Warriors and Paladins Block About the Same in Build 8905

All things related to the expansion

Moderators: Fridmarr, Worldie, Aergis

Postby Garath.Gorefiend » Sat Sep 13, 2008 7:06 am

Sartor wrote:
Tisiphone wrote:http://wotlk.wowhead.com/?spell=2565


Buff:
Shield Block
Block chance and block value increased by 100% for next attack.
10 seconds remaining


Outdated!

Now blocks ALL attacks for 5 seconds, and increase all damage blocked by 100%. ;)

*cough* check the stickies *cough*


Not sure where you got this...MMO champ, Wiki and all the rest say:

Blocks all attacks for 10 seconds and increases all damage blocked by 100%.

Talent improves it from a cooldown of 1 min to a cooldown of 40 seconds.
Garath.Gorefiend
 
Posts: 406
Joined: Mon Aug 25, 2008 2:07 pm

Postby Ferrix » Sat Sep 13, 2008 7:30 am

The only problem I have with warriors blocking as much as us is, warriors are still slightly ahead in mitigation, are getting more avoidance on their gear, and are still ahead of us in terms of tanking casters.
For the most part, we can still tank, but in a min/max guild, the warrior is going to be the preferred tank unless something changes.
Ferrix
 
Posts: 77
Joined: Wed Jun 25, 2008 1:10 pm

Postby moduspwnens » Sat Sep 13, 2008 7:51 am

Exewu wrote:I actually have different conclusions than you. I found that for similar gear paladins mitigate a lot less melee damage than warriors. Mainly because paladins have slighlty less avoidance (difference in tier sets) and RF is only 6% while warrior's defensive stance is 10%.

http://rapidshare.com/files/144372560/M ... k.xls.html

linky to the spreadsheet.


Hehe, well firstly, I'm referring ONLY to blocking. My findings state "in the same gear" they "block about the same amount of damage" at common avoidance levels.

You're also calculating Shield Block incorrectly. This is a similar mistake I was making early on. A Shield Block with no talents or abilities active is for 1 * your block value. A Shield Block with Shield Mastery (30% block value) is for 1.3 * your block value. A Shield Block WITH Shield Mastery AND Shield Block is for 2.3 * your block value. Critical blocks double these values.

Over in Cell A21:
=((Unmitigated hit*(Reduction by Armor)*Modifier from Defensive Stance)-Block Value * 2 * 1.3)*Avoidance

That'd be an error that favors the warrior, and would close the gap a little, but I also can't find a Critical Block option, which would favor the Paladin. I didn't see Redoubt in there. Did you calculate the stats on the gear to find that we'd be uncrushable in it with Holy Shield up?

I need to build in actual hits and DR to the simulator.
I rule.
moduspwnens
Moderator
 
Posts: 6211
Joined: Tue Nov 06, 2007 12:28 pm
Location: Shattered Hand

Postby Macha » Sat Sep 13, 2008 7:56 am

Ferrix wrote:are getting more avoidance on their gear,


This is a non-issue, because we can just use offset items with equal or higher avoidance.
Also, at the beginning we benefit more from knocking Hit out of the attacktable than from just stacking dodge.

But yes, a warrior right now is clearly the prefered tank. But remember, we have more changes to come - two new "cool deep prot talents" for once.

Mainly because paladins have slighlty less avoidance (difference in tier sets)


Why do people still say that? You can use non-tier set items, people. It's really not a problem.
Last edited by Macha on Sat Sep 13, 2008 7:57 am, edited 1 time in total.
Macha
 
Posts: 625
Joined: Sat Jul 26, 2008 12:50 pm

Postby moduspwnens » Sat Sep 13, 2008 7:57 am

Ferrix wrote:The only problem I have with warriors blocking as much as us is, warriors are still slightly ahead in mitigation, are getting more avoidance on their gear, and are still ahead of us in terms of tanking casters.
For the most part, we can still tank, but in a min/max guild, the warrior is going to be the preferred tank unless something changes.


Well, right now, they have 4% more flat DR, more health, a ranged slot for (at least throughout TBC) better tanking stats, are blocking for about the same amount (from what I've found), and have several gimmicks (Disarm, Fear break, Spell Reflect). They're not finished with tanks yet, though.
I rule.
moduspwnens
Moderator
 
Posts: 6211
Joined: Tue Nov 06, 2007 12:28 pm
Location: Shattered Hand

Postby Murbella » Sat Sep 13, 2008 1:04 pm

Macha wrote:
Ferrix wrote:are getting more avoidance on their gear,


This is a non-issue, because we can just use offset items with equal or higher avoidance.
Also, at the beginning we benefit more from knocking Hit out of the attacktable than from just stacking dodge.

