Myths of Riposte

[kiztent]

Spent 2 hours in Crypt of Decay noodling around with various mobs. Enchanter on board so was single combat at all times, mobs shaman slowed (except for 3 mobs that couldn't be slowed). Fully hasted (36 percent worn and 66 percent for speed of the brood to hit haste cap). Weapon was Velium gemmed warhammer.

Myth 1:
weapons with bigger hit are better since they do more riposte damage.

Parse:
I swung 4292 times total and riposted 59 times. Ripostes were 1.3 percent of the swings taken. (4292 swings * .4841 accuracy * 97 damage (magic number) = 201542 total damage)

With a Windblade this would be 2752 swings total and 59 ripostes. Ripostes would be 2.1 percent of the swings taken. (2752 swings * .4841 accuracy * 145 damage (magic number) = 193175 total damage)

95.8 percent damage, compared to the expected from MN/Delay of 95.6 percent damage. About .2 percent increase in damage output from extra riposted. I guess it's not a myth technically.

My swings:
2058 hits
41 crits (CF1)
1638 misses
181 dodged
203 parried
171 riposted (I was riposted)

Myth 2:
slower weapons are better because of fewer ripostes.

Actually this is true as well. I was swung at 1415 times and of those 171 were from me being riposted. WIth a 44 delay weapon, I would have been riposted 108 times, for a total of 1352 swings. This results in a 4.5 percent reduction in damage (1352/1415).

This is exactly offset by the 4.6 percent increase in damage from a War Marshall's Bladed staff (dead things don't hit back).

Swings at me (with LR3-4):
681 hit me
550 missed me
53 I dodged
72 I parried
59 I riposted

Conclusion: Anyone who pays more for a Windblade than a War Marshall's Bladed Staff got ripped off. Unless 20 AC is worth that much to you (70k versus 6k).

[Lickity]

20 AC is not worth 66k, but that doesn't make it worthless. I would gladly pay several thousand more for a higher AC weapon (assuming AC is important to the class, i.e. a tanking class.)

Warriors/paladin/SKs are not taken into a group for their ability to deal dps. They are brought in to stay alive and hold aggro so that the DPS classes in the group can work at their full potential. This is a rather unparseable stat (ability to hold aggro,) but it is one that needs full investigation before the case of "War Marshall vs. Windblade - Which is a better deal" is set in stone.

[kiztent]

This is a rather unparseable stat (ability to hold aggro,) but it is one that needs full investigation before the case of "War Marshall vs. Windblade - Which is a better deal" is set in stone.

What's there to investigate? I could hold aggro with 2 opal steins if I felt like it.

You own weapons for one reason: DPS. There's 20 slots that can raise your AC and HP and stats, there's only 1 slot you can raise your DPS in (barring ATK items, which are a constant raise of DPS, dual weilders and ranger's bows).

[Nhin Impious]

This is a rather unparseable stat (ability to hold aggro,) but it is one that needs full investigation before the case of "War Marshall vs. Windblade - Which is a better deal" is set in stone.

Wouldn't that be done by counting the corpses of other party members?

I could hold aggro with 2 opal steins if I felt like it.

If I hadn't tried and failed to get aggro from you on several occasions, I would take this for sheer bravado.

Nhinx Aphsion
Paladin of Innoruuk who can get aggro from anybody...except Kiztent

[Itzena]

Also: Windblade is fugly, WMBS is not.

[kiztent]

This is a rather unparseable stat (ability to hold aggro,) but it is one that needs full investigation before the case of "War Marshall vs. Windblade - Which is a better deal" is set in stone.

Followup: It may be unparsable, but the amount of hate from weapons isn't reallly significant. I believe it's been found to be 20ish hate per swing. In comparison, EB is 700 hate, terror of thule is 600 hate, anger 3 is 600 hate, taunting blow (Cord Hilted Spike Driver) is 450, pique is 100 hate and 50 damage (Barnacle Covered Axe) - no other spells have been accurately measured for hate, those are the only direct hate procs/spells I know off the top of my head.

But, at 20 hate per swing and 700 hate from an EB proc, you'd need 35 swings (it would take most of a LDoN mob to swing that much). Even if you double melee hate to 40 a swing, it's not a lot next to procs and spells.

Oh, and thanks Nhinx for not mentioned all the times the mob seems to be slow to die and I notice at 30 health or so I forgot to turn on attack

[KyrosKrane]

With a Windblade this would be 2752 swings total and 59 ripostes. Ripostes would be 2.1 percent of the swings taken. (2752 swings * .4841 accuracy * 145 damage (magic number) = 193175 total damage)

(Highlight mine.) Did you parse this or just assume it? As stated, it seems you just assumed linearity and calculated from there. Actually, there's a few different assumptions here.

1) You assume that the rate of ripostes you swing at the mob is independent of the weapon. This may indeed be the case, and I'll concede that it's likely, but it still needs to be established.

