WAR Stats: Debuffing Initiative

Thanks to a comment in my last WAR Stats post from the venerable TzuDevil I'm going to discuss Initiative debuffs and their effect on your chance to be crit.

TzuDevil's comment was that debuffing Initiative is HUGE. He's right. Here's an example:

I am aloofly striding into battle with my Swordmaster and start attacking the insidious Evillyn. Using Nature's Blade enchantment that is core to all Swordmaters and with points in the right tree I have a chance on any attack to steal 120 pts from a random stat, including initiative.

A reminder that the formula before bonuses from items for Chance to be Crit is 35/Initiative



Evillyn has an initiative of around 200 and one -3% Crit item on her Armour Set giving her a base chance to be crit of about 15%. I would say this is typical for most people.

I hit Evillyn a few times, get lucky and steal 120 points from her initiative reducing it from 200 to 80. Her chance to be crit has now increased from 15% to 45%! If she gets focused by DPS, even just a little, she'll probably cark it.

The good news for Evillyn is the SM initiative debuff is a bit random. The bad news is that a Witch Hunter can debuff Initiative too. Blessed Bullets of Purity debuffs Initiative by 80. The Archmage has a 80pt debuff called Mistress Of The Marshes.

By the same yardstick, buffing your own or group's Initiative is very advantageous. It's all to do with the steepness of the curve at the low end. You don't want to be sitting on the steep bit, it's treacherous. Unless you stack Initiative up to around 300 a debuff is really going to hurt, so protect yourself and your allies by buffing it if you can.

Conclusions
If you hold true to the maxim that Bursts Kills and Crit FTW! then do what you can to debuff your opponent's initiative - it'll make a difference.

About Swordmasters
For a Swordmaster looking to assist the DPS classes the potential for Nature's Blade to debuff Initiative by 120 points makes it very hard to overlook in my eyes, and it's actually a stat steal so also buffs your entire group by the amount it steals. Very useful. The Black Orc has the mirror for Destro. It doesn't just steal iniative either, it can steal any stat.

People will argue that a Knight can do the same job of increasing crit chance much more reliably than a Swordmaster, and they may well be right. For example, the KotBS has a tactic called Dirty Tricks that increases their group's Crit chanceby 10% for 10s every time they block. This is more reliable when compared to Nature's Blade, which is more random but has a higher reward. The odds are hard to calculate but in a warband with 3 or 4 SMs all running Nature's Blade I would think that there would be enough initiative getting debuffed it would leave your enemy very vulnerable to crits, not to mention stealing and buffing your group with all the other stats.

WAR Stats: This is Critical

In previous posts I have looked at damage totals from combat logs but this ignores something very important - the heat of the moment, what can happen in just a few seconds of combat. There is one thing I dread to see flash up on my screen and that's a string of big crits hitting me in quick succession.

Looks like this...

-1050
-1637
-1320

When I see just one big number like this I think "oh shit", nevermind three at once. That's over 4000 damage in the blink of an eye. Numbers like this are enough to kill you dead if you are already at low health. When people see you with low HP they go for the kill and you get focused. A string of big crits will either kill you straight off or leave you vunerable and put pressure on healers. Things tend to go awry.

DPS classes are crit-crazy. Go ask one. I believe BWs and Sorcs can get there crit chance up to 75%, other DPS classes to around 50%. They love the stuff.

On top of what your opponent is stacking you also have a base chance to be crit, based primarily on your initiative that has nothing to do with you opponent.

At Level 40 the formula for 'Chance to be Crit' = 35 / Initiative
E.G.
Initiative @ 100 (35/100) = 35% chance to be crit
Initiative @ 200 (35/200) = 17.5% chance to be crit
Initiative @ 300 (35/300) = 11.6% chance to be crit
Initiative @ 400 (35/400) = 8.75% chance to be crit
Initiative @ 500 (35/500) = 7.0% chance to be crit

Most people's initiative is around 200 so a 17.5% chance to be crit. This means that a Sorc with their own +75% chance to crit will actually have a chance to crit you of 92.5%!! This is scary.

Reducing your base chance to be crit to 0% is not too difficult but what's the point if a Sorc still has a 75% chance to crit you?

It's the chance to not be Crit multiple times in a row that's important. You can survive one big crit. Surviving three is a different story.

Imagine a situation where the DoTs have added up and you have 4000 HP left. A Sorc or two focuses on you with +75% Crit. Added on top of your 17.5% base chance to be crit you will be crit 92.5% of the time.

The probability you will be crit twice in a row is 92.5% x 92.5% = 86%
Thrice in a row = 92.5% x 92.5% x 92.5% = 79%. You will survive 2 out of 10 times.