But yes, a warrior right now is clearly the prefered tank. But remember, we have more changes to come - two new "cool deep prot talents" for once.

Mainly because paladins have slighlty less avoidance (difference in tier sets)


Why do people still say that? You can use non-tier set items, people. It's really not a problem.


I'm kind of confused why you think it is a good idea for paladins to have to use offset gear "designed for warriors" to get the same avoidance as warrior tier armor which would likely have set bonuses on it... Wouldn't the simple solution just be to make them equal in the first place (i haven't looked over the stats a ton, but from glancing over them, i dont think i would use the paladin set over the warrior given the choice)?
User avatar
Murbella
 
Posts: 85
Joined: Wed Apr 16, 2008 8:38 am

Postby Garath.Gorefiend » Sat Sep 13, 2008 5:06 pm

Well, frankly, I cant agree.

Warriors do not block similarly to Paladins. Their mechanic is superior in all aspects

I ran full simulations:
Assumptions:

Armor mitigation at 63.7%
Pure Avoidance: 53.78% - figure this should be reasonably attainable
Blk Chance: 20.8% (25.8% for warriors - they get the good talents)

Now, the simulator works in HS and Redoubt procs..basically casting HS every 8 seconds and checking for a redoubt proc on any damaging hit.

For warriors, it spams Shield block once every 40 seconds.

Assumptions for Warrior shield block is that it blocks for 2 * block value when active, so if the warriors tool tip shows he blocks 553 it assumes 1106

Every successfull block has a 30% chance to double the amount blocked, so Critical blocks when Shield block is up means a block of 4* the normal blocked amount.

Basically, what the simulator shows is the warrior outperforming the paladin at every level, regardless of the number of mobs attacking.

I broke this down, and on a fixed cycle of 250,000 attacks, the warrior had about 79,800-79,000 blocks on average. However, their number of effective blocks based on the block amount multiplied works out to around 123,800 blocks. In other words, it more than offset the number of hits that got through.

For the record, the paladin, single mob on that scenario blocked about 115,400 attacks (no hits get through).

I then ramped up just the block rating. And, as I surmised, the paladin did not improve.

The warrior, on the other hand, continued to benefit from the critical block mechanic even once normal hits were totally pushed off the table (the point when redoubt and HS become pointless).

In other words, the warrior mechanic continues to improve as gear gets better. The paladin one suffers.

The other thing truly noticeable about the warrior mechanic, is that it is exactly flat line. The number of mobs attacking the warrior is irrelevant, he will block exactly the same amount of attacks on average and get hit the same number of times.

The paladin mechanic, however, deteriorates in AoE situations. The more mobs attacking, the more normal hits get through.
Garath.Gorefiend
 
Posts: 406
Joined: Mon Aug 25, 2008 2:07 pm

Postby moduspwnens » Sat Sep 13, 2008 5:19 pm

Garath.Gorefiend wrote:Assumptions for Warrior shield block is that it blocks for 2 * block value when active, so if the warriors tool tip shows he blocks 553 it assumes 1106


THIS is where it is wrong. You'll notice on page 1, I assumed the same thing, and came out with the same results, but it is wrong. 100% increased block value is NOT the same as blocking for double. Skim through the thread for where I found this out.

Let's say you block with no talents or buffs. You block for 1 * Your Block Value. Now, add in Shield Spec/Mastery. You're now blocking for 1.3 * Your Block Value. When you add in Shield Block, you do NOT block for (1.3 * Your Block Value) * 2. You block for 2.3 * Your Block Value. This is a significant difference. Critical blocks DO work that way, but 100% increased shield block value is NOT the same as blocking for double.

Garath.Gorefiend wrote:Every successfull block has a 30% chance to double the amount blocked, so Critical blocks when Shield block is up means a block of 4* the normal blocked amount.

Again, not accurate.

Garath.Gorefiend wrote:Basically, what the simulator shows is the warrior outperforming the paladin at every level, regardless of the number of mobs attacking.

This is the same thing I found (incorrectly) on page 1, and it's because Shield Block and Critical Block both have nothing to do with charges. The number of mobs for it won't change, but Holy Shield will only get worse.

Garath.Gorefiend wrote:I then ramped up just the block rating. And, as I surmised, the paladin did not improve, except in multi-mob scenarios and once normal attacks were totally pushed off the table, he straight-lined.

