Theck's MATLAB thread - Cataclysm/4.x

Warning: Theorycraft inside.

Moderators: Fridmarr, Worldie, Aergis, theckhd

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby Zibey » Tue Jan 11, 2011 2:27 am

I am not sure it's right topic to post in, but I can't seem to find any comparison between TPS from reckoning, and grand crusader with glyph of focused shield (or how is it called). I should also probably mention, that I also won't have 10 expertise from SoV glyph (or how is it called now - I dinged like 2 days ago after a year break, so forgive me if I don't know new spell names), which probably will affect reckoning a bit (for the most of time at least - its supposed to be self healing heavy build).
Gniev - The Maelstrom EU
User avatar
Zibey
 
Posts: 129
Joined: Sun May 25, 2008 6:04 am

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby Durability » Tue Jan 11, 2011 6:29 am

Focused Shield would increase GC's effect by 30% on a single target (and I assume that's the only place you'd consider using it). Reckoning would lose approximately 5% value from enemy dodges and parries. But since the gap in the first place is well over 2:1, Reckoning still wins out by a substantial margin - it's still 80-90% better, even against a single target.
Durability
 
Posts: 43
Joined: Wed Jun 16, 2010 11:15 am

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby theckhd » Tue Jan 11, 2011 6:52 am

The Talent Comparison on the first page has exactly the information you're looking for.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 6.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 6211
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby Sarkan-ZdC » Tue Jan 11, 2011 8:44 am

Vengeance Calculation
The Vengeance calculation is working correctly as far as we can ascertain. Vengeance caps at a number equal to Stamina plus 10% of base health, which is lower than 10% of health. When we buffed Stamina late in Cataclysm development, we specifically did not buff Vengeance because we didn’t think tanks needed 40% additional attack power in order to maintain threat or contribute to damage done by the group. The Vengeance tooltip is a little misleading, but in this case we think that’s fine because a more technically accurate tooltip would be quite long. Astute theorycrafters will understand the difference, while most tanks won’t actually be affected by the delta. (Source)


Was that information known before? Does it change the threat stats (specifically how good stamina is as a tps stat?)
Sarkan-ZdC
 
Posts: 297
Joined: Tue May 13, 2008 4:00 am

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby theckhd » Tue Jan 11, 2011 9:27 am

All of the calculations currently up are based on pure hit-point scaling. We only became aware of the error last week. It's effectively like running at ~80% vengeance instead of 100%. So the value of stamina as a threat stat is about 20% lower than what's in the plots. The corrected Vengeance formula is already updated in the code base, but isn't going to be reflected in the posts until I update everything for 4.0.6.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 6.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 6211
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby kristoferpally » Wed Jan 12, 2011 10:30 am

Has anyone any new information on the bugged block meta gem? No mention of it was made in the 4.06 notes I've seen thus far :(
kristoferpally
 
Posts: 13
Joined: Thu Oct 14, 2010 3:34 pm

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby Arees » Wed Jan 12, 2011 7:06 pm

kristoferpally wrote:Has anyone any new information on the bugged block meta gem? No mention of it was made in the 4.06 notes I've seen thus far :(


Thats because it might not actually be bugged, but suffers from an unclear tooltip.

The "bug" is that the tooltip says it increases the amount blocked by 5%, but when you equip it, your character screen block value does not go to 35%. This is because 30 * 1.05 = 31.5. I can't remember if the character screen truncates this to 31 or rounds up to 32. My memory says it truncates it, so only shows 31%, though my memory could be faulty.
Image
User avatar
Arees
 
Posts: 301
Joined: Tue Dec 25, 2007 9:50 pm

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby Durability » Wed Jan 12, 2011 9:28 pm

It truncates, both on the tooltip and some testing I did on beta.

If the tooltip is accurate, it still does it even with HS's 10% bonus, where 40*1.05 should be 42 even.
Durability
 
Posts: 43
Joined: Wed Jun 16, 2010 11:15 am

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby kristoferpally » Thu Jan 13, 2011 11:24 am

so worth wearing or no? my gut says yes but what do I know :oops:
kristoferpally
 
Posts: 13
Joined: Thu Oct 14, 2010 3:34 pm

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby theckhd » Thu Jan 13, 2011 11:55 am

Durability wrote:It truncates, both on the tooltip and some testing I did on beta.

If the tooltip is accurate, it still does it even with HS's 10% bonus, where 40*1.05 should be 42 even.


