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

Warning: Theorycraft inside.

Moderators: Fridmarr, Worldie, Aergis, theckhd

Re: Theck's MATLAB TPS analysis

Postby chinoquezada » Sun Mar 21, 2010 12:09 am

The Ability TPS/DPS Breakdown is based on a gear set with 2pc t10, any chance you can quickly model this without 2pc t10?
It would basically be removing 20% of the damage from HotR. My guess is that it would move ShoR to top DPS/TPS ability.
User avatar
chinoquezada
 
Posts: 102
Joined: Thu Jun 19, 2008 2:47 am

Re: Theck's MATLAB TPS analysis

Postby theckhd » Mon Mar 22, 2010 6:39 am

chinoquezada wrote:The Ability TPS/DPS Breakdown is based on a gear set with 2pc t10, any chance you can quickly model this without 2pc t10?
It would basically be removing 20% of the damage from HotR. My guess is that it would move ShoR to top DPS/TPS ability.

It should be a simple multiplicative factor. So it would drop HotR's net damage from 4561 to 3649, slightly below ShoR.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
MATLAB 5.x, Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 5.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 7956
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB TPS analysis

Postby Awyndel » Thu Apr 08, 2010 5:33 am

I have a tps ( well actually dps ) question about trinkets.

Yesterday I was lucky enough to pick up Tiny abomination in a jar for my ret offspec. I also have the str greatness card, and the mark of supremacy for hit.

Now I was wondering for the prot spec, asuming the hit trinket is fully usefull, wich 2 trinkets would do the most tps/dps. So basicly I'm asking where tiny abomination stands, don't know if there is prot math on it yet.

And of course it gets a lot more complicated. Coz the numbers will ofc be different for both our seals.

Did I miss any other good threat trinket options?
User avatar
Awyndel
 
Posts: 672
Joined: Sat Feb 14, 2009 8:49 am
Location: The Netherlands

Re: Theck's MATLAB TPS analysis

Postby theckhd » Thu Apr 08, 2010 6:48 am

Awyndel wrote:I have a tps ( well actually dps ) question about trinkets.

Yesterday I was lucky enough to pick up Tiny abomination in a jar for my ret offspec. I also have the str greatness card, and the mark of supremacy for hit.

Now I was wondering for the prot spec, asuming the hit trinket is fully usefull, wich 2 trinkets would do the most tps/dps. So basicly I'm asking where tiny abomination stands, don't know if there is prot math on it yet.

And of course it gets a lot more complicated. Coz the numbers will ofc be different for both our seals.

Did I miss any other good threat trinket options?

Quick napkin math, feel free to double-check this. Using the stat weightings in the Stat Threat Analysis

Code: Select all
        TPS per 10
Stat    stat
Hit     31.80
STR     30.80
AP      9.10

Mark of Supremacy = 128 Hit rating, 1024*20/120=171 avg AP.
128*3.18=407 TPS
171*0.91=156 TPS
Total of 563 TPS

DMC:G = 190 avg STR
190*3.08=585 TPS

So DMC:G is still a little ahead of Mark, though not by a lot. Mark also has the advantage of giving you more consistent threat (since you'll get fewer misses). So it's a toss-up between those two.

I've seen various reports on TAiaJ, so I'm not 100% sure on what exactly triggers it, especially for prot. It seems like DS, CS, Seal procs, and Judgement all proc motes for Ret; presumably Judge/HotR/ShoR and Seal procs all trigger it for prot as well. However, we don't know whether HV/JotJ/Vindication applications proc it or not.

We could probably do extensive testing to figure out every potential proc trigger like we did with Last Word/Vindication/Mongoose/BW, but in this case we don't care about uptime, just results. So it's probably easiest to just wail on a Dummy for a while and upload the log to WoL. Since it should be relatively frequent, a 10 minute session should give you decent enough statistics to have a good average proc rate for a 969 rotation.

It would probably be a good idea to check that Consecration ticks can't proc it, so that you can drop Cons from the rotation to conserve mana. That will probably be enough to let you keep going for 10 mins straight. May want to repeat the test twice to test the dependence on hit rating as well - once with nearly 0% hit rating and once with 8%.

You also may want to test it with SoCom, since I've read that each SoCom proc can trigger the buff. My guess is that you use SoCom and attack 2-3 dummies you'll get a noticeably higher proc rate.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
MATLAB 5.x, Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 5.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 7956
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB TPS analysis

Postby Boyfriend » Thu Apr 08, 2010 7:05 am

IIRC Apart from what you listed (CS/DS/Judgement/Seal/Melee) only HV proccs it for ret, which accounts for a large part of it's dps for ret. (Up to 4 motes per attack)

I assume Vindication cannot but JotJ might proc it for prot, hope I can get a TAJ myself someday to try it out.
Image
User avatar
Boyfriend
Maintankadonor
 