2) You assume that the number of swings with a Windblade compared to the VGWH would change linearly with the delay. This is not necessarily correct. There are a number of other factors you ignore, such as swings lost due to casting and swings lost to being stunned (due to ripostes taken from a faster weapon). This can only be determined by parsing a Windblade and comparing the actual accuracy against the theoretical accuracy you use above.

3) Although it's not relevant to the mathematical analysis, you also have to take into account how many additional hits could have been avoided or mitigated due to the WB's extra AC. (It's not relevant to an analysis of ripostes dealt to the mob, but it is relevant when calculating ripostes taken, and hence when determining which is the better weapon.)

For your first myth, you test whether "weapons with bigger hit are better since they do more riposte damage." This looks at how often you riposte a mob, and it's determined by looking at the average size of a riposte hit. You don't give the needed numbers, but assuming the magic number also holds for ripostes, a Windblade deals about 50% more damage per riposte hit, on average, than a VGWH. (As you note, though, the actual damage increase is insignificant.)

For your second myth, you test whether "slower weapons are better because of fewer ripostes." This deals with how many ripostes the mob swings at you, and this is where the linearity assumption could cloud your results. This has to be tested to get a meaningful result.

In any case, your numbers do establish that the number of ripostes both dealt and received is all but insignificant. The defensive aspect comes into play only when you are fighting a borderline mob -- that one extra riposte due to a fast weapon may allow the mob to one-round you whereas you would have avoided the riposte with a slower weapon.

And if you did parse the WB, forget all I said above. =)

Edit: Since you have a parse of the VGWH handy, you may as well parse your average hit and use that to calculate your true magic number (it looks like you used 2.0 as a factor). The factor to determine the magic number has been shown to vary wildly based on your attack and the mob's AC and defense, ranging from about 2.0 to 5.0, if memory serves. The same factor should hold for both weapons, though

[Lickity]

I could hold aggro with 2 opal steins if I felt like it.

You failed to mention your class as well. I will assume you are an SK by the responses from the pally who can get aggro from anyone but you. Remember that while you are casting, you are dealing zero dps, your weapon isn't swinging. So, your ability to hold aggro without a weapon is irrelevent if you are speaking from a DPS standpoint. You want a weapon that CAN hold aggro, even at 20ish per swing, so you have to cast less. If you have a weapon with a significantly slower swing, it is doing much less for your aggro, that is just a fact. The effect is two fold, the weapon is doing less for your aggro, so you have to cast more, thereby making your weapon do even less for you in terms of aggro and dps.

Also, stun has been parsed, at least in reference to stun components on weapon procs to be 400 hate (CHoS, Sword of the Bloodsworn etc.)

EDIT: Reading your signature might have given me a good idea of your class as well =) /blush

[kiztent]

Remember that while you are casting, you are dealing zero dps, your weapon isn't swinging.

Well, yes and no. Weapon swings are discrete events, unlike real life, so if I'm casting a spear (.5 second cast), clinging darkness (1.75 second cast), I'm not going to significantly delay my swing, if I delay it at all. Touch of Innoruuk (3.2 second cast - less hybred haste) may delay 1 swing a little, but honestly, I'm not casting it that often.

I do want to look at how many swings I take in a given fight, but with double and triple attacks, this might not even be helpful. Not to mention that a shaman who decides it's fun to blow the mob halfway across the room and blame the paladins stunning (don't think I don't know what you guys do on raids ) is going to impact my damage a lot more than having a swing delayed a second to cast a lifetap.

Kyros:

Had a longer response, it got eaten.

Ripostes have been parsed to be linear (which seems intuitive, since it implies every swing has a fixed chance to be riposted).

"n" was 2.9 for my CoD parse, which still gives a WMBS a 5 percent edge over a WB. ~1350 ATK with 285 strength, 55 worn ATK and FA.

Bazaar geared people don't tank stuff that can 1 round them (and they use a sword and board to do it). Bazaar geared (mostly) I can break 7K fully buffed. How far would I need to progress to find something that could 1 round me?

20 raw AC (30 worn) would yield 1.5 percent less damage (AC up to 2200 in PoP elementals has been shown to linearly yield 5 percent mitigation per 100 points).

[Lickity]

Stupid shaman giving us all a bad wrap. Virulent Paralysis should NEVER be used in standard grouping. It does the group no good to have a mob knockbacked all over the zone for a brief stun.

WTS VP AA for something more useful.

As for what "you guys" do on raids, if my shaman ever use VP without a good reason on a raid, they are likely to get themselves into a lot of trouble.
(BTW: it is more often the REAL nuking classes that use knockback nukes to blow the mobs across the room =) )

/derail off

As for the casting vs. weapon dps, the scenario applies much more to Paladins than Shadowknights.