Does the character sheet read 42% block with Holy Shield active? I was under the impression it was still 41%, indicating that it wasn't multiplicative, but was simply adding a flat 1% to the blocked amount.

My work in the derivations thread leads me to believe that the meta is not 'bugged' at all, but that it was determined to be too powerful at 5% late in beta and was adjusted so that its benefit was closer to what the 2% armor meta grants. In all likelihood, they changed it to a flat 1% and just never updated the tooltip (or they made it a multiplicative 5% and left the tooltip ambiguous, depending on what the HS test tells us).
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 6.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 6211
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby sculder » Thu Jan 13, 2011 1:34 pm

theckhd wrote:
Durability wrote:It truncates, both on the tooltip and some testing I did on beta.

If the tooltip is accurate, it still does it even with HS's 10% bonus, where 40*1.05 should be 42 even.


Does the character sheet read 42% block with Holy Shield active? I was under the impression it was still 41%, indicating that it wasn't multiplicative, but was simply adding a flat 1% to the blocked amount.

My work in the derivations thread leads me to believe that the meta is not 'bugged' at all, but that it was determined to be too powerful at 5% late in beta and was adjusted so that its benefit was closer to what the 2% armor meta grants. In all likelihood, they changed it to a flat 1% and just never updated the tooltip (or they made it a multiplicative 5% and left the tooltip ambiguous, depending on what the HS test tells us).


I've been using the BV meta for a few weeks now and plan on parsing some logs when I have time. Also, I'll test the HS question when I get home.
User avatar
sculder
 
Posts: 113
Joined: Sun Jul 25, 2010 9:34 am
Location: Massachusetts

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby sculder » Thu Jan 13, 2011 7:41 pm

41% on tooltip with hs active
User avatar
sculder
 
Posts: 113
Joined: Sun Jul 25, 2010 9:34 am
Location: Massachusetts

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby theckhd » Thu Jan 13, 2011 8:55 pm

If you have time, take a look at the Call to Arms thread. I think we still want a data set where you take a beating from a mob (hit size >1000 probably) with the meta active so we can determine exactly what the effect of the meta is. Though I suspect based on the tooltip it's a flat 1%.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 6.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 6211
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby Panemtorvra » Thu Jan 13, 2011 10:15 pm

I've taken only a quick look, but given block value's consistency I think it's plenty. I block 44.9% of all damage with BV meta and HS active, likely due to rounding. I've been using the BV meta for a while, I believe it to have been hotfixed several patches ago.
Panemtorvra
 
Posts: 1
Joined: Thu Jan 13, 2011 10:12 pm

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby theckhd » Fri Jan 14, 2011 5:50 am

Do you have parses that I can browse to confirm that? If the meta were actually giving a full 5% block amount, it would be vastly superior to the armor meta.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 6.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 6211
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby sculder » Fri Jan 14, 2011 7:11 am

from the first 20 blocked melee hits from last night's Cho'gall:

Code: Select all
hit   absorb   block   blocked
25682   2197   19374   0.410005714
18494   3543   15314   0.41000241
22811   1186   16676   0.410001721
23856   6600   21165   0.410007555
31761   1086   22826   0.410001257
26572   0000   11938   0.309997403
29068   0000   20200   0.410002436
27737   0000   19275   0.410001702
32265   0000   22422   0.410006034
27639   0000   19207   0.410002989
17730   12997   21353   0.41000384
26181   0000   18194   0.410005634
27597   0000   19178   0.410005345
38098   0000   26475   0.410001084
31997   10550   29567   0.410003605
39096   10826   34692   0.410003073
41972   0000   18857   0.310000164
34989   1819   25579   0.41000529
32829   0000   22814   0.41000665
26833   0000   18647   0.410004398


As you can see, all of the blocked percentages are 41%. (except for the two where I taunted and didn't have HS up)
User avatar
sculder
 
Posts: 113
Joined: Sun Jul 25, 2010 9:34 am
Location: Massachusetts

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby theckhd » Fri Jan 14, 2011 10:04 am

Thank you, that's exactly what I needed.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 6.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 6211
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby sculder » Fri Jan 14, 2011 10:13 am

theckhd wrote:Thank you, that's exactly what I needed.


I'll check a few other fights from this past week just to be sure, but I'm fairly confident the results will be the same
User avatar
sculder
 
Posts: 113
Joined: Sun Jul 25, 2010 9:34 am
Location: Massachusetts

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby cappytoi » Sun Jan 16, 2011 7:35 am

Seal of Truth: All single target attacks (including Judgement, Hammer of Wrath, Exorcism, and Templar's Verdict) can now trigger this seal.