Posts: 588
Joined: Wed Feb 24, 2010 8:16 am
Location: Sweden

Re: Theck's MATLAB TPS analysis

Postby Tintalle » Thu Apr 08, 2010 6:30 pm

From what I've seen from about 10 minutes at a target dummy everything listed does proc TaJ including JotJ, and 2 chances from just judgement alone without JotJ or SoV. This means white swings and HotR are up to 3 motes each, ShoR is up to 2 each and judgement is up to 4 each assuming SoV is up and seal hits go along with each one. Consecrate does not proc it. Also went afk for 306 swings and had 152 motes proc, so I'd say close to 50% proc rate as well. Edit: it seems SoC isn't proccing TaJ, the hit on your target or the cleaves. Bug or not, that's what I'm seeing. Angling to not melee but use specials, using ShoR and HotR only got 1 mote per use over 50 uses of each and ShoR cleaving the dummy behind me for 150 seal procs with no additional motes.
Overall very impressed so far from this proc rate, with a slow weapon could be a very solid tps increase.
Last edited by Tintalle on Thu Apr 08, 2010 11:26 pm, edited 2 times in total.
Tintalle
 
Posts: 2
Joined: Thu Apr 08, 2010 5:40 pm

Re: Theck's MATLAB TPS analysis

Postby Rasmfrackn » Thu Apr 08, 2010 10:07 pm

Wow, lots to catch up on.

I'm expecting to use my 4pt10 bonus, so I'm trying to choose between Kraken gloves and Cataclysmic chest to sub in. They're nearly identical on survival stats, so I thought I'd look at threat to make a decision. It ends up being 60 str vs. 82 expertise and 54 hit. I was assuming the expertise and hit would win, but I'm right at 26 expertise with the SoV glyph, and I'm at 262 hit with lingering vortex boots. With those breakpoints, it looks like the 60 str is stronger.

Here's my question, though... why is the Str window on the stat values so high? I'm pretty light on Str because I'm still using some lower ilvl stuff, but even the T10 progression set on page 80 something, and Theck's own gearset in the armory, are only around 2000 str unbuffed *with* the SoV libram adding 200. It seems like using 1900 as the low end is pretty optimistic for unbuffed str. I'm only at about 1600, or 1800 with the libram procced.

BTW, the new icy touch makes me feel like less of a tank. :(
Rasmfrackn
Dwarf Paladin
Icecrown Server
Eng/Scribe/Masochist
User avatar
Rasmfrackn
 
Posts: 1014
Joined: Fri Aug 17, 2007 12:11 pm

Re: Theck's MATLAB TPS analysis

Postby Meloree » Thu Apr 08, 2010 11:32 pm

Rasmfrackn wrote:Here's my question, though... why is the Str window on the stat values so high? I'm pretty light on Str because I'm still using some lower ilvl stuff, but even the T10 progression set on page 80 something, and Theck's own gearset in the armory, are only around 2000 str unbuffed *with* the SoV libram adding 200. It seems like using 1900 as the low end is pretty optimistic for unbuffed str. I'm only at about 1600, or 1800 with the libram procced.


Raid buffs.

EDIT: I'm dumb, Theck apparently did everything unbuffed. Personally, I think that's a mistake. Unless the calculation adds in raid-buffs. But something a bit funky is going on, because the T10 progression set only has 1650 Str unbuffed, but the plot starts at 1900ish. So I'm thinking the calculation adds raidbuffs in, but doesn't subtract them out, and Theck isn't really referring to "unbuffed" anything.

EDIT2: Theck references the unbuffed character sheet str of the default gearset being 2070, but in the gearset breakdown it says 1649. (1649+(155+51)*1.15)*1.10=2074. That's Horn of Winter + Mark + Kings/Sanc. Also, said default gearset includes Libram of Valiance, for another 200*1.15*1.10=253 str, which should mean that the total buffed str is closer to 2327. Theck, something is inconsistent somewhere.
Meloree
Maintankadonor
 
Posts: 1420
Joined: Wed Mar 12, 2008 10:15 am

Re: Theck's MATLAB TPS analysis

Postby xstratax » Fri Apr 09, 2010 12:30 am

Rasmfrackn wrote:BTW, the new icy touch makes me feel like less of a tank. :(


That. My DK OT can hit like 20k threat in one GCD, I pop wings, Taunt+AS macro, and HotR (with 2pT10). Icy Touch still pulls threat. Im not the best geared, far from it, but equally geared its ridiculous.
xstratax
Maintankadonor
 
Posts: 118
Joined: Fri Aug 14, 2009 1:18 pm

Re: Theck's MATLAB TPS analysis

Postby theckhd » Fri Apr 09, 2010 7:49 am

Meloree wrote:Theck, something is inconsistent somewhere.


The plot axis is generated by this line:
Code: Select all
x(j,:)=floor((Base_STR+Gear_STR+extra_STR).*DivStr);

Base_STR is 151
Gear_STR is 1649
extra_STR is an array that varies from -200 to +200 to give you a range.

The unbuffed character sheet STR of the base gear set should then be:
(151+1649)*1.15 = 1800*1.15 = 2070

Your calculation came out to 2074 by coincidence it seems, since you made a few errors. The 1649 STR from gear didn't get the Divine Strength modifier and base health was omitted. Also note that the 200 STR from LoValiance is already included in the 1649 attributed to gear.

