasts only one swing.

Things that are unequal:
* primary hand damage bonus (but this thread is about secondary)
* push desirable? you'll push more with faster weapon
* push undesirable? you'll push less with slower weapon
* target DS? you'll take less damage with slower weapon
* target hits very hard? you'll take less riposte damage with slower weapon


Things that you might think are unequal but are actually equal:
* comparing damage numbers without delay? um, seriously, simple math like this is not even slightly hard.
* weapon procs? weapon proc rate per swing is normalized to achieve a target rate vs. time (usually referred to as procs/minute), your chance of a proc on any given swing is affected by weapon delay and haste/slow (google shakerpaging for an amusing use/abuse of this during the PoP era.)


It is, of course, theoretically possible that the p99 devs have implemented a combat system that is fundamentally different from the one that was parsed to death on live. But I've never seen any evidence of that. And really, despite the 1960's era talking Barbie doll's claim, math is not hard.


Not taking a swipe at anybody with the "math isn't hard" comment, btw. Just trying to emphasize that this is pretty straightforward once you take the time to be methodical about your analysis.

[Danth]

Optionally, look at it another way: If there's a difference, it's so minute that nobody has ever noticed after all these years.

Danth

[Synthlol]

Please ignore interactions of mechanics like damage shields, push, riposte checks, damage bonuses, damage caps, etc. I'm just looking for proof that equal ratio weapons have the same theoretical damage output when considering only the effects of their damage (uncapped) and delay (no mainhand bonus).

Wouldn't abilities like Double/Triple Attack and Riposte play into a weapons overall DPS?

I don't know the rate at which DA/TA procs, or if it's modified by speed.. but Riposte seems to be pretty clearly enhanced by having a slow weapon in the main hand.

I'm not talking about overall dps, I'm talking about autoattack dps. Dual wield and double attack skill checks may very well be crucial variables, as they directly control the number of autoattacks in a given time frame.

Ripostes are not autoattacks. The benefits of a slower weapon with regards to Riposte have no place in this discussion.

[Synthlol]

Things that are unequal:
* push desirable? you'll push more with faster weapon
* push undesirable? you'll push less with slower weapon
* target DS? you'll take less damage with slower weapon
* target hits very hard? you'll take less riposte damage with slower weapon



I guess I'm just going to have to constantly repeat myself throughout this entire thread, but these are all things I understand which also have zero relevance to my question.

What I'm looking for is clarification of the foundation of the assumption that the ratio of weapon base damage to weapon delay is the premier metric of offhand weapon quality, when in fact swing damage is distributed in a manner unknown to me within a damage curve which is calculated by a formula (which again I don't know specifically) that takes into account not only weapon base damage, but also things like character STR value, character ATK rating, and target AC rating.

I know faster weapons push more.
I know faster weapons riposte for less damage.
I know faster weapons cause the player to eat more ripostes and damage shields.
I know faster weapons apply the static mainhand damage bonus more frequently.
I know there's a damage cap limiting swing damage at low levels.


Please exclude these kinds of topics from the discussion, as they are not the focus. I'm really at a loss on how to clarify any further what I'm getting at. I'm not trying to argue why judging by ratio is wrong. I'm saying I don't know why it's right.

[Jimjam]

I've noticed when playing low level characters that sometimes my weapon will do more than it's 'max hit' (i.e. greater than 2xDMG). It tends to happen more when my Str is buffed.

Can this bonus damage from Str apply to the off hand? And do we know if this bonus damage is a factor of DMG or is it a simply added on bonus?

If the rare bonus damage from Str applies to the offhand AND it is not a factor of DMG then it stands to reason that lower delay weapons will benefit more from it, in the same way that low delay primary weapons benefit from damage bonus (just to a lesser extent).

Also, to repeat my previous question, how does EQ handle rounding fractions when working out damage/hit? This could produce errors that might benefit different delays.

[curtischoy]

The topic I'm trying to gain insight on is whether ceteris paribus, do equal ratio weapons truly produce the same average damage output?

i did not read any other posts after you said this because this is a very easy question to answer. All you have to do is take something like, as you suggested, a 10/20 wep and a 15/30 wep and parse them.

Parse them on the same type of mobs, or if you can, the same mobs from a static spawn. (Brutes in WW caves near dalnir will always spawn at the same level, meaning that in the room with 3, the one on the left will aways be lvl x, middle will always be lvl y, right will always be lvl z) Parse them and look at the data by using a two sample t test for means to see if they are statistically different or not.

This method completely eliminates any other varaibles and it does not care how ac is calculated etc. It only cares about the number of swings, mean, and standard deviation of each sample. The only variables that have to be controlled are the weapons you are using, your skill using those weapons, your class, and the level of the mobs; all of which are very easy to control.

Problem solved.

[curtischoy]

Oh and control for str as well. There may be a few others you have to control for that I am missing, but they should be super easy to control.

[Synthlol]

i did not read any other posts after you said this because this is a very easy question to answer. All you have to do is take something like, as you suggested, a 10/20 wep and a 15/30 wep and parse them.

Parse them on the same type of mobs, or if you can, the same mobs from a static spawn. (Brutes in WW caves near dalnir will always spawn at the same level, meaning that in the room with 3, the one on the left will aways be lvl x, middle will always be lvl y, right will always be lvl z) Parse them and look at the data by using a two sample t test for means to see if they are statistically different or not.

This method completely eliminates any other varaibles and it does not care how ac is calculated etc. It only cares about the number of swings, mean, and standard deviation of each sample. The only variables that have to be controlled are the weapons you are using, your skill using those weapons, your class, and the level of the mobs; all of which are very easy to control.

