Synopsis
If you can equip a best in slot item with expertise and are over expertise cap rather then equiping something else remove the talent point from Primal Precision.
Sadly there is no real alternative.
[edit]Corrected some math[/edit]
Recently I have heard from 2 different sides that Primal Precision (PP) is not really worth it. So I decided to do some thinking again.
The first was honest mistake. The person thought that the expertise cap was at 15 for tanking. It is however at 15% (generally accepted cap) which for druids is 60 expertise. If you are below 26 expertise any expertise is double the value since until you are dodge capped it counts towards both caps.
The second one was the following: It stated that when expertise capped by gear Primal Precision is not worth it. I wondered how much it was worth really.
So how can I see what it is worth? If you look at gems you will see that 16 expertise = 16 agility = 16 strength = 32 AP.
Now those gems all have different values for each class, making them perfect for one class, more worthless to an other.
If you look at enchants you will see that on gloves for instance 15 Expertise holds the same slot as 44 AP. You can not look at this for real though, since the expertise enchant is way cheaper. An expertise enchant on gloves is of would be a "green" enchant where 44Ap would be a "blue" enchant (as in green "uncommon" gems and blue "rare" gems). To have the same value expertise enchant it should be 22 expertise (if you follow me).
I did the comparison in a different way: I used Rawr (raidbuffed).
I am expertise capped by using Inevitable Defeat (78 exp). To make that useless I equipped Chestguard of the Recluse (83 expertise).
The dps value from Inevitable defeat dropped from 1583 to 1523. So 78 expertise is worth about 60 dps.
I did the same for Chestguard of the Recluse. I unequipped Inevitable Defeat and was still expertise capped. The chastguard's value went up from 314 to 385. That is 71 dps. So it is about the same.
1 expertise ~= 0.8 dps.
So 10 expertise = 2.5% = 2.5* 32 rating = 80 expertise rating.
PP therefore (for me) is worth about 80 dps on gear. When, and only when, you are expertise capped (26 expertise). If I would not have PP I would have to get it on gear loosing 60 dps.
Total dps is 6363 what rawr tells me. If I had to get that by gear alone I would have 6303 dps. 60 / 6303 = 0.95%%, that is 0.48% per talent point.
So that is not really much. It is that there are not really any other talent points for cat that are worth more.
Once again, don't be mistaken. When you are not capped by gear alone the talent point is worth much more since 1 talentpoint will increase dps with 2.5% alone (double parry and dodge cap). I am only saying that if you find a better place for your talentpoint and your best in slot items have expertise plenty... only then you can remove a point in PP.
Showing posts with label Theory Crafting. Show all posts
Showing posts with label Theory Crafting. Show all posts
Sunday, April 26, 2009
Saturday, April 25, 2009
Correction! iMangle does not own
Synopsis
Tarskin has pointed out that I forgot the difference in Maul and Lacerate dots when they get 4% MMS damage. So I did the math again. MMS is better by 1.3%.
Let me state first that this post is a comparison between a very popular tanking spec 0/55/16, holding SS and MSS, and an alternative spec 0/60/11 holding iMangle and no MMS.
I forgot Maul and Lacerate dots in the rotation. The damage for MMS and not MMS is different of course.
So I went back to rawr. Because some values had changed I redid the spreadsheat.
Some weird things I saw; Windfury totem changes the damage per maul. Of course it changes the total damage by Maul, but now also changes the average damage and threat done by Maul apparently. If anyone has an explanation for this I'd like to hear it.
New spreadsheet:
It shows that with WF totem the threat you loose with iMangle rotation is 105 threat per second. This is 1.3% threat loss.
So if you are looking for max threat, MMS spec is the way to go. Well, to be honest, drop something like IW and ask someone else to apply the same debuff. Get iMangle too. That does the most threat.
However you can also ask the few top dpssers to hold back a little bit or use some threat reducing effects (FD) or get hunters and rogues to misdirect to you.
Still, in the end, my conclusion stands I think. 1.3% threat loss on a boss is something not that bad. You get 5% more uptime from Idol of the Corruptor it halves it's downtime. That is worth it I think. If you can dodge an attack that otherwise would have killed you it saves you from a wipe.
PS. This all made me think to what things can you expect normally for being raidbuffed (25 man).
So my next post will hold the raidbuffs you can normally expect in a raid and should be checked while comparing items to eachother.
Tarskin has pointed out that I forgot the difference in Maul and Lacerate dots when they get 4% MMS damage. So I did the math again. MMS is better by 1.3%.
Let me state first that this post is a comparison between a very popular tanking spec 0/55/16, holding SS and MSS, and an alternative spec 0/60/11 holding iMangle and no MMS.
I forgot Maul and Lacerate dots in the rotation. The damage for MMS and not MMS is different of course.
So I went back to rawr. Because some values had changed I redid the spreadsheat.
Some weird things I saw; Windfury totem changes the damage per maul. Of course it changes the total damage by Maul, but now also changes the average damage and threat done by Maul apparently. If anyone has an explanation for this I'd like to hear it.