[kiztent]

Stupid shaman giving us all a bad wrap. Virulent Paralysis should NEVER be used in standard grouping. It does the group no good to have a mob knockbacked all over the zone for a brief stun.

Ask Sanna how many times an unslowed mob that resisted mez has been VPed off the top of her before you dismiss VP

She just breaks 700AC and 3k HP. Which is about what my SK has naked (though she does look better in a dress than I do naked).

[Abazagaroth]

I just typed a huge long post, but after spending an hour running numbers and using my parses I decided the differences weren't big enough and the numbers were going to be confusing. So I decided to go another route that would be cleared in explaining what I have posted every **** time this has come up on this board (by the same poster). What I found on a large sample (188000+ of my swings) when extrapolated to a WMBS and a Windblade is that you takes 4.1% less dps using windblade and give 3.9% more damage using WMBS. That jives with the end numbers you posted above.

This is NOT taking into account dps lost due to spell casting. A fully hasted WMBS swings about every 1.4 seconds . A fully hasted windblade swings about once every 2.2 seconds.

WMBS: 42.86 swings per minute
WB: 27.27 swings per minute

What happens if you are using a 1.0 sec spell and cast it between 2 and 6 times during that minute?

WMBS
timer: 1 2 3 4 5 6 7 8 9 10 11 12 13 14
lost to casting: 0 0 0 0 1 2 3 4 5 6 7 8 9 10
= 0.32 worth of a swing lost per spell cast (on average, in practice it could be 0 or or up to 0.72)

# of casts | swings lost | dps lost
1 | 0.32 | 0.75%
2 | 0.64 | 1.50%
3 | 0.96 | 2.25%
4 | 1.28 | 3.00%
5 | 1.60 | 3.75%
6 | 1.92 | 4.50%

What about a 2.0 sec cast time spell?

WMBS
timer: 1 2 3 4 5 6 7 8 9 10 11 12 13 14
lost to casting: 7 8 9 10 11 12 13 14 15 16 17 18 19 20
= 1.35 worth of a swing lost per spell cast (on average, in practice it could be 0.7 or or up to 2.0)

# of casts | swings lost | dps lost
1 | 1.35 | 3.15%
2 | 2.70 | 6.30%
3 | 4.05 | 9.45%
4 | 5.40 | 12.60%
5 | 6.75 | 15.75%
6 | 8.10 | 19.10%

Now, if we took the law of averages, those dps numbers would not change with windblade at all (because in the hypothetical, we are mashing the spell casting randomly). In reality a good knight will time those casts in order to either lose the least dps (when just casting for aggro) or to reduce the dps taken (like paladins stunning not right after their mob has swung, but right before they will swing again).

The higher the delay the greater leeway for human timing. For example, with a 2.2 delay between swings and a 1.0 cast time spell it is very possible to lose almost no dps at all, while when using a 1.4 delay between swings weapon you go from haveing 1.2 seconds to react to 0.4 seconds to react.

That is where the dps comes from. It has NOTHING to do with WB gaining dps, it is entirely to do with WMBS losing dps due to spell casting.

I will repeat for you what was stated many times:

If you are fighting from behind (not tanking) WMBS will do more damage. If you are not casting and tanking you will do more dps with WMBS. If you are a knight that is continuously casting while you are attacking the Windblade will do more dps.

If you are tanking ANYTHING that is not trivial (like a hard ldon for most tanks that are using either weapon) then the dps taken decrease is dramatic and the windblade far and away outstrips the WMBS.

People don't say "WMBS sucks, Windblade is the win!", they say "Windblade is an overall superior weapon for a tank, but like all of the best droppable items, your return per pp spent gets lower as your approach the best items, and if you can't afford it there is nothing wrong with a WMBS".

To claim that anyone buying a Windblade is getting "ripped off" is patently ridiculous.

Soloing you might take a WMBS over windblade if your mitigation is not that great such that the 4% decrease in dps taken is less than the dps decrease you'd get from killing 4% faster. In a group setting your 4% decrease in dps is now where near as important as the 4% increase in avoidance (i.e. your tanking ability) while tanking.

As I said, I have OWNED both weapons. I PARSED both weapons under the same conditions over a 2 week period (and I play 40 hours a week). Windblade came out slightly ahead (on the order of 1%) in dps given, and was far and away better while tanking in terms of damage taken.

You are quite simply wrong. And no matter how you try to spin the numbers, actually parsing both weapons under the same conditions shows that you are wrong.

[Lothay]

Closing.

Statemtents such as "you are wrong," are personal attacks, even though mild.

edit: Additionally, the term spin seems to be perjorative, since that is the general context in modern discussion.

It is possible to argue a conclusion, or point without attacking the perosn beihind the point.

Since this thread has turned the corner to personal attacks quite abruptly, I'm disinclined to allow continued conversation. Take the argument to some other message board.