Remove Advertisements

Theck's MATLAB TPS analysis - WotLK/3.x

Warning: Theorycraft inside.

Moderators: Fridmarr, Worldie, Aergis, theckhd

Re: Theck's MATLAB TPS analysis (A Jonesy derivative work)

Postby Venoseth » Mon Aug 17, 2009 11:33 am

Thalastor82 wrote:I wonder at which point will be more useful (avoidance-wise speaking) to gem parry instead of agi/dodge with the new DR calculation for these stats.
It's worth considering, on top of DR calculations, that the speed increase coming for parry could work as a nice TPS boost with the new SoV. Hard to calculate and to compare with the TPS coming from increase crit from Agi, though.


First: this information was posted previously in this thread
The changes to the DRs of dodge/parry means that if you have more than 1.87% dodge for every 1% parry [the #s right from your character sheet, but will be different w/ raid (agi) buffs] then you should use a parry/stam instead of dodge(agi)/stam for maximum avoidance.

Second:
Parry is not a threat stat, in the traditional sense. We can't "gear" for parry, barring a few exceptions (Enchanting your gloves w/ armsman and socketing primarily). That said, even if you could gear for parry for threat, you'd still be losing large amounts of avoidance because of parry's stronger (still) DR; refer to the previous paragraph. ^^

Last:
On the TbtL changes, and spellpower/threat: Tanks are stacking stam anyway, and never for the threat it provided

ie:
Theck's 3.2 Stat Threat Analysis

What does that mean to me? Well, I think that our threat scaling (the 10% nerf to RF's bonus holy threat) going down makes sense as we were ahead by a good margin in most situations. We triple-dip on threat from STR (our #1 threat stat before) w/ a cap to how much BV we can take advantage of (for threat).

So, tanking difficult content? Stam like usual (barring the regular exceptions). Way outgeared your content, or have a threat set? STR's an even better threat stat, and you're no longer losing out on some AoE threat (Consecrate's 32% spellpower coefficient) while not wearing your top stam pieces even tho' total AoE threat may be slightly lower on packs of mobs much larger than 3~, which I'm 100% fine w/. :D

Gearing choices haven't changed all that much. If this is a 5%+ AoE threat loss while maintaining our maximum EH/avoidance/HP/etc then I'd call this fair. :D
"Pleasure in the job puts perfection in the work."

"Education is an ornament in prosperity and refuge in adversity."
Image
Venoseth
 
Posts: 240
Joined: Thu Oct 02, 2008 11:05 pm

Re: Theck's MATLAB TPS analysis (A Jonesy derivative work)

Postby culhag » Mon Aug 17, 2009 4:28 pm

I've updated my RatingBuster modification using the latest 3.2 compatible version and Theck's latest threat stats value calculation for 3.2.

I've also added an EH comparison that really was missing from RB.

Check it out !
User avatar
culhag
Maintankadonor
 
Posts: 1728
Joined: Wed Aug 06, 2008 7:50 am
Location: France

Re: Theck's MATLAB TPS analysis (A Jonesy derivative work)

Postby Thalastor82 » Wed Aug 19, 2009 5:41 am

Venoseth wrote:First: this information was posted previously in this thread

yeah, sorry but I found it only afterwards... I actually put in my favorites the "Some preliminary conclusions concerning avoidance in 3.2" post, I love this site ;)

Venoseth wrote:The changes to the DRs of dodge/parry means that if you have more than 1.87% dodge for every 1% parry [the #s right from your character sheet, but will be different w/ raid (agi) buffs] then you should use a parry/stam instead of dodge(agi)/stam for maximum avoidance

yes, thanks. In my case it happens that, fully raid buffed, I have actually 2.1 as a ratio (once substracted the naked parry and dodge from the percentages), which make more efficient to gem for parry

Venoseth wrote:Second:
Parry is not a threat stat, in the traditional sense. We can't "gear" for parry, barring a few exceptions (Enchanting your gloves w/ armsman and socketing primarily). That said, even if you could gear for parry for threat, you'd still be losing large amounts of avoidance because of parry's stronger (still) DR; refer to the previous paragraph

