Remove Advertisements

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

Warning: Theorycraft inside.

Moderators: Fridmarr, Worldie, Aergis, theckhd

Re: Theck's MATLAB TPS analysis

Postby theckhd » Thu May 20, 2010 9:09 am

tlitp wrote:Can't a man edit a post without Theck interfering ? :twisted:

fixT (as in, re-quoted)
"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: 7710
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB TPS analysis

Postby tlitp » Thu May 20, 2010 9:13 am

theckhd wrote:
tlitp wrote:Can't a man edit a post without Theck interfering ? :twisted:

fixT (as in, re-quoted)

Also re-quote the edited second post. :lol:
theckhd wrote:field-like structure

For the greater win. I had intended to bug you about a SVN-like system, but lotsofwork+noavailabletime=no-no.
User avatar
tlitp
 
Posts: 556
Joined: Mon Jul 27, 2009 3:25 pm

Re: Theck's MATLAB TPS analysis

Postby badgermonkey » Wed May 26, 2010 4:39 am

When you did your big exhaustive weapon analysis, you said you did it with the assumption Last Word had an 80% uptime. As it's now far closer to 100% (98% or so maybe), is it possible to run that through again?

Most specifically I'm trying to compare it's threat comparative to Bloodvenom Blade, but having it up to date and accurate would be nice :)

(Also, sorry if this WAS redone and I missed it)
Image
badgermonkey
 
Posts: 329
Joined: Tue Apr 28, 2009 6:36 pm

Re: Theck's MATLAB TPS analysis

Postby theckhd » Wed May 26, 2010 5:00 am

It was.

Theck wrote:# Last Word now has a very high uptime. I've modeled it at 100% for these calculations.
"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: 7710
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB TPS analysis

Postby badgermonkey » Thu May 27, 2010 4:38 am

Yay Theck then! :P

Cheers :)
Image
badgermonkey
 
Posts: 329
Joined: Tue Apr 28, 2009 6:36 pm

Re: Theck's MATLAB TPS analysis

Postby theckhd » Fri May 28, 2010 7:01 am

Announcement:
I'm not going to be doing much more in the way of TPS calculations for 3.x. I assume we won't see any major overhauls to our class with Ruby Sanctum, and the gear is only a slight upgrade or sidegrade for many of us. I might run a quick weapon DPS/TPS analysis that includes the new weapons, but I may not. It really depends on what the loot tables look like. If I do, it will be an abbreviated analysis that only considers ICC and Ruby Sanctum gear, in part because the code has gotten so unwieldly that I don't want to mess with it too much.

So that's the "bad news," as such. The good news is the reasoning behind it - I'm starting to work on the 4.0 code. This is more of a heads-up post than anything, because I haven't gotten very far yet. However, there are a few major changes in store, and a few things I'm going to ask for assistance with.

Changes:
  1. Major code revamp.

    1. Clean-up: I'm going to be doing a lot of re-writing to clean up the code and make it more usable. This includes basic things (like making sure variable names are consistent) and structural things (trying to standardize some of the array representations I use). And of course, it also includes re-writing the code to make it consistent with Cataclysm mechanics.
    2. New gear system: Complete re-write of the way the code handles gear, including a "Gear Database" that contains all of the stats of any item we want to include. Loading or invoking a gear set will be much easier with this method, as it will just require a function call, like "equip(cataclysmicchestguard)" or "equipset(theck_t11)"
    3. Better "out-of-the-box" experience: Hopefully, an easier experience for beginners who want to mess with the code. My goal is that someone with access to MATLAB, Octave, etc. can download the code, make very simple changes to a "gear" file (i.e. change "equip(cataclysmicchestguard)" to "equip(sanctifiedlightswornchestguard)", and then run a basic file that spits out a bunch of useful information about your character.
  2. Collaboration improvements

    1. New repository: To make it easier for other people to contribute to the code, I'm moving to an SVN system on Google Code. So far, the code has been handled locally; I keep my "master" copy here, and upload it to the google sites repository every so often for others to use. This has a number of downsides, especially when other people want to suggest or submit changes or improvements. With SVN, the master copy will reside on the server, making it easier for me to work on it in multiple locations and easier for others to contribute.
    2. Given that, I'm hoping to encourage more people to help out. If I can make the code more readable, and make it easy for people to submit changes or improvements, then I can hopefully spread the workload around and get things done faster. Even if all someone does is help maintain a single module, it gives me more time to work on the rest of the code.
  3. Call for Assistance

    There's a lot of work to be done in the next few months to get everything ready for Cataclysm. Here's the two major jobs I'll want help with:
    1. People willing to work on and/or maintain code modules or calculation files. This would require some basic knowledge of how MATLAB works, as well as an understanding of the mechanics you're trying to model. I'm hoping to keep everything well-commented so that it's easy to understand even for a novice coder. This job would involve downloading the code from the SVN, making and testing changes, and uploading new builds that work.
    2. Testing the Cataclysm mechanics on the beta servers, both to confirm that old mechanics haven't changed and to determine how to properly model new mechanics. This doesn't require any coding knowledge, but will be fairly time intensive.
If you're interested in helping with the code, get in touch with me and I'll set you up with SVN access once we get a few more of the foundation built. Until we find out who has beta access, it will be hard to volunteer for testing, but if you're willing feel free to speak up.