New spreadsheet:
It shows that with WF totem the threat you loose with iMangle rotation is 105 threat per second. This is 1.3% threat loss.
So if you are looking for max threat, MMS spec is the way to go. Well, to be honest, drop something like IW and ask someone else to apply the same debuff. Get iMangle too. That does the most threat.
However you can also ask the few top dpssers to hold back a little bit or use some threat reducing effects (FD) or get hunters and rogues to misdirect to you.
Still, in the end, my conclusion stands I think. 1.3% threat loss on a boss is something not that bad. You get 5% more uptime from Idol of the Corruptor it halves it's downtime. That is worth it I think. If you can dodge an attack that otherwise would have killed you it saves you from a wipe.
PS. This all made me think to what things can you expect normally for being raidbuffed (25 man).
So my next post will hold the raidbuffs you can normally expect in a raid and should be checked while comparing items to eachother.
Friday, April 24, 2009
Breaking news: iMangle rotation owns!!!
Synopsis
Get yourself to an iMangle rotation, even if you don't have the idol!
[edit]
See next post. The breaking news was broken. It is not true. Thanks go to Tarskin for that. Thank you very much. :-( XD
[/edit]
Either iMangle roation own, or rawr lies.
Inspired by Tarskin's post on Faerie Fire / iMangle rotation I took my head our of my ass and saw that FFF is huge threat. Definately do it every time you can.
I took his post and extended it a bit.
This is a screenshot of my calculations.
First of all the rotation. All rotations are based on a continues cycle. It doesn't say what you should start out with. They all start at the moment lacerate is renewed, thus the first action is Lacerate.
Instead of only looking at Mangle / Fearie Fire I took Lacerate into the cycle. Lacerate can not be allowed to drop, so do that at least every 9 actions. Atfer 10 actions it will have dropped.
6 sec Mangle:
If you are doing Mangle / FFF every 6 seconds, you have to do lacerate every 12 seconds. If you do that every 13.5 seconds if will inevitable collide with Mangle or FFF. It would mean the cycle you would have would be very long, and hard to execute.
That means the rotation would look something like this:
8 moves; 12 seconds: 2 Mangles, 2 FFF, 1 lacerate and 3 swipes.
For refreshing Lacerate it has to be done at least once every 13.5 seconds.
That is coincidentally exactly 3 Mangles. This gives us a 9 move cycle.
9 moves; 13.5 seconds: 3 Mangles, 2 FFF, 1 lacerate, 3 swipes.
It is not disputed that iMangle does more damage. However the fact that if you take iMangle means you will have to drop Master shapeshifter. And everything does less damage.
So from Rawr, using my own gear and fully raidbuffed I took the following average values. I took the average values because you have to count crits too. Here I only discuss threat though damage can also be seen in the spreadsheet:
iMangle spec:
An iMangle 4.5 sec Mangle cycle gives me 28078 threat = 3120 TPS
So taking iMangle over master shapeshifter is 100% the right thing to do and it even leaves 2 extra talentpoints you can spend somethere you like.
Unless ofcourse my rawr data is wrong.
Get yourself to an iMangle rotation, even if you don't have the idol!
[edit]
See next post. The breaking news was broken. It is not true. Thanks go to Tarskin for that. Thank you very much. :-( XD
[/edit]
Either iMangle roation own, or rawr lies.
Inspired by Tarskin's post on Faerie Fire / iMangle rotation I took my head our of my ass and saw that FFF is huge threat. Definately do it every time you can.
I took his post and extended it a bit.
This is a screenshot of my calculations.
First of all the rotation. All rotations are based on a continues cycle. It doesn't say what you should start out with. They all start at the moment lacerate is renewed, thus the first action is Lacerate.
Instead of only looking at Mangle / Fearie Fire I took Lacerate into the cycle. Lacerate can not be allowed to drop, so do that at least every 9 actions. Atfer 10 actions it will have dropped.
6 sec Mangle:
If you are doing Mangle / FFF every 6 seconds, you have to do lacerate every 12 seconds. If you do that every 13.5 seconds if will inevitable collide with Mangle or FFF. It would mean the cycle you would have would be very long, and hard to execute.
That means the rotation would look something like this:
8 moves; 12 seconds: 2 Mangles, 2 FFF, 1 lacerate and 3 swipes.
- Lacerate
- Mangle
- FFF
- Swipe
- Swipe
- Mangle
- FFF
- Swipe
For refreshing Lacerate it has to be done at least once every 13.5 seconds.
That is coincidentally exactly 3 Mangles. This gives us a 9 move cycle.
9 moves; 13.5 seconds: 3 Mangles, 2 FFF, 1 lacerate, 3 swipes.
- Lacerate
- Mangle
- FFF
- Swipe
- Mangle
- Swipe
- FFF
- Mangle
- Swipe
It is not disputed that iMangle does more damage. However the fact that if you take iMangle means you will have to drop Master shapeshifter. And everything does less damage.