Well, as we are talking here about fine tuning I think that we could in some way take into account that parry accelerates the next melee attack, which also generates a proc of the SoV.
So, I wonder how many additional TPS (of fraction of TPS, if it is really small) would generate a parry gem instead of a dodge gem on a given boss (say Vezax)

Said that, of course I would not gem for parry if it made me gain less avoidance than Dodge. But in my case I actually have to gem for parry in order to have the optimal avoidance, and I'm wondering how much TPS I will gain from this

BTW, nice consideration about STR being even more useful as a AOE stat in lower content, I didn't think about that
Thalastor82
 
Posts: 20
Joined: Mon Jul 20, 2009 6:27 am

Re: Theck's MATLAB TPS analysis (A Jonesy derivative work)

Postby jere » Wed Aug 19, 2009 6:30 am

Thalastor82 wrote:
Venoseth wrote:The changes to the DRs of dodge/parry means that if you have more than 1.87% dodge for every 1% parry [the #s right from your character sheet, but will be different w/ raid (agi) buffs] then you should use a parry/stam instead of dodge(agi)/stam for maximum avoidance

yes, thanks. In my case it happens that, fully raid buffed, I have actually 2.1 as a ratio (once substracted the naked parry and dodge from the percentages), which make more efficient to gem for parry


As a note, what Venoseth said had one minor quirk: you need to subtract 10% from both your dodge and parry %'s in your character sheet before you apply the 1.875 ratio. The ratio is DR'd dodge to DR'd parry, and you have 10% non DR parry and *about* 10% non DR dodge.

For example, if you look at my horribly out of date signature, I have 25.55% dodge and 20.735 parry, I would do:

(25.55-10)/(20.73-10) = 1.449, so I need to keep going with dodge in my case. As Venoseth noted, that will be different with raid buffs, but since I know my parry won't increase with those, I can go ahead and see what my dodge will need to be before I start looking at parry:

(x-10)/(20.73-10) = 1.875 => x = 30.12

So I should continue to stack dodge until my raid buffed dodge is 30.12, and then I stack them at 1.875:1 to maintain the ratio.
Image
User avatar
jere
 
Posts: 2957
Joined: Fri May 11, 2007 5:12 pm

Re: Theck's MATLAB TPS analysis (A Jonesy derivative work)

Postby Gamingdevil » Wed Aug 19, 2009 1:23 pm

Culhag wrote:I've updated my RatingBuster modification using the latest 3.2 compatible version and Theck's latest threat stats value calculation for 3.2.

I've also added an EH comparison that really was missing from RB.

Check it out !


For the threat part, do you change the value of expertise after the soft-cap is reached?
If it is weak, kill it before it gets stronger. If it is strong, weaken it.
Image
User avatar
Gamingdevil
 
Posts: 1738
Joined: Mon Jan 14, 2008 8:16 am
Location: Belgium

Re: Theck's MATLAB TPS analysis (A Jonesy derivative work)

Postby Nadir » Wed Aug 19, 2009 1:27 pm

This is a fairly accurate approximation of the effect of raid buffs on your dodge %.

((Base_AGI+AGI_from_gear)*(0.10)+(Strength_of_Earth_Totem+Mark_of_the_Wild)*(1.10))*(1 dodge % / 59.8958333 AGI)

Let's assume you're using a non-agi weapon enchant and are a Draenei.