I plan on starting a new thread for discussion about the new code once work has started in earnest. I will probably leave this one open until Cataclysm's release in case there are questions about the old results and simulations.
"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: 7710
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB TPS analysis

Postby Awyndel » Fri May 28, 2010 10:58 am

You know this code is probably one of the reasons maintankadin has some of the best tank theorycrafting out there.

We all appreciate the work a lot. Can't wait for 4.0 stuff :) .

Unfortunately I can't code or get a beta key, but I can keep assisting with silly questions, as usual :D .
User avatar
Awyndel
 
Posts: 672
Joined: Sat Feb 14, 2009 8:49 am
Location: The Netherlands

Re: Theck's MATLAB TPS analysis

Postby Skideflink » Mon May 31, 2010 12:38 am

Hello

I'm new here on maintankadin, I was looking around in this threat after something special, which is how much tps a paladin tank in almost endgame gear (277 mostly, and 271 onehand) should make, without any assistance from rogues/hunters. Unfortunately I couldn't find the one number Í'm looking for.

I'm a fustrated feral druid which happens to be overaggroing alot, and I made this fine little post to see if there's something wrong with me, or if our maintank is slacking.

as a feral druid on sindragosa hardmode I need to get 2 salvations, intervene spam, and vigilance from a war, still holding abit back on dps, for not overaggroing in the first phase 1. Our MT even got a MD from the begging.

I'm at almost same gearlvl as out paladin tank + this is with the 20% buff.

But basicly: How much tps should a paladin do on an encounter as Sindragosa (or Deathbringer even, just a fight were u stand still in the begging) the first minute, with a 271wep and mostly 277 gear?

Sorry for the worst grammar EU.
and once again sorry, if I'm changing the subject from something important, though it seems there haven't been so many new posts the last week.
Skideflink
 
Posts: 3
Joined: Mon May 31, 2010 12:25 am

Re: Theck's MATLAB TPS analysis

Postby Meloree » Mon May 31, 2010 1:37 am

Sindragosa is one of those fights where you actually want your rogues and hunters to tricks/MD the MT on cooldown, as opposed to using it to boost DPS. I personally consider myself to be pretty decent at threat output, but Sindragosa can still be a challenge. Most people wear some Frost Resist gear, and flush out the gearset with max survival, leading to abnormally low TPS, especially when combined with the attack-speed debuff, which reduces threat by some margin. I don't have a handy number for Sindragosa "time on target" threat output in the gearset I use, but it's pretty anemic, in the ballpark of 10-12k tps. Two salvations and chain intervenes plus vigilance sounds like there's probably a more severe threat problem, though. Are you getting Tricks with your berserk, or some other threat-boosting gimmick that's fairly unwise in the first P1 at Sindragosa? If not, your tank may need to work on his or her threat output. If you want to link us a WOL/WMO parse, we could look into it further.

For Saurfang, I actually have useful numbers handy: At 0%, I did roughly 12k TPS in full progression gear at 0% (roughly 14.5k TPS equivalent now, at 20%). In the threat-heavy gearset I use now, roughly 17k sustained TPS, and in the gimmick "lol tps parses" set I used this week at 20%, around 22k tps.

So, TLDR: Sindragosa is a bit of a special case, to which your raid might need to adapt. TPS can be an issue there. If you're having threat problems at Saurfang, though, something is wrong.
Meloree
Maintankadonor
 
Posts: 1420
Joined: Wed Mar 12, 2008 10:15 am

Re: Theck's MATLAB TPS analysis

Postby Boyfriend » Mon May 31, 2010 3:04 am

Is your tank stopping his attacks on Sindragosa to stop the Chilled to the Bone debuff from stacking?
He should not, especially if he's wearing frost resist gear the damage it does even with 7 stacks is fairly inconsequential on a tank; especially since the attack speed slow usually causes it to drop anyway.

I have never had issues with threat on sindragosa, even back when I used Frost Res/Max Survival gear, the fact that I was attacking through Blistering Colds while everyone else ran out; and that I was generating threat with HoR/AS/Judge before Sindragosa landed while dps were still busy with Ice Blocks allowed me to compensate for low TPS gear.