So from Rawr, using my own gear and fully raidbuffed I took the following average values. I took the average values because you have to count crits too. Here I only discuss threat though damage can also be seen in the spreadsheet:
iMangle spec:
- Mangle: 4552
- FFF: 3434
- Lacerate: 1128
- Swipe: 2142
- Mangle: 4734
- FFF: 3434
- Lacerate: 1153
- Swipe: 2228
An iMangle 4.5 sec Mangle cycle gives me 28078 threat = 3120 TPS
So taking iMangle over master shapeshifter is 100% the right thing to do and it even leaves 2 extra talentpoints you can spend somethere you like.
Unless ofcourse my rawr data is wrong.
Wednesday, April 22, 2009
Go for improved Mangle!
Synopsis
Is it better to get improved mangle or master shapeshifter? With Idol of the corruptor; I think so. iMangle here I come!
Today I just thought of this when I saw the following item on wowhead. Idol of the Corruptor.
As if it was it's destiny, today we got General Vezax down. And guess what he dropped. And guess who got the item. Yep yours truly has the idol now. Now I don't have the time to do this at some point, but I have to do it now, tomorrow Ulduar is waiting again. (Btw more on the Vezax fight later. It is a really nice druid cat fight).
I tested it very shortly, just before we went Yogg-Saron. I did it on cat, and it has no cooldown and a very high proc rate. I guess it is a copy of Idol of Terror, but then with way higher stats. Also there is no cooldown what soever. I think the Idol of terror used to have one. The duration is also 12 seconds and not 10.
Ok so we know it is nice. The thing is, is it worth it taking iMangle for it?
The proc rate for me calculating miss is 61.5%. (see math block 1 below).
With this we can calculate the uptime on a 6 sec cycle and on a 4.5 sec cycle with a 12 seconds duration. This is a really complex thing. I did an attempt to it (see math block 2 and 3 below). If you have real math send it my way!
6 second cycle:
9.1% downtime.
4.5 second cycle:
4.2% downtime.
The difference is 4.9%
Now simplified; (It is late):
What does it give: 2.44% dodge and 1.98% crit for me.
4.9% extra downtime comes to 0.12% dodge and 0.10% crit.
In bossfights you have to consider the fact that you will be not able to dps some time. Meaning the buff drops often. The sooner it can be up the better. It is a stab in the dark but I would double those values. Then it is pretty significant.
For iMangle we would sacrifice Shapshifter and Master Shapeshifter. You dont want to drop anything else really.
2 points in Master Shapeshifter give 4% damage done = 4% threat.
iMangle gives threat back in a dps cycle. Taking a quick look at an old post by Kalon we can see that iMangle does 2.8% more threat.
I don't now how valid the stats currently are. But I want to believe that what he posted then is in general true.
In total we would loose 1.2% threat and gain dodge and a bit more SD.
Also we can go back to 5/5 furor which is really nice since you can't always choose the moment the pull is gonna be. Having a 100% certain 10 rage can avoid some nasty stuff.
As I have seldom threat issues I am willing to trade the threat for some extra dodge, especially when progressing in Ulduar. This will be my spec as of tomorrow: Link
Below you will find the math I used to calculate the values as mentioned in Block 1, 2 and 3.
Math Block 1
First of all the proc rate. Idol of Terror is said to have 68% proc rate (wowhead.com) I did a quick test on a lvl 83 target dummy. For me I got 52 procs in 51 hits, 28 crits and 6 misses. On hit/crit for me that comes down to 52 / (51 + 28) ~= 66%. So I'll take the 68% as a fact for now.
I also, with current gear, miss 2.33% percent of the time. I get parried (15% cap) 7.25% of the time. So that is a 100 - 2.33 - 7.25 = 90.42% hit chance.
Total percentage of proc = 68% * 90.42% ~= 61.5%. The chance to miss is 38.5%.
Math Block 2
When it procs it has to be renewed within 6 seconds or 12 seconds. So there are 2 chances not to drop it. The chance it drops is when they both miss = 38.5% ^ 2 = 14.8%
Every time mangle is off cd it can be renewed at 38.5% chance.
1 time: 38.5% ^ 1 = 38.5%
2 time: 38.5% ^ 1 = 14.8%
3 time: 38.5% ^ 1 = 5.7%
4 time: 38.5% ^ 1 = 2.2%
5 time: 38.5% ^ 1 = 0.8%
Now there is a 0.8% is will go to 5 misses.
This means that there is a 2.2% - 0.8% chance = 1.4% chance it will only go to 4 misses. This has been implemented below:
So if it drops is will be down for:
30 seconds 0.8% of the time
24 seconds 1.4% of the time
18 seconds 3.5% of the time.
12 sec 9.1% of the time.
6 sec 23.7% of the time.
Average downtime calculation:
time * percentage
-> 6 * 23.7% + 12 * 9.1% + etc.
-> 6 * 0.237 + 12 * 0.091 + etc.
The result is the average downtime.
Average downtime = 3.72 seconds IF it is down. Chance on that is 14.8% so average downtime with 12 seconds is 3.72 * 14.8% = 0.55 seconds per 6 seconds = 9.1% downtime.
Math Block 3
When it procs it has to be renewed within 4.5 seconