Tuesday, May 4, 2010

Number Crunching Blade Warding

Unless you happen to be a DK (since they are spoiled brats) or a druid (they should use Mongoose), finding a nice tank wep enchant can prove to be diffcult. None of the ones out there are really -the- enchant to have, and debates are going high and low about which one to choose. So which ones are the nominees? Well, most of the time people either go for Mongoose, Blade Warding or Blood Draining (and in rare occassions Agility). There are some different attributes to look at which make people decide for one and not the other. For instance, Mongoose has said to have a much higher proc chance than Blade Warding, therefore the overall gain is better even though one proc of Mongoose is less good than one proc of Blade Warding. Some say Blood Draining is the better choice since it is the only one giving a proc which is beneficial 100% of the times (unless you are oneshot from +35% health to 0% in one blow).

Since both Mongoose and Blade Warding work with ratings (parry and agility which increases dodge), even if you get a proc there is a high chance it won't do anything. Say you've got 20% parry, get a Blade Ward proc and get 24% parry, there is still a 76% chance you won't parry at all. Some people say you can't think about "chance" that way. A higher chance is always better than a lower. Others say they'd rather work with definites. And I understand them, since in the world of WoW chances are dangerous things. Lives depend on them. But even if Blood Draining has a 100% chance to proc when you go below 35% hp, we're still only talking about around 2500 hp. Considering most well geared tanks go well over 40k hp nowadays, it feels like a fart in a storm. But then again, Blood Draining wards against all kinds of damage, whereas Mongoose and Blade Ward only works against physical. The problem about parry and dodge are that unless you have quite alot of them, they don't increase your survivability much, which is why alot of people rather go with "safe" stats like stamina instead. All these are issues to consider when choosing a proper tank enchant. But since there is no enchant that gives stamina (*raise fist at DKs*), we'll have to find out which of the other enchants is the best substitute.

Let's agree that none of the enchants are crazily good. Which one you choose will probably make a minor difference. And maybe none of them are the best choice, maybe it all depends on what you want to tank. Here I'll just support you with some easy facts so that you can make that choice yourself.

I thought I'd start out with Blade Warding since I've heard everything from "it sucks" to "it rocks" when asking people about it. I do agree that in the beginning I thought it had a low proc chance, but recently I feel like it must've gotten a "stealth" buff? (Maybe there was a post about it, but nothing I've read). It procs more alot often than I felt it did before, so I decided to write down some numbers for some hobby-number-crunching of myself. Now these are very simple figures and won't give you any exact facts but an easy overview about what this skill seems to do.

First of all, what does it do?
Permanently enchants a weapon to sometimes grant Blade Warding when striking an enemy. Blade Warding increases your parry rating by 200 and inflicts 600 to 800 damage on your next parry. Lasts 10 sec.

This ability can stack up to 5 times. 200 parry rating is like I said about 4%. 4,4% to be exact. Getting a stack of 5 would therefore give you more than 20% parry, which probably will double your parry. Some people say that they've removed it's stackability, but I can tell you that it does still stack, at least 3 times (never gotten more than that). It doesn't say however that the buff is removed if you parry. But let's look a little bit closer at the tooltip before we go to the numbers. Since it procs from "striking an enemy" it will be somewhat better for a warrior than a paladin. Prot paladins don't, as of yet but that will change in Cataclysm, have an instant strike without cooldown like Devastate. Every strike they do, except their autoattacks is on cooldown, which is actually something that annoys me with palatanking. That means a paladin will do alot less striking than a warrior, who not only has devastate, but heroic strike too.

This of course depends on what Blizzard actually mean by "striking". If they mean "deal damage to" I'd say the situations are reversed. With spells like Consecration and Hammer of Righteousness, a paladin probably spanks out more damage attacks (not necessarily more damage though, notice the difference), than a warrior. This enchant has only been tested on a warrior by me, so I can't say for sure what skills will proc this, since warriors don't have any magical attacks. I strongly doubt Blizzard would create an enchant that would be practically worthless for paladins, so it probably procs off of any direct damage (but probably not dots). But in any case that means these data are best applied to warrior tanking.

This also means of course that if you're stunned or feared, this won't proc at all. Any time you're unable to strike your opponent, this will be worthless. That is also worth keeping in mind.

So let's jump to the numbers, shall we? Now, there aren't any addons that track procs anymore, since Procodile isn't maintenanced. It doesn't work at all, and I haven't found a good substitute (I'd love a tip if you have one!). These figures were therefore collected by me counting procs manually, and since I tank while I do this I admit they might be off by one proc here or there. I had an addon play a loud noise everytime Blade Ward procced, so I definitely noticed it though. Afterwards I checked my time in combat using Recount, and also checked how many times I had actually parried with a Blade Ward proc. This is easily seen as it deals damage when that happens. That way I can see it's actualy usefulness, although that is completely dependant on how much I am being struck of course. Rough estimations, here I come.

I did 6 different instances, with completely different "combat situations" (tanking alot of packs or tanking alot of single mobs). The instances were randomly chosen as I simply queued for a random pug to collect these data. Keep in mind also that not all time in combat actually means fighting anything. In instances like HoR you're kept in combat by the event although there isn't anything to fight, this is true for some other instances as well. Each combat time is therefore a little bloated, which subsequently will diminish the ppm figures. Keep this in mind. I'm rounding the figures to one decimal. Oh and also remember that I suck at maths.

- PoS: 13,3 min in combat + 23 procs (15 parries) = 1,7 procsperminute (1,1 parriesperminute, due to Blade Warding procs).
- HoL: 10,5 min in combat + 16 procs (7 parries) = 1,5 ppm (0,6 ppm)
- CoS: 11,1 min in combat + 10 procs (6 parries) = 0,9 ppm (0,5 ppm)
- HoR: 11,6 min in combat + 21 procs (16 parries) = 1,8 ppm (1,4 ppm)
- UK: 9,1 min in combat + 16 procs (8 parries) = 1,7 ppm (0,9 ppm)
- VH: 7,8 min in combat + 7 procs (4 parries) = 1,1 ppm (0,5 ppm)

Looking at these data, a rough estimation would be that it has a ppm of 1,7 on a warrior tank going about his business. There are some drops at CoS and VH, which might just be unlucky or due to some nature about the combat situations in those instances. CoS for instance is quite "unfair" as you're held in combat by those pesty zombies about the whole instance, although you don't actually strike many of them. So when removing this one faulty data, the numbers seem pretty clear.

But those are only proc times. How many times did the Blade Ward proc actually make me parry? Yet again removing the CoS data, a mean value of the actual parries would be 0,9 parriesperminute. An effective extra parry per combat minute (or at least close to it).

What conclusions can we make from this? Well, we can't do much comparing as of yet, since we really have to test Mongoose before we can say anything. Blood Draining is only useful when you drop below 35% hp, and you rarely do that in a heroic. That shouldn't happen to a tank in raids either of course, but we all know that's an ideal hard to live up to at times. One problem with my data are that they aren't tested in a raid setting, although I honestly believe the figures would be about the same, that is about one extra parry per minute. Since parry still removes 100% damage, the overall hpgain in a raid would be more than with Blood Draining, for sure, considering raid bosses hit for thousands of damage. But yet again, there will always be a chance that you don't parry at all, even with several stacks of Blade Warding up. Blood Draining will definitely be there (with its pathetic heal) when you need it the most.

One parry per minute doesn't sound like much, and in a heroic it isn't either. Considering Mongoose gives agi, which also is crit I'd rather go for this one if you intend to mostly do heroics. For raids, one extra parry per minute could be quite some hp considering how hard raid bosses strike. For raiding, I'd actually go with Blade Ward over the other two, as it looks now. With the above numbers you might be able to make your own opinion in the matter. I will return with datas on the other enchants too however (and some raid data on this one as soon as I can), and maybe I will come to another conclusion then.


  1. interesting. I am trying to compare the overall effectiveness of blade warding vs mongoose. Agi rolls alot of tanking stats together. Dodge, Crit and Armor. You get about 2.5% dodge, 2.5% crit and 240 armor which is nothing to sneeze at. But I haven't looked up the PPM of it either or it's uptime, which would be very important.

    anyway, good looking up :)

    tanksalot/benevölence - silverhand

  2. Yes, I'll definitely look at Mongoose as well, which I currently use on my Paladin (and my paladin and warrior use the same weapon, Rimefang's Claw from PoS hc). Since they proc from the same thing, namely weapon swings, this matters of course. And as you say, the agi from Mongoose gives alot of good stats. I'll do a post on Mongoose in some days, just have to collect some data first :)