Problem solved.

I've thought about this, but there's a roadblock that always stops me short of doing my own testing, which is also why I came here to see if anyone has an answer.

I don't know how to 'parse'.

[emdubwubwub]

Download and read about Gamparse.

[Ele]

I don't know how to 'parse'.

Turn your logs on and use Gamparse.

[koros]

Been saying this forever. Thank you. Who's to say 1 dmg = 1 delay when it comes to raw ratio? I'd love to see a mythbusters series for P99 with all the bro science floating around... (hybrid penalty... AC mechanics...)

What? You're misinterpreting what a ratio is. 1/2 = 2/4. The the damage and delay are never compared. only damage to damage and delay to delay.

[Clark]

If people's enthusiasm to know matched people's enthusiasm to go ahead and parse and test...we'd have answers ;)

Personally I like to relax and play, if my offhand does marginally less per hour than some other rogue's off hand with the same ratio I'm not going to lose any sleep over it.

:)

[Foggon]

Synth. Here is my answer to you, and it is not very scientific so-to-speak. I don't run a parse or anything(Kind of takes away from the fantasy right?) but I have asked friends to parse me from time to time and this is what I find...

Using faster weapons with less damage I notice my DPS is generally always about the same.. give or take about 3-4 dps..

Using harder hitting slower weapons of or about the same ratio, I tend to have about the same DPS on average as the quicker weapons, but with a bit more variance... maybe 15-20 more or less dps.

Before I had my epic I played with many weapons. Bags of weapons. My quick weapons(and my overall average dps) was about 45 dps. I feel like I saw that number everytime I asked for a parse, 44 or 45.. sometimes 48. But when I pulled out the meat luggers.. sometimes I would reach up to 68 dps, and other times I had seen it as low as 34 dps.

Some nut somewhere probably has all this calculated out, and like I said I don't have the fractions and numbers to min/max these tests of mine.. But I think simply enough it is because if you miss with a high delay weapon then you just lost a ton of play time. If you miss twice with a big delay weapon you might as well start the bus and go home.

Of course on the other hand you have the chance to hit two doubles in a row for maximum damage(Mmm dat feels) and that's the kind of stuff that causes the fluctuation in dps.

[Raev]

The topic I'm trying to gain insight on is whether ceteris paribus, do equal ratio weapons truly produce the same average damage output?

I did some parses for my monk, and my conclusion was yes, but as you are being told the only way to convince yourself is to download gamparse and give it a shot.

[GinnasP99]

Synth. Here is my answer to you, and it is not very scientific so-to-speak. I don't run a parse or anything(Kind of takes away from the fantasy right?) but I have asked friends to parse me from time to time and this is what I find...

Using faster weapons with less damage I notice my DPS is generally always about the same.. give or take about 3-4 dps..

Using harder hitting slower weapons of or about the same ratio, I tend to have about the same DPS on average as the quicker weapons, but with a bit more variance... maybe 15-20 more or less dps.

Before I had my epic I played with many weapons. Bags of weapons. My quick weapons(and my overall average dps) was about 45 dps. I feel like I saw that number everytime I asked for a parse, 44 or 45.. sometimes 48. But when I pulled out the meat luggers.. sometimes I would reach up to 68 dps, and other times I had seen it as low as 34 dps.

Some nut somewhere probably has all this calculated out, and like I said I don't have the fractions and numbers to min/max these tests of mine.. But I think simply enough it is because if you miss with a high delay weapon then you just lost a ton of play time. If you miss twice with a big delay weapon you might as well start the bus and go home.

Of course on the other hand you have the chance to hit two doubles in a row for maximum damage(Mmm dat feels) and that's the kind of stuff that causes the fluctuation in dps.

This is spot on, its made drastically evident using the imbued fighters staff 38/40. Sometimes you double attack for 200 several times in a row, other times you'll miss for 7-10 seconds straight. Quicker weapons will parse much more consistently but you can get really lucky with an IFS or something similar and put out great damage. Riposte is also alot better with a heavy hitter

[fastboy21]

its really not very complicated. if you can't trust the ratios and math to predict to what your dps should be then you need to parse. the more you parse the more data. the more variables you control for the cleaner your data will be. you may also take into account that depending on class, level, skills, stats that certain weapons will produce different results due to the attack/damage formulas changing for a character as they progress. you may also want to control for whether your are tanking or not (reposte damage, etc.).

generally speaking, when i am parsing i am generally only looking for my pure melee dps when I am not tanking. I do not care about holding aggro, I do not care about reposte damage, etc. My "benchmark" figure is just pure auto-attack on a mob while somone else is holding aggro. Obviously, in real EQ, there is more to take into account, but I use pure dps as my standard to compare weapon setup for my own information.

as stated above, the more attacks you are making the tighter the spread on your data will be. with less attacks the bigger the spread. averages are still averages, but the consistancy will be more reliable with faster weapons. Example, I know fighting any golem in CoM with my regular weapon setup on my 50 war (sarnak warhammer and troc skean) I can basicly go afk and come back a minute later to find a dead golem. When I equip my ogguk cleaver, I might not always win the same fight, regardless of my expected average DPS...I might win big too. The very slow (and potentially heavy damage) attacks create a very very big chance that I might win big or die if I go afk.

the short of it is, that if you want to play the min/max numbers game (and many people don't---in fact, many group mates will HATE if you are parsing them while you fight) then you need to parse. if it makes the game less magical for you to reverse engineer the combat formulas then don't do it. some folks find playing with data to be a fun part of the game though.