Only Boss I am having Threat issues on is LK, if a Shambler spawns in swipe/maul range of our druid tank I am usually incapable of getting aggro on it. LK-spec + full EH gear + using GCD's for non threat things just results in an inability to beat Druid AE threat with my Single target threat :<

Though 'issue' is an exaggeration since I just taunt them before they go active anyway, but sometimes I forget ...
Image
User avatar
Boyfriend
Maintankadonor
 
Posts: 588
Joined: Wed Feb 24, 2010 8:16 am
Location: Sweden

Re: Theck's MATLAB TPS analysis

Postby Anorian » Mon May 31, 2010 3:21 am

Reading the last few responses I'm getting the feeling I'm doing something wrong. Not that I'm losing aggro but the tps numbers you are producing are far higher than my omen says.

I scored several top 10 parses so I'm fairly sure it's not my ability to do dmg or 969 for that matter. But in my survival set I am mostly seeing around 10-11k tps, maybe I'm not watching omen at the right times but I'm definitely not seeing 17k ish sustained in a threat set even.
Image
User avatar
Anorian
 
Posts: 361
Joined: Tue Jun 03, 2008 2:38 am

Re: Theck's MATLAB TPS analysis

Postby Boyfriend » Mon May 31, 2010 3:32 am

Anorian wrote:Reading the last few responses I'm getting the feeling I'm doing something wrong. Not that I'm losing aggro but the tps numbers you are producing are far higher than my omen says.

I scored several top 10 parses so I'm fairly sure it's not my ability to do dmg or 969 for that matter. But in my survival set I am mostly seeing around 10-11k tps, maybe I'm not watching omen at the right times but I'm definitely not seeing 17k ish sustained in a threat set even.


I would not trust omen; open your parse and calculate:

((Total Damage - Melee Damage) * 1.8 + Melee Damage) * 1.43 / Fight Length

You can check other Paladins logs in the same way and compare. Always compare Apples with Apples. (so even if the calculation is not 100% correct, you minimize errors by doing the same calculation on all data)

Checking what number you see in Omen is not at all accurate.
Image
User avatar
Boyfriend
Maintankadonor
 
Posts: 588
Joined: Wed Feb 24, 2010 8:16 am
Location: Sweden

Re: Theck's MATLAB TPS analysis

Postby Skideflink » Tue Jun 01, 2010 2:07 am

Our MT only have 1 set of gear, which is based on highest item-lvl (atleast that's my guess).

On sindragosa he didn't use any frost resistance gear, and we only had 1 hunter and 0 rogues, so I did not get any Tricks, and our MT gets the MD on CD.

From what I can read, it appears our Maintank is abit slacky when it comes to tps.

Here's the log for that day we took him. I do not req. any of you reading it or posting back, but some1 mentioned he would like to look into it, though don't do it just for me, I'm perfectly fine with what I got so far.

http://www.worldoflogs.com/reports/rt-d8wmbs8psue0uc2r/

and thanks for all the replies <3
Skideflink
 
Posts: 3
Joined: Mon May 31, 2010 12:25 am

Re: Theck's MATLAB TPS analysis

Postby lythac » Tue Jun 01, 2010 4:13 am

Skideflink wrote:and our MT gets the MD on CD.


No he doesn't. For your kill only 1 MD.

http://www.worldoflogs.com/reports/rt-d ... 26&e=12894 [{"spellNames": ["Misdirection"]}]
(Show events where spell is Misdirection)



Also you only had Vigilance for 5 of 20 attempts.

http://www.worldoflogs.com/reports/rt-d ... 0uc2r/log/ [{"spellNames": ["Vigilance"]}]
(Show events where spell is Vigilance)



The paladin's rotation looks to be decent enough. And for most of the attempts AW is used before the pull. It does not look like the paladin is playing badly, looks pretty decent.

http://www.worldoflogs.com/reports/rt-d ... 0uc2r/log/ [{"spellNames": ["Avenging Wrath"], "sourceNames": ["Adea"]}, {"spellNames": ["Hand of Reckoning"], "targetNames": ["Sindragosa"], "sourceNames": ["Adea"]}]
(Show events where source is Adea and spell is Avenging Wrath
Show events where source is Adea and target is Sindragosa and spell is Hand of Reckoning)
Ryshad / Lythac of <Heretic> Nagrand-EU
User avatar
lythac
Moderator
 
Posts: 2670
Joined: Wed Sep 24, 2008 8:10 am

Re: Theck's MATLAB TPS analysis

Postby Skideflink » Tue Jun 01, 2010 6:23 am

Thanks once again.

About vigilance, we switched tanks now and then, so I only had it when he was tank, when he wasn't, I only had the intervene's left.

About the MD, that sounds sad :p gotta ask our hunter to assist on that more.

I'm glad, that it looks our tank is doing decent, since I doubt he even did.
Skideflink
 
Posts: 3
Joined: Mon May 31, 2010 12:25 am

PreviousNext

Return to Advanced Theorycraft and Calculations

Who is online

Users browsing this forum: No registered users and 1 guest


Remove Advertisements

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