With a base chance to be crit of 0% you will get crit 75% of the time. The probability you will be crit twice in a row is 75% x 75% = 56%
Thrice in a row 75% 75% x 75% = 42%. You will survive 6 out of 10 times.

By reducing your base chance to be crit from 15% to 0% you are 3 times more likely to survive focus fire. Not only are this but there are lots of nasty procs and debuffs linked to critical hits that you will avoid.

-Crit also obviously gives you a direct overall damage reduction. For most classes a crit is 50% more damage than a non-crit. If a normal attack hits for 400, a crit will hit for 600. Therefore, a 10% reduction in Crit chance = 5% reduction in damage. Sorcs/BW however do 150% damage per crit, so a 400 attack will hit for 1000 when it crits. A 10% reduction in Crit is therefore equivalent to a 15% reduction in damage versus a BW or Sorc. As I have shown in a previous post, playing a SM and being in the front line Sorcs do me a lot of damage.

Getting to 0%
If you look again at the formula and examples above of initiative and base crit chance you'll notice a diminishing return on the stat. Going from 100 initiative to 200 reduces your Crit chance by 17.5% but going from 300 to 400 only reduces it by 3%. I would advise getting to around 250-300 initative and get the rest from -Crit% on items. Most armour sets have a piece that gives -3% Crit so mix and max from different sets you want to use to get the best -Crit.

Conclusions
Recommended.

WAR Stats: Getting Disruptive

In my last WAR stats I had done some testing of Disrupt and Parry, the chance to Disrupt a Caster attack and the chance to Parry a melee attack.


I chose Disrupt specifically because as far as I know the chance to Disrupt does not depend on your positioning, unlike Parry and Block which only defend from the front. You can Disrupt an attack whether it hits you in the front or back so it makes a very good test case for working out how much your enemy's offensive stats will reduce your defensive stats.

Last time I tested a build that had 18% Parry and 18 % Disrupt (0% Block) and my conclusions were mixed. My Disrupt rate was reduced by 12% from 18% to just 6%.
I have now tested some different amounts of Disrupt and Parry.

The value on the left is what was shown on my character screen, on the right the amount actually avoided from looking at my combat log over 1000s of hits.

The results for 18% and 25% Disrupt show that a 12-13% reduction can more or less be considered as the typical reduction to your defense, so whatever your character screen says take 12% off that. Some opponents will reduce that more, some less. I know I have access to two tactics as a Swordmaster that can reduce my opponent's chance to defend by 10% each. I'm not familiar with Caster tactics but I have seen high level set bonuses and glyphs each with 4% chance to disrupt on them so I would not be surprised to see people reducing my defense as much as 20%. Then there are the "Undefendable" proc items.

The testing of 10% Disrupt and 10% Parry also shows that there seems to be a minimum amount to which your defense can be reduced. I think this might be based on your actual stat and disregard the modifiers.

There's are some equations out there for working this stuff out...

(defenders willpower or weapon skill / attackers intelligence or strength) * 15 + defenders '+ defense' bonuses - attackers 'reduced chance to be defended' = Disrupt % or % Parry

My willpower is 150 so assuming an attack from a 1000 Int. Caster with no modifiers taken into account on either side my Disrupt % would be 2 to 3%, and similarly my Weapon Skill is 350 vs 1000Str = 5%.

So I'm guessing that your chance to disrupt or parry can't fall below the first bit of the equation: (defenders willpower / attackers intelligence) * 15

A Closer Look at Parry
Parry isn't as straight forward as Disrupt due to the positional considerations and that both high and low Str people will be hitting me so I don't want to over analyse it. But (But!) looking closer at the attacks parried against certain classes when I had a 18% base chance, I parried around 10% attacks from Black Orcs compared to only 2% from Witch Elves.

It's also important to note that an attack will actually give your enemy a Parry bonus if they have higher Weapon Skill than your Strength. So a Black Guard with 500 STR attacking a Slayer with 700 WS will actually give the Slayer a 21% chance to Parry before any modifiers.

Conclusions

I had really wanted to find that avoidance and Disrupt in particular were worthwhile investments. There is a lot of guff on forums about avoidance being "completely useless" and if I'm honest wanted to do some testing and prove this wrong. I believe people often latch on to a popular idea and it becomes accepted as the "truth" without the facts actually being looked at.

But from these latest results I now think avoidance isn't something you can count on as a reliable defense UNLESS you can get you're actual in-game avoid chance something like 50%.

The positional requirements of Parry and Block also make me question the merit of these. In a 6v6 I can imagine positioning would be easy to keep up but when you start getting 12v12 (e.g. scenarios) you often get mobbed and attacked from all sides and Block/Parry won't make much difference. Them Witch Elves just love to get behind you! And I notice that I'm getting AoE attacks from Choppas from all directions!

I still think Disrupt has some hidden value due to the potential to avoid large spike damage from Sorcs, especially for the ever popular 2H Tank . If I could stack up my base Disrupt chance towards 40-50% I would. At the moment I sacrifice too much to get there but Warlord has a +5% Disrupt set bonus that I'll be interested in when I reach that level.

The Swordmaster does have access to a few abilities to increase Disrupt. Aerythic Armour is +5% and Calming Winds tactic gives +10% in Improved Balance and +20% in Perfect Balance. Plus Hold the Line or Wall of Darting Steel are channelled abilities granting +45%/+50% Disrupt. With these I can get my Disrupt to +95%/100% and I notice this is very, very effective. It Disrupts a lot!!

I am still interested in doing more testing of avoidance. I will try out some of the Swordmaster's Disrupt tactics mentioned above and I will try a maximum avoidance build too (e.g. shield + tactics).

But I am also soon to be taking a look at initative and reducing your chance to be crit. I think this may be the answer I'm looking for. Stay with me ;)

WAR Stats: Avoidance and Onions

I've been having a good 'ol think about combat and avoidance for some time now and have done some testing in-game. It's been a bit like an onion really. There are many layers to peel and it makes my eyes cry so I'll be breaking it down into a few parts.

In this part I'll be looking at some basics and testing a balanced Parry & Disrupt build. In part 2 I'll go all out and stack more avoidance and throw a big shield into the mix and see how that fares. I want to look at how avoidance works with Guard damage too. Once I'm done with testing I hope to offer some insights to the worth of avoidance vs. mitigation as there is much more to avoidance than just reducing overall damage, such as avoiding CC and damage spikes.

Basic Mechanic
Stats and Items give a character a percentage chance to avoid attacks.
Rok did an awesome job of posting the theory of combat in his post 'From Mace to Face'. Check it out for a more in-depth look at combat. I thank him for the loan of the equations :)

Weapon Skill gives a % chance to parry a melee attack. My modest 350 Weapon Skill gives me a base chance to parry of 7.4% on the character screen.

Similarily, Willpower offers a % chance to disrupt an attack from a Caster and Initative gives a % chance to Dodge ranged attacks.

Block gives a % chance to avoid any type of attack. A decent T4 purple shield with a block rating of 370 translates as a %Block value of 21.1% on my character screen.

However an attacker's offensive stat (Str, Int or BS) reduces the defender's chance to defend. A melee attack made with 1000 Strength will reduce my base chance to block from to 21.1% to 14.8% and my chance to parry will drop from 7.4% to 5.3%.

There are other bonuses in-game (Tactics, Items, Buffs) that influence the chance to attack and defend, e.g. +5% to Block or 5% reduced chance to be Evaded. 2-Handed weapons reduce chance to defend by 10%.

And very importantly Block and Parry are only useful when the attack is made to your front, a 180 degree arc I think.

Please TheoryHammer! Don't Hurt Us!
As you can see there is a lot going on with avoidance and a base chance to avoid on the character screen of 10% may or may not translate to that in combat. So on to some testing...

Through a combination of gear, tactics and renown points I got both my Parry and Disrupt to 18%. I ran with a 2-hander so my Block chance was 0%. I didn't use any abilities which provided a buff to these stats.

I specifically chose Parry and Disrupt to test as Parry is only effective against Melee attacks and Disrupt is only effective against Casters. The two can be kept seperate. Also, as Parry is positional and only defends against attacks from the front and Disrupt defends from all directions I can get an idea how much Parry I lose out on by getting attacked from behind. I must add that the current situtation in-game will dictate my positioning rather than trying to keep my attackers in front of me all the time. If I need to move to defend an ally in some way then I usually will even if that means exposing my back (not to a witch Elf though!!).

Results with 18% Parry and 18% Disrupt

• I was attacked 1260 times by Casters and Disrupted 85 times. 6.7% Disrupt Rate.
• I was attacked 1478 times by Melee and Parried 81 times. 5.5% Parry Rate.

Furthermore from Initative I had about an 8% chance to Dodge. I was hit by a Squig Herder 123 times and didn't Dodge any of them!! 0% Dodge Rate.

This is really pretty depressing at first glance. It appears that approx. 10-12% of your avoidance is wiped off your base stat straight away by your atatcker's offensive stats and gear.

I have had a go at working out the amount of potential damage I avoided. I would appreciate some comments on this comparison as it's not an easy one.

Looking just at the Sorcerer for starters, those bastards hit me 407 times and I disrupted 26 hits (6.4% avoidance). I can see which individual abilities I avoided so can make a good estimate of the damage those 26 Disrupts saved me from. Sorcs did 154,352 damage to me overall and I estimate that I avoided an extra 13,826. That's an overall damage reduction of 8% or 34 dmg per hit. If I've got this right I would need ~170 extra Toughness (34*5)to get the same overall reduction in damage, which is quite a bit actually. 34 renown points buys you 102 Toughness. A tactic slot all Tanks get called Rugged gives 160 toughness.

With the Sorc being a high DPS class I also took a look at the damage I avoided from wimpy Tanks. I was hit by Tanks 675 times and parried 38 hits (5.6%) avoiding an estimated 4067 damage, a reduction in damage of 5% or 6 Dmg per Hit. This is equivalent to the overall damage reduction I would gain from an extra 30 Toughness.

Conclusions
A common perception in Warhammer is that avoidance is worthless in PvP as your chance to avoid an attack gets reduced far too much by your attacker's setup. I've demonstrated this is true to an extent with my both my Parry and Disrupt rates getting reduced from 18% to ~6%. But by comparing the potential damage avoided against a Sorc and against Tanks it suggests the benefit of avoiding an attack get proportionally better the higher the damage of the attack. This is a bit like in PvE when Tuten McBosspants is hitting you for 8000 per hit - it's better to avoid that hit entirely than just mitigate it down to 7500 or something.

I'm starting to think now that avoidance might be the best path for me. In my other WAR Stats posts I've shown that it's the high damage attacks from the likes of Sorcs and Witch Elves that do me the most harm, and I know from experience it's getting big hits in quick succession that gets me killed in a hurry. Having a chance to avoid these big hits rather just soak them up sound good to me. Plus there's the intangilble benefit of avoiding some CC or a nasty armour debuff.

Quite a bit to chew on. I hope this has not made your eyes water and makes some sense. Next up I'm going to try and stack even more Disrupt and see if the extra X% that I stack on top of the 18% I've tested here translates to an extra as X% in-game.

~Aeo

WAR Stats: Hard Hitting

This is a look at the abilites which do the most damage per hit on my Karak Norn Swordmaster. From my combat logs I have totalled up the damage done by each ability and then divided this by the number of times the ability actually hit me.

The numbers are before any mitigation from Toughness, Armour or Resists to give them more relevane to all classes, rather than slanting the numbers with my high tanky mititgation.

I've only included abilities which hit me more than 20 times (out of 2000 total hits). There are certain abilities, such as Morales, which hit hard but they might only have hit me once or twice so I'm not counting them as typical.




Remeber this is average damage per hit and before any mitigation. It's an average of all normal and all critical hits together and this can smooth over the potential actual damage some of these abilities can do.

For example, from my logs I can see Agonizing Wounds can crit for ~1500 (before mitigation) and, as it ignores armour if struck from behind, I will often take ~1000 actual damage from an Agonizing Wounds crit.

Saying that though, Sorc crits regularily do me for closer to 2000 actual damage. I mitigate about 50-60% of Sorc damage meaning the total hit before mitigation would be scarily around 4000. I dread to think what this would do to a low toughness class with debuffed Spirit resistance.

Conclusions: Ouch and Double Ouch
As I've been posting my other results and seeing the Sorcs taking the damage biscuit I've been mindful that a lot of this damage could be so called "meaningless damage", AoE or DoTs, that are easily healed away, and it's the burst abilites that get the real job done.

The results show that Sorcs has the potential to inflict more single big hits on me as well as most overall damage. I have to stop them somehow, for the greater good.

I have been tailoring my spec a bit for greater survivability vs. Sorcs. and it has worked out well, but I had been holding some back until now as I was fearful of losing to much to melee classes, particularily the Choppa and Witch Elves.

Swordmaster's were initially portrayed by Mythic as the "anti-magic Tank" and there are a few useful abilities for this which are often ignored on the SM forums.

I am going to explore further a Sorc-Killing role for my SM.

WAR Stats: Damage Received per Class

The chef at the Blurring Shock cafe is all out of pie but has cooked up another tasty treat. This time it's damage bars (may contains nuts).

Following my look at the damage I received in game by damage type (i.e. Spirit, Physical, etc) I have now looked at the damage done to me by each class.

My reason for doing this is to make me more survivable in RvR, and to make me a better tank. One of my favourite things about WAR is the wealth of different stats available to the player, offensive and defensive. With such variety it isn't always easy knowing which stats are the best - hence a closer look at the numbers.

I can't easily determine who auto-attacked me. In the combat logs there are lots of hits simply described as 'attack'. I could go through it by hand and figure each one out but it's a ball-ache so I'm not going to. In the data below auto attacks from all classes are listed together as 'All auto-attack' and, as far as I know, all auto-attacks do physical damage only.

The results are before any mitigation from armour or resists. I play a Swordmaster on Karak Norn and as was pointed out last time the type of damage received will depend on who attacks me. As a tank I am naturally in the firing line.


This doesn't mean that a Chosen and a Choppa do the same damage, but that these classes did the same damage to me over a couple hours gameplay. I can't tell how many Chosen or Choppas there were out there. One would assume that there were more Chosen. Remember too that these values are before mitigation so Chosen would've done more actual damage to me than Choppas as my spirit resist (40%) is lower than my physical (75%)

Conclusion: Can't Ignore the Sorc
There is a lot more to PvP than total damage. Burst damage might not amount to much in total but it puts prssure on healing and kills. In my search for a better defense however, I can't ignore the amount of damage done to me by the Sorc. +Block and +Disrupt and good Spirit and Corp resists will be important.

There is more work to be done! Next week I am going to have a look my avoidance stats (Block/Parry/Dodge/Disrupt) as well as how much damage I take from Casters vs. Melee vs. Ranged.

I'll need to look at individual high-damage abilities at some point soon too.

~Aeo

WAR Stats: Damage Pie meets Premade

I was fortunate enough :p last night to PuG some scenarios and run into Destro premades on KaraK Norn and along with my fellow Order PuG-in-arms proceeded to get the mauling I deserve for going solo.

My damage pie from yesterday was based on what I would consider typical combat, verus all types of groups, and including scenarios won and lost, oRvR, skirmish, keep battles and "zergfare". I thought it would make an interesting excercise to compare this with the damage dished out by the premades.

This new damage pie (now with extra cherry!) is from 5 scenarios. Destro were very, very heavy melee and well backed up well by healers, with the odd caster or ranged. This was the distribution of classes across all 5 scenarios.

DoK 10, Chosen 10, Choppa 6, Shaman 6, Zealot 6, Witch Elf 6, Black Orc 5, Sorc 5, Squig Herder 4, Black Guard 4. Note: no Marauders or Magus!

Below left is a breakdown of damage type collected from the whupping I took in these 5 scenarios compared to the more general "overall" damage on the right that I posted yesterday.

Conclusions: There is more than one way to skin a Tank
It'll be stating the obvious to a lot of people but the type of damage received will vary depending on who you are up against. I knew this but I wasn't quite expecting it to be that different and ironically, since yesterday I had re-geared/specced slightly to have a better defense against caster and magic damage. My mistake!

The lesson I will learn is make sure I have different options available to me and make decisions based on who the opposition are. Having an armour pot handy against melee trains for example, some +disrupt rings against caster heavy groups.

~Aeo

WAR Stats: Damage Pie

I was messing about with the different armour sets I've got for my Swordmaster (Anni/Sent/Conq) and trying to figure out which Set Bonuses to plump for.

I could not look past the 3 Piece Bonus of +320 Spirit Resist on Annihiator and the 2 Piece Bonus of +344 Corporeal Resist without wondering exactly how much more defense this would give me over the other defensive stats, like +66 Toughness Bonus from Sentinel.

So I decided to parse my combat logs to see exactly how much whack I took from the 4 different damage types.

Here's the results.

This is total damage inflicted on me (a Tank) during 45 mins combat time before any mitigation or absorbs are taken into account.














How I did this
I used combat log parser and saved the file as text and imported that into excel. I then sorted the spreadsheet by ability name and created a pivot table to sum all damage from each ability. Then the tedious bit. I looked on WarDB at each ability and made a note of the damage type. Then added it all up. I'll be happy to provide better instructions to anyone who's interested.

My Conclusions
I was suprised to see how much damage came from magic compared to physical, thought it would've been more like 33%:66%. Also, just how little elemental damage I took. No need for +Elem on items then but definely need to keep the Spirit and Corp high.

There's a hell of a lot more useful information to come out of this and I'll be making some more posts in the coming days, things like the amount of Burst and Crit damage, the amount of damage from each archetype and my %mitigation.

Wait 'til you see the breakdown of damage per class and just how much Sorcs accounted for! Stay tuned for that one.

~Aeo