Yeah, once you start eating into the Paladin's advantage (30% increased block at all times), the warrior benefits.

Garath.Gorefiend wrote:In other words, the warrior mechanic continues to improve as gear gets better. The paladin one suffers.

Yes, but I wasn't too concerned about that. The Paladin will already have much better AoE threat, and I've yet to see a challenging AoE tanking encounter (mitigation-wise).

Garath.Gorefiend wrote:The paladin mechanic, however, deteriorates. The more mobs attacking, the more normal hits get through.

Well, yeah, Holy Shield will get worse, but Redoubt scales up and gets better very slowly. Either way, Shield Block and Critical Block win as far as multiple mobs.
I rule.
moduspwnens
Moderator
 
Posts: 6211
Joined: Tue Nov 06, 2007 12:28 pm
Location: Shattered Hand

Postby Garath.Gorefiend » Sat Sep 13, 2008 5:36 pm

hmm, when i have time, I will rerun the simulations with the 30% increase on BV taken out on the critical calculations...however, a quick raw calculation shows that the warriors are still out performin us, simply not by as large a margin.
Garath.Gorefiend
 
Posts: 406
Joined: Mon Aug 25, 2008 2:07 pm

Postby moduspwnens » Sat Sep 13, 2008 5:42 pm

Garath.Gorefiend wrote:hmm, when i have time, I will rerun the simulations with the 30% increase on BV taken out on the critical calculations...however, a quick raw calculation shows that the warriors are still out performin us, simply not by as large a margin.


Alright, upload or give me a link to your simulator and I'll try it out.
I rule.
moduspwnens
Moderator
 
Posts: 6211
Joined: Tue Nov 06, 2007 12:28 pm
Location: Shattered Hand

Postby Garath.Gorefiend » Sat Sep 13, 2008 7:25 pm

Heh, I would, but unfortunately I did not write it in a portable language...right now if I compiled it, it would probably be an 8MB lib with appropriate run-time libs...

Once the talent trees settle down though, I can either re-write it or provide a link to a downloadable version.

I did a quick correction of the multiplicative factors.

Factor on shield block becomes 1.76923 (1+1/1.3) on shield block blocks, 3.53846 (2*(1+1/3) on crit shield block blocks and 2 on critical normal blocks

25% of blocks are from shield block as it would over-ride normal block. Of those 30% are going to be critical, which gives us 22.5% rate for shield block blocks and 7.5% for critical shield block blocks.

Normal blocks are going to occur 75% of the time of which 52.5% will be fully normal blocks and 22.5% will be critical.

So we can assume that the number of effective blocks is
about:

(.175*1.76923)+(.075*3.53846)+(.225*2)+.525=1.55

effective blocks per actual block. This assume that your blocking all attacks (in other words, avoidance plus block rating=102.4).

As your avoidance drops, this number changes..and actually increases because the contribution margin from the shield block mechanic increases, rather than decreases. In other words, if your taking more hits, the number of hits blocked by the shield block mechanic (up 25% of the time) increases in value as compared to the number of blocks generated by your normal chance to block.

Examining my code actually led me to fix two problems. One was the magnification factor, which, as you pointed out, I had wrong.

However, the other mechanic I had wrong, was that I was not giving the Shield Block mechanism precedence over normal block. (was treating it like your calc on HS/Redoubt and only counting it if normal mechanics failed).

The Shield Block mechanism has to be given precedence because it provides higher mitigation values.

Net of it is, given equivalent raid level gear, we are worse off.

Using current beta Naxx 10 gear for a sample, we are running at 50.27% pure avoidance, and 20.11 block rating.

Single target:

We get 49.82% blocks, 20.11% normal, 29.71% from HS

Warriors again have 25.11 block rating.

They get 32.46% blocks and given the calculations above, an effective block rating of 57.153%, note the magnification factor is greater than 1.55 because of their blocks, 17.6% came from block rating, while 14.86% came from Shield block.

But the 17.6% with crits is about 22.88%, and the 14.86% from Shield Block is equivalent to 34.18% normal blocks, thus yielding the very high effective block rating.

Please verify my math, if you would, but unless I am missing something basic, the number show a fairly uneven scenario that I am not liking.