I don't really understand this patch note. According to your work on matlab this is already the case for single target attacks (I checked the ability_model.m updated in 7th December). The only thing missing in code now is to add HoW to proc seals as the latest revision doesn't have that but that doesn't really make too much difference in talent comparison, I guess. Is this patch note meaning that single target attacks also stacks the censure dot (maybe to help for target switching to fix ret on that area?)
cappytoi
 
Posts: 6
Joined: Tue Jan 11, 2011 9:08 pm

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby tlitp » Sun Jan 16, 2011 7:40 am

Let's see some numbers for Mending. For that, we'll use the default raid set (average ilvl ~359), and two common frameworks, CS/SotR and CS/WoG :
Code: Select all
ph=physical hit (%)
exp=expertise (skill)

               Mend.hps@SCSC   Mend.hps@WCWC
ph=0,exp=0          216             206
ph=8,exp=0          236             223
ph=8,exp=26         263             247

Note that :
  • the default talent build doesn't have EG; in order to assess the WoG framework, I've switched from 2/2 SotP to 2/2 EG
  • the default talent set has 3/3 Divinity, increasing Mending's output
  • although these represent the average output, they should be taken as theoretical maxima; in real world scenarios, the overhealing cannot simply be ignored
User avatar
tlitp
 
Posts: 487
Joined: Mon Jul 27, 2009 3:25 pm

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby theckhd » Sun Jan 16, 2011 9:57 am

cappytoi wrote:
Seal of Truth: All single target attacks (including Judgement, Hammer of Wrath, Exorcism, and Templar's Verdict) can now trigger this seal.


I don't really understand this patch note. According to your work on matlab this is already the case for single target attacks (I checked the ability_model.m updated in 7th December). The only thing missing in code now is to add HoW to proc seals as the latest revision doesn't have that but that doesn't really make too much difference in talent comparison, I guess. Is this patch note meaning that single target attacks also stacks the censure dot (maybe to help for target switching to fix ret on that area?)


Judgement, Exorcism, and Hammer of Wrath don't poc seals on live. And nothing but auto-attacks stack Censure on live.

Testing on the PTR so far indicates that both of these will change - Censure will get stacked by all of our single-target attacks, and all three of those abilities should proc seals. This is all subject to change, of course; I'm not sure that getting doubleprocs and double stacks on Judgement (one from Judgement, one from JotJ) will last once Blizz notices it.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 6.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 6211
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby vexryn » Sun Jan 16, 2011 10:59 am

Theck,
I've got a question about the interplay of tps talents from your models on the Talent Comparison.


I'd like to pick up Rule of Law, not for the threat value, but for the 15% crit chance on WoG. I'm trying to get a handle on the threat implications, though.

What I'm finding is that to pick up RoL, I will be dropping both points from Seals of the Pure, and 1 point from Reckoning. You've got values nicely listed for the impact of SotP and Reck per talent point.

But here's my uncertainty: aren't these two actually interrelated? Because my Reckoning swings proc my Seals, doesn't that actually mean that Reckoning is somewhat less valuable without SotP, or conversely that SotP is somewhat less valuable due to fewer Seal procs if I don't have 2/2 Reckoning?

I suppose it's not a big deal either way, but it just struck me that because of the overlapping value of these, removing both points from SotP and 1 point from Reckoning will reduce my tps by somewhat less than would be implied by the tps-per-talent-point calculations shown in the Talent Comparison.
vexryn
Maintankadonor
 
Posts: 154
Joined: Tue Oct 28, 2008 3:12 pm

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby theckhd » Sun Jan 16, 2011 11:32 am

Your intuition is correct. The talent comparison is designed to give "best-case" values for each talent. In some cases that means that a talent is slightly less effective than its maximum value.

For example, in the absence of Reckoning, each point in SotP is worth somewhat less. So if you're comparing a build with 2/2 SotP, 0/2 Reck, and 0/3 Rule of Law to a build with 0/2 SotP, 0/2 Reck, and 2/3 RoL, each point in SotP is actually a smaller loss than the value in the comparison.

