DPS with Stam line... 1h and shield or 2h/DW?

Given that the first 4 points are wasted in each line, going beyond 3 lines seems inefficient. Optimally, you would want to go deep into only 2 lines. If you plan to be a tank (1H+shield), I would forget INT because 1H are very susceptible to the haste cap (delay cannot be reduced below 1.0 so a 2.0 delay weapon will not not benefit from a haste value over 125 (if I understand the curve correctly.)

There is no 1.0 sec delay cap on weapons. The only cap is that weapon delay can only be reduced by a max of 125% (200 Haste).

I like the recommendation for:
4/4/8 STR
4/4/8/4 WIS
4/1 INT

But, what if instead of the 4/4/8 STR I went with 4/4/8 STA since I use an <Ancient Velium Battle Hammer> and and <Ironplate Shielding> buckler. Would this be good then:
4/4/8 STA
4/4/8/4 WIS
4/1 INT

Getting the haste off of Accelerated Strike in the INT line requires a sword.

Where do you get this info? It seems every week or so someone is posting about a limit. There is NO limit. Dev/Coder said so right here in the forums, and I've tested it myself.
For quite some time I believed myself that 0.8 was a limit, because some knowledgeable leaders in the game, and others here on the forums said so. I was also having a problem with the /weaponstat command, where I couldn't get it to show less than 0.8 seconds. But regardless of what this says, I've used a parser, and auto attack only, and found myself swinging faster than 0.8 seconds.

That said, it's still always better to have a longer delay weapon, because haste is a percent. So 100% haste on a 3 second weapon, you save 1.5 seconds on each swing, where as 100% haste on a 1 second weapon, you only save 0.5 seconds per swing. Of course how much total actual damage output you gain will vary, depending on how fast (or slow) you spam your CA's, as this can interupt your auto attack some.

It was consensus before DoF. If it changed, great. If consensus was always wrong, then great.

Hey guys, there is no 1 sec delay for weapons, as Fight Game said.

Grab yourself a dual wield sword, put ur int buff on, open wounds and then engage in combat for berserk and type /weaponstat

Got to 0.6 delay last night testing it

My 2 hander when im solo and in combat without open wounds drops from 2.5 to 1.3 which is nice, in a group im normally about 0.9

***removed to prevent miss-info***
Thanks for clarification
Message Edited by RufusDeMarko on 02-01-2007 01:05 PM

Blocking "power" is overrated anyway, if you're looking for evasion you can get much more potent returns in other areas of evasion.
The order of defensive checks is: Enemy's Hit Roll -> Parry/Riposte -> Block -> (Dodge) -> People's cover of you

Here's some data from a fight I tanked Zylphax the Shredder, limited to melee only:
Total attacks: 265
55 Misses (20.7%) Shredder just flat out failed his attack roll, or I "dodged." This is with roughly 430 defense.
26 Riposte (9.8%) Riposte is from my testing, uncontested. I have full STA ripote and the riposte adorn.
35 Parries (13.2%) This is with roughly 450 parry, with a dirge. This makes sense as I have 6.5% riposte rate base, 4.5% from achievement, 2 from adorn. Then 20% of parry become ripostes. 3.3/(3.3+13.2) = 20%, 6.5+3.3 = 9.8% which was my riposte rate above.
11 Blocks (4.1%) This is all my 10.9% block chance came out to be, Buckler of the Howler + Sta block spec.
11 Other-Blocks (4.1%) This is the number of times the paladin covering me blocked/parried/riposted.

Of interest:
42 Stoneskin procs (Templar/Dirge combo)

So
Evaded: 55+26+35+11+11 = 138, 138/265 = 52% of Shredder's attacks were shut down by evasion.
Nullified: 42, 42/265 = 15.8% of Shredder's attacks were nullified by stoneskin.
Total: 180, 180/265 = 67.9% of Shredder's Attacks were ENTIRELY ineffectual.

Of that remaining damage my mitigation ate roughly 60-65% of it through the fight, rough average of 63% from mit buffs/other buffs.
32.1% damage checked by mitigation, reduced by 63% = 11.88% of his total damage got through.

So in the end, 52% of shredder's DPS was shut down by evasion, 15.8% by stoneskin, and 20.22% by mitigation, allowing only 11.877% of his total melee dps in. My evasion was almost two and a half times more potent than my mitigation as a raid tank, take this for what you will, but also realize I am built for evasion. It is not uncommon for my evade rating to top 9k while raid tanking.

Of all of that evasion, blocking was only worth a whopping 4.1% total reduction. A tower shield blocks about twice as often as a buckler, but you give up the 4.5% riposte as well. That means your net gain in defensive potential is about 5% evasion. Your block may go up 8-9%, but your riposte would drop around 4-5%.

Because riposte comes earlier in defensive checks (and seems uncontested) it is also more potent. The maximum advantage of a tower over a buckler is about 5% by my estimates. Now this may seem small, but if you're barely hanging on to a fight, the difference between 1000 dps and 1080 dps being done to your tank can be life or death. Couple that with the ability to spec into defensive skills or parry while holding a tower with your now free points and you can get another 2-4%, making the maximum defensive advantage around 8-9%.

That's the tradeoff for the offensive advantage with the buckler.

As much as me and kemt may disagree with a lot of aspects, he's always spot on with his knowledge of the Buckler line and tanking in general.

Hes been tanking for an alliance of guilds for quite some time, and if you ask anyone who particpates i'm yet to come across many with negative comments.