Of course, the simulator does assume that you are never even critically shield-blocking the entire hit.
Garath.Gorefiend
 
Posts: 406
Joined: Mon Aug 25, 2008 2:07 pm

Postby moduspwnens » Sat Sep 13, 2008 7:57 pm

Obviously nobody can check your code if it's a private simulator. One of the reasons I made my simulator like it is, is so all of its calculations can be confirmed. You turn on rolls, time, and effective blocks, and the only thing you DON'T see is the random generation of numbers, which is done through PHP's rand() function (which can be confirmed through the source code, which is available).

It's effectively transparent and simple. What this means is that you don't have to rely on any formulas or theoretical guesses. The only math the simulator actually does is clearly laid out through in-game mechanics (like, 10% chance to proc a 30% block chance for 10 seconds, or the single-roll attack table) or the mechanics for "effective block," which have been clearly proven via screenshots. I did this intentionally to avoid theorycrafting formulas (like you're doing) and making assumptions (like that we'd be uncrushable with Holy Shield).

And, if anything is questionable, it's easy to understand so anyone can make sure it's working. It may not be written in the best language, or be super efficient, but it has those other benefits. I'll probably make a spreadsheet style simulator sometime (in fact, that'd be much more appropriate for PHP), but for now, I don't see any reason to believe what this simulator is doing is wrong.
I rule.
moduspwnens
Moderator
 
Posts: 6211
Joined: Tue Nov 06, 2007 12:28 pm
Location: Shattered Hand

Postby Garath.Gorefiend » Sat Sep 13, 2008 8:05 pm

My simulator is doing the same thing.

Doing rolls for each attack etc..and am using a C rand function. Difference is, since I am running in C doing simulations with 250000 or more attacks happens damn quick

What I noticed in your simulator was what I mentioned in my last post.

Are you prioritizing the shield block mechanic when its up over normal block? (when I look at your simulator, it looks like its not)

The calculations I provided effectively give you a benchmark to check your conlusions against. If your numbers are varying widely from what I put out, then there is either an issue in my implementation of the mechanics or yours.

Any simulation run against enuff samples, will always come out to pure number theory in the end, as the randomness drops out.

My issue was with a few of your statments, in that at no place do I find the paladin mechanic superior to the warrior mechanic once both are in a non-crittable state, given equal gear.
Last edited by Garath.Gorefiend on Sat Sep 13, 2008 8:13 pm, edited 1 time in total.
Garath.Gorefiend
 
Posts: 406
Joined: Mon Aug 25, 2008 2:07 pm

Postby moduspwnens » Sat Sep 13, 2008 8:07 pm

Garath.Gorefiend wrote:Are you prioritizing the shield block mechanic when its up over normal block?


Well yeah, have you tried it? I mean, it's in my signature.
I rule.
moduspwnens
Moderator
 
Posts: 6211
Joined: Tue Nov 06, 2007 12:28 pm
Location: Shattered Hand

Postby Garath.Gorefiend » Sat Sep 13, 2008 8:18 pm

moduspwnens wrote:
Garath.Gorefiend wrote:Are you prioritizing the shield block mechanic when its up over normal block?


Well yeah, have you tried it? I mean, it's in my signature.


Yup I have...here let me paste

Shield Block cast.
Mob 1: PARRY
Mob 1: MISS
Mob 1: BLOCK (SHIELD BLOCK)
Mob 1: MISS
Mob 1: BLOCK (PASSIVE)
Shield Block fades.

Shield Block cast.
Mob 1: DODGE
Mob 1: PARRY
Mob 1: DODGE
Mob 1: BLOCK (PASSIVE)
Mob 1: BLOCK (SHIELD BLOCK) - Critical Block proc.
Shield Block fades.

Shield Block cast.
Mob 1: DODGE
Mob 1: BLOCK (SHIELD BLOCK)
Mob 1: BLOCK (PASSIVE) - Critical Block proc.
Mob 1: PARRY
Mob 1: MISS
Shield Block fades



The bolded lines are examples from your simulator, clearly showing you are prioritizing normal blocks, normal crit blocks over SHIELD BLOCK and SHIELD BLOCK Crit blocks

And as I was trying to say, the various blocks and criticals have differing mitigation values
Garath.Gorefiend
 
Posts: 406
Joined: Mon Aug 25, 2008 2:07 pm

PreviousNext

Return to WotLK

Who is online

Users browsing this forum: No registered users and 1 guest

Who is online

In total there is 1 user online :: 0 registered, 0 hidden and 1 guest (based on users active over the past 5 minutes)
Most users ever online was 380 on Tue Oct 14, 2008 6:28 pm

Users browsing this forum: No registered users and 1 guest