((76 AGI+22 AGI)*(0.10)+(152 AGI+52 AGI)*(1.10))*(1% dodge / 59.8958333 AGI) = 3.91% dodge (pre-DR)

This is roughly 2.61% dodge from buffs post DR assuming you have 25% dodge on the character sheet.
Image
Nadir
 
Posts: 266
Joined: Mon May 04, 2009 8:37 am

Re: Theck's MATLAB TPS analysis (A Jonesy derivative work)

Postby tlitp » Wed Aug 19, 2009 5:13 pm

Thalastor82 wrote:(...) parry could work as a nice TPS boost with the new SoV. Hard to calculate and to compare with the TPS coming from increase crit from Agi, though.

parry rating as a threat stat is really in a world of its own; without points in Reckoning, 1 parry rating is roughly equivalent to 1 haste rating (all raid (de)buffs assumed available, using SoV); having points in Reckoning is where the things get messy, as there are cases in which additional parry rating does actually lower the overall output

threat-wise, 1 parry rating should be considered strictly inferior to 1 agility
Nadir wrote:((Base_AGI+AGI_from_gear)*(0.10)+(Strength_of_Earth_Totem+Mark_of_the_Wild)*(1.10))*(1 dodge % / 59.8958333 AGI)

gear_AGI is fully subjected to DR, whereas base_AGI is not :
((base_agi*0.1+gear_agi*1.1)+ ...
User avatar
tlitp
 
Posts: 556
Joined: Mon Jul 27, 2009 3:25 pm

Re: Theck's MATLAB TPS analysis (A Jonesy derivative work)

Postby lythac » Wed Aug 19, 2009 5:45 pm

tlitp wrote:without points in Reckoning, 1 parry rating is roughly equivalent to 1 haste rating (all raid (de)buffs assumed available, using SoV)


Have you a source on this?

Combat Level ratings

Haste 32.79
Parry 45.25

As they both affect the same attacks and parrying only makes your next attack come around 24% faster (although is an old post) I would think that 5.75 parry is roughly equivalent to 1 haste unless I have missed something.
Ryshad / Lythac of <Heretic> Nagrand-EU
User avatar
lythac
Moderator
 
Posts: 2673
Joined: Wed Sep 24, 2008 8:10 am

Re: Theck's MATLAB TPS analysis (A Jonesy derivative work)

Postby Nadir » Thu Aug 20, 2009 12:21 am

tlitp wrote:
Nadir wrote:((Base_AGI+AGI_from_gear)*(0.10)+(Strength_of_Earth_Totem+Mark_of_the_Wild)*(1.10))*(1 dodge % / 59.8958333 AGI)

gear_AGI is fully subjected to DR, whereas base_AGI is not :
((base_agi*0.1+gear_agi*1.1)+ ...

Yes, this is true. Base_AGI is not subjected to DR. That's not quite what I'm after.

I'm merely trying to approximate the effect of raid buffs post DR relative to your unbuffed dodge that appears on your character sheet. The DR on your gear_agi has already been taken into account in the character sheet number. However, the DR on the 10% AGI gain from Kings on your base_AGI and gear_AGI haven't been taken into account. Therefore base and gear AGI are subjected to 0.10 rather than 1.10.
Image
Nadir
 
Posts: 266
Joined: Mon May 04, 2009 8:37 am

Re: Theck's MATLAB TPS analysis (A Jonesy derivative work)

Postby Arees » Thu Aug 20, 2009 11:18 am

Hey Theck (or anyone else that can answer this). In this post, you examined librams. That was in 3.1 though when LoO did not have 100% uptime. I'm probably going to be getting the new dodge libram soon and I'm just curious what my dps/threat loss will be.
Image
User avatar
Arees
 
Posts: 546
Joined: Tue Dec 25, 2007 9:50 pm

Re: Theck's MATLAB TPS analysis (A Jonesy derivative work)

Postby Jasari » Thu Aug 20, 2009 11:36 am

WATERBOYsh wrote:Hey Theck (or anyone else that can answer this). In this post, you examined librams. That was in 3.1 though when LoO did not have 100% uptime. I'm probably going to be getting the new dodge libram soon and I'm just curious what my dps/threat loss will be.

It's pretty easy to estimate by hand:

Assuming you're not BV capped, LoO adds a flat (352 * 1.3) = 457.6 BV.

This adds 457.6 damage to your ShoR, which you cast once every 6 seconds.