Note that the gear set was chosen as a "25-man hard-mode progression" gear set, which is basically the gear level I'm at now (roughly full 264 gear). My character sheet is using a few different pieces than the progression set, but I'm at 1793 before the 230 from the Libram and 115 from the Last Word proc (or 94 from the Bonebreaker Scepter). That would put me at 2117 before raid buffs or kings.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
MATLAB 5.x, Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 5.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 7956
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB TPS analysis

Postby Meloree » Fri Apr 09, 2010 8:08 am

theckhd wrote:
Meloree wrote:Theck, something is inconsistent somewhere.


Your calculation came out to 2074 by coincidence it seems, since you made a few errors. The 1649 STR from gear didn't get the Divine Strength modifier and base health was omitted. Also note that the 200 STR from LoValiance is already included in the 1649 attributed to gear.


Sounds about right that it's me being inconsistent. I assumed the 1649 included Divine Strength, then it seemed low, so I figured the proc wasn't up. Serves me right for posting at 2am.

That said, why don't the calculations include raidbuffs? Everyone is going to be ~400str further along the line, with ~11% more crit, sunders/WF/CoE, etc. Seems like it would make a pretty significant difference in the stat valuations?
Meloree
Maintankadonor
 
Posts: 1420
Joined: Wed Mar 12, 2008 10:15 am

Re: Theck's MATLAB TPS analysis

Postby theckhd » Fri Apr 09, 2010 8:14 am

Meloree wrote:That said, why don't the calculations include raidbuffs? Everyone is going to be ~400str further along the line, with ~11% more crit, sunders/WF/CoE, etc. Seems like it would make a pretty significant difference in the stat valuations?

The calculations do include raid buffs. The axis doesn't. I've plotted it against your unbuffed character sheet STR, because that's something you can easily read off and compare to. Not everyone remembers their full raid-buffed STR, but you can easily check your unbuffed char sheet STR in-game or via the armory.

In other words, you look at the armory and see that your unbuffed character sheet STR is 2000. The stat values the plot gives you are what you'd expect if you took that armory profile and fully raid buffed it (and thus would have well over 2000 STR).
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
MATLAB 5.x, Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 5.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 7956
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB TPS analysis

Postby Meloree » Fri Apr 09, 2010 8:22 am

theckhd wrote:
Meloree wrote:That said, why don't the calculations include raidbuffs? Everyone is going to be ~400str further along the line, with ~11% more crit, sunders/WF/CoE, etc. Seems like it would make a pretty significant difference in the stat valuations?

The calculations do include raid buffs. The axis doesn't. I've plotted it against your unbuffed character sheet STR, because that's something you can easily read off and compare to.


Ahh. Serves me right. Sorry.
Meloree
Maintankadonor
 
Posts: 1420
Joined: Wed Mar 12, 2008 10:15 am

Re: Theck's MATLAB TPS analysis

Postby theckhd » Fri Apr 09, 2010 8:29 am

Meloree wrote:Ahh. Serves me right. Sorry.

No problem. I had to go back to the code to make sure I hadn't made a mistake. And I'm sure that others have been confused by that subtlety in the past.

Also, I probably explained it reasonably well the first time I used that axis system and then got lazy about describing it in later posts. So I can make it slightly clearer in the future.

Though I don't expect to do much serious number crunching in this thread until either Ruby Sanctum (if we see any major mechanics changes) or Cataclysm beta.


Somewhat Unrelated Thought:
On the bright side (for you guys, anyhow), I may have to cut back in my raiding schedule in the next few months. What that most likely would mean is less raiding 25's and/or raiding with a different guild at more convenient or abbreviated times. That might leave me with a lot more in-game free time, which could be spent doing paladin testing in the beta (provided I can beg/borrow/steal a beta invite).

Either way I'll be keeping abreast of the beta changes and updating the models for 4.0. This TPS thread is still and will always be my primary project, so all of the survivability/defensive stuff will be set aside until everything here is updated.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
MATLAB 5.x, Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 5.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 7956
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB TPS analysis

Postby Awyndel » Fri Apr 09, 2010 11:57 am

Well don't really have the time to make a log right now. But I did some playing around with taiaj. I can pretty much confirm what the previous poster said.

it procs from white swings, shor and hotr. I don't think it procs from additional hotr targets.

It does not proc from socom at all. It procs from the sov applications and from the sov proc. I don't think it procs from sov applications on secondary targets.

As for judgement, I have been able to get 2 motes from it with seal of wisdom. So I don't know where that puts judgement, jotj and hotc. I don't know about vindication either.


I'm sure you could get a beta key if blizz knew what you where doing here Theck :) . Just tell them you have a blog or w/e. Would be really great to get a beta report with a math point of view.
User avatar
Awyndel
 
Posts: 672
Joined: Sat Feb 14, 2009 8:49 am
Location: The Netherlands

PreviousNext

Return to Advanced Theorycraft and Calculations

Who is online

Users browsing this forum: Exabot [Bot], Yahoo [Bot] and 1 guest

cron

Who is online

In total there are 3 users online :: 2 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: Exabot [Bot], Yahoo [Bot] and 1 guest