Note though that it goes both ways. If you start with a build that has 2/2 SotP and 2/2 Reckoning, which produces X DPS, and you compare to a build with 0/2 SotP and 1/2 Reckoning, you won't get X-2*S-1*R. It will actually be slightly higher than that, because S and R both include a little bit of that interaction.
"Theck, Bringer of Numbers and Pounding Headaches," courtesy of Grehn|Skipjack.
Simcraft 6.x, Call to Arms 6.0, Talent Spec & Glyph Guide 6.x, Blog: Sacred Duty
User avatar
theckhd
Moderator
 
Posts: 6211
Joined: Thu Jul 31, 2008 3:06 pm
Location: Harrisburg, PA

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby cappytoi » Sun Jan 16, 2011 11:51 am

theckhd wrote:
cappytoi wrote:
Seal of Truth: All single target attacks (including Judgement, Hammer of Wrath, Exorcism, and Templar's Verdict) can now trigger this seal.


I don't really understand this patch note. According to your work on matlab this is already the case for single target attacks (I checked the ability_model.m updated in 7th December). The only thing missing in code now is to add HoW to proc seals as the latest revision doesn't have that but that doesn't really make too much difference in talent comparison, I guess. Is this patch note meaning that single target attacks also stacks the censure dot (maybe to help for target switching to fix ret on that area?)


Judgement, Exorcism, and Hammer of Wrath don't poc seals on live. And nothing but auto-attacks stack Censure on live.

Testing on the PTR so far indicates that both of these will change - Censure will get stacked by all of our single-target attacks, and all three of those abilities should proc seals. This is all subject to change, of course; I'm not sure that getting doubleprocs and double stacks on Judgement (one from Judgement, one from JotJ) will last once Blizz notices it.


You are indeed right as usual as I forget to count the judgement procs. The stacking censure part is good news especially for ret and to us from a point. Thanks for your clarification.
cappytoi
 
Posts: 6
Joined: Tue Jan 11, 2011 9:08 pm

Re: Theck's MATLAB thread - Cataclysm/4.x

Postby tlitp » Sat Jan 22, 2011 5:59 am

Let's discuss environmental modifiers (check the supported buff classes in [BM]). Because of synergies, one must consider two approaches : a. no buffs, then adding one buff class at a time; b. full buffs, then subtracting one buff class at a time. They're sorted out, from highest to lowest, based on the average value.

Setup : SoT, regular CS/SotR, default gear set (~359), default talent/glyph configuration.
Code: Select all
high Vengeance : veng=1.0
low Vengeance : veng=0.3
high hit/exp : 8% physical hit, 26 expertise skill
low hit/exp : 0% physical hit, 0 expertise skill


@ high V, low HE
          dps@buffs.off   dps@buffs.on   dps@avg
unrage    723             973            848
soe       494             902            698
coe       484             631            557
lotp      212             530            371
bok       293             445            369
arctac    280             388            334
pwf       252             346            299
savcom    132             185            158
sund      130             181            155
wf        115             161            138
tw         24              35             29
isb         6               9              7



@ low V, low HE
          dps@buffs.off   dps@buffs.on   dps@avg
soe       474             786            630
unrage    392             544            468
coe       310             405            357
lotp      139             342            240
bok       182             281            231
arctac    182             251            216
savcom     87             121            104
sund       86             118            102
pwf        76             104             90
wf         76             104             90
tw         25              35             30
isb         6               6              6



@ high V, high HE
          dps@buffs.off   dps@buffs.on   dps@avg
unrage    882            1184           1033
soe       607            1063            835
coe       559             726            642
bok       356             538            447
lotp      254             598            426
arctac    344             474            409
pwf       308             422            365
savcom    179             249            214
sund      176             243            209
wf        152             210            181
tw         27              38             32
isb         7              10              8



@ low V, high HE
          dps@buffs.off   dps@buffs.on   dps@avg
soe       580             934            757
unrage    478             663            570
coe       358             467            412
bok       220             339            279
lotp      166             387            276
arctac    223             307            265
savcom    118             163            140
sund      116             159            137
wf         99             136            117
pwf        92             126            109
tw         27              38             32
isb         5               7              6

A probably redundant clarification : for 5-man dungeons, check the @buffs.off columns; for unoptimized raids, the @avg columns; for optimized raids, the @buffs.on columns.
User avatar
tlitp
 
Posts: 487
Joined: Mon Jul 27, 2009 3:25 pm

PreviousNext

Return to Advanced Theorycraft and Calculations

Who is online

Users browsing this forum: Bing [Bot] and 1 guest

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: Bing [Bot] and 1 guest