Then factor in hit/miss/crit rates as well as +% damage raid buffs:

457.6 * (hit chance) * ((100 + crit chance) / 100) * 1.1(OHWS) * 1.03(Crusade) * 1.13(CoE) * 1.03(SancRet) * (94.8 / 100)(DR from boss partial resists) / 6

So if we assume you're hit capped and have 20% crit raid buffed:

457.6 * (1.00) * (1.2) * 1.1 * 1.03 * 1.13 * 1.03 * (94.8 / 100) / 6 = 114.41 DPS or (114.41 * 1.43 * 1.9) = 310.85 TPS (pre 3.2.2)

This will change slightly depending on your actual hit and crit % and whether or not you spec'd for Crusade. Also, against Undead that'd go up since you'd get another 3% from crusade and 1% from Glyph of Sense Undead
User avatar
Jasari
 
Posts: 2195
Joined: Sun Apr 27, 2008 12:32 am
Location: Boston, MA

Re: Theck's MATLAB TPS analysis (A Jonesy derivative work)

Postby Arees » Thu Aug 20, 2009 11:40 am

Jasari wrote:457.6 * (1.00) * (1.2) * 1.1 * 1.03 * 1.13 * 1.03 * (94.8 / 100) / 6 = 114.41 DPS or (114.41 * 1.43 * 1.9) = 310.85 TPS (pre 3.2.2)


Why is that number lower than the 333 TPS theck said that LoSS adds in the thread I linked? I thought since 3.2 LoO was superior.
Image
User avatar
Arees
 
Posts: 546
Joined: Tue Dec 25, 2007 9:50 pm

Re: Theck's MATLAB TPS analysis (A Jonesy derivative work)

Postby Jasari » Thu Aug 20, 2009 11:42 am

WATERBOYsh wrote:
Jasari wrote:457.6 * (1.00) * (1.2) * 1.1 * 1.03 * 1.13 * 1.03 * (94.8 / 100) / 6 = 114.41 DPS or (114.41 * 1.43 * 1.9) = 310.85 TPS (pre 3.2.2)


Why is that number lower than the 333 TPS theck said that LoSS adds in the thread I linked? I thought since 3.2 LoO was superior.

Was that post before the nerf from 130% BV to 100% BV for ShoR?
User avatar
Jasari
 
Posts: 2195
Joined: Sun Apr 27, 2008 12:32 am
Location: Boston, MA

Re: Theck's MATLAB TPS analysis (A Jonesy derivative work)

Postby Arees » Thu Aug 20, 2009 11:48 am

Jasari wrote:Was that post before the nerf from 130% BV to 100% BV for ShoR?


I totally forgot about that... it was before. So your numbers are probably correct.

Thats not to bad of a tps loss... But my next question is about the EH loss. I know 450BV is not as good as 40ish stam because BV doesn't work on magic damage, but its still 450 less damage you take on every physical swing. I'm thinking though that I'll still want to get the libram, because thats a lot of dodge from just one item. Ratingbuster tells me its almost 3% for me after DR.
Last edited by Arees on Thu Aug 20, 2009 11:54 am, edited 2 times in total.
Image
User avatar
Arees
 
Posts: 546
Joined: Tue Dec 25, 2007 9:50 pm

Re: Theck's MATLAB TPS analysis (A Jonesy derivative work)

Postby Jasari » Thu Aug 20, 2009 11:53 am

For general survivability purposes, I think the dodge libram is superior. I also think I remember seeing a blue post recently that said they were planning to fix the fact that LoSS is worse than LoO.
User avatar
Jasari
 
Posts: 2195
Joined: Sun Apr 27, 2008 12:32 am
Location: Boston, MA

PreviousNext

Return to Advanced Theorycraft and Calculations

Who is online

Users browsing this forum: twilytgardnfaery and 1 guest


Remove Advertisements

Who is online

In total there are 2 users online :: 1 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: twilytgardnfaery and 1 guest
?php } else { ?