So all in all, i think he just closed the argument about tanking with the buckler line and its capabilities

I've been using the STA line as a raid tank for a little while now. With the 4th ability in that line, my chance to block sits around 10.4% with a legendary buckler equipped (Badge of Service). My best shield at the moment is the Crested Mistmoore one from the EoF heritage quest with a 16.2% block (I've been so unlucky with drops that I'm beginning to think I'm cursed). That's a 5.8% difference for me. That means if the random number generator holds true, I will have blocked an additional 5.8 hits out of 100 (or roughly 1 out of every 17). Even with a good fabled shield with a 20% block chance, that would still only be an additional 1 out of 10 hits blocked. Against a single target when a flurry of hits isn't coming in all at once, the buckler is enough. It's even enough against groups of mobs as long as a wave of hits will not spike higher than your total health and the wards placed on you. Don't get me wrong. IMHO, block is still very important, but in most situations I find myself in, the buckler is enough.

EDIT: Some people also seem to be confused regarding contested versus uncontested avoidance. Uncontested doesn't mean 20% block against an even con mob will be 20% block against a level 75^^^x4. The epic mob will still have hit bonuses. There just is no skill check. There is no block skill. At least, that's how I've come to understand it. So, if that is true, then my numbers above are being generous.

Switching from a 2H to a 1H without a significant boost to 1H DPS is a big DPS loss as well. It's also 2 pieces of gear you have to swap in vs just changing shields. With the stamina line, switching from offense to defense is pretty much just switching stances. It's a very nice hybrid build in my opinion.

EDIT: Dimglow is spot on.

Message Edited by uux on 02-01-2007 12:53 PM

Pretty much what Uux just said, if you want pure DPS, then 2 hander is the way to go, but be prepared to never tank anything worth while. E.g. Nizara or raid zones.

However, if u want to be able to tank the harder stuff, and still do the DPS, then buckler and 1 hander is the way for you

The number look totally different if u use a shield. i am specced str/int and use 2 block adornments + the parry adornment on the weapon.
i checked the log from inner sanctum and got the following
291 blocks
76 parrys
46 counters
104 misses

the numbers maybe be a bit off because i used a text editor to count, but it shows that blocking is extremely important if u use a tower shield.


*edit* for autoattack only
272 hits
358 misses
- 182 blocks
- 46 parries
- 26 ripostes
- 104 misses
the block rate is really high. i thought it should be something like 20%, but it is much more.






Message Edited by hakedu on 02-01-2007 10:44 AM

(edited out some math) Either you had some extreme buffs on you (including a second chance to avoid) or you have to be the luckiest SOB I have ever seen when it comes to the random number generator. Yes, streaks like that are very possible, no matter how unlikely they may occur. This is why I still think block is important. However, using the STA line is a trade off to still receive enough block to suffice for most cases.

Message Edited by uux on 02-01-2007 02:42 PM

Check your text again. Specifically make sure you're only counting instances of "name tries to slash you, but YOU block" type text.

No offense but that many blocks is completely nonsense, and goes against years of parsing done by hundreds of people as well as a lot of testing regarding defensive check order from brawlers to crusaders to warriors.

I wouldn't be surprised if you're counting enemy blocks, or some other text entry of block as your own.

You should ONLY be counting situations of "but YOU block" "but YOU parry" "but YOU riposte" type text. Otherwise you're getting the entire raid's evasion checks and quite possibly your enemy's. Also make sure that your hits incoming and hits evaded match up to a real parser's evaluation, IE ACT's incoming damage, slashing (500 attacks, 300 hits, 200 misses) you should be able to find exactly 200 "but YOU blah" or "but NAME blah" where Name is a fighter covering you.

i agree that the numbers are very high, but thats what the logfile says. maybe someone else with tower shield can double check. my group was me, templar, warden, mystic, dirge, coercer. dont know if they have some aa that increase blocking chances.

i have a german logfile, what i count is:
versucht EUCH zu .*waffen -> all auto-attack misses
versucht EUCH zu .*waffen.* IHR blocken -> blocks
versucht EUCH zu .*waffen.* IHR parieren -> parrys
versucht EUCH zu .*waffen.* IHR kontert -> ripostes
versucht EUCH zu .*waffen.* verfehlt aber -> misses

trifft YOU for .*waffenverletzung -> auto-attack hits

I deleted some math from my post previously because I noticed the first set of numbers you posted didn't included the number of hits which throws the total number of swings a little out of proportion. Even so, the second set of numbers for autoattack puts your zone wide block rate over 30%. Your group make up didn't include a class with a second chance to block for you either. While the random number generator can be streaky, it's very unlikely to be so for the duration of a zone wide parse. It should fall very close to the listed chance to block in the persona window. Also, being only autoattack the total number of swings may also be off.

unfortunately i have no parser that computes hits/misses/blocks/parrys.
but i think i found the reason why my blocking rate was so incredibly high: i forgot to count the hits that didnt do damage (because of shields). this were 330 hits.
so i have blocked 182/960 = 18,96 %.
this seems reasonable. and it shows that blocking is extremely important if u use a tower shield:
- overall avoidance was about 65% (persona window, dont know the exact number)
- block chance 21.3%
-> but half of the misses came from blocking

30% block rate is insane. Sorry, but Dimglow is right, your not parsing correctly. Just remember that block is an outright, unmodified ability, e.g. it doesnt follow the standard rule of offensive skill vs defensive skill etc. It is also not affected by the con of the mob, so your not doing something correctly.

.6 actually seems to be the cap, i have a 1.2 delay weapon with my monk and i grouped up with an illusionist to get 170ish haste, only brought me to .6, same as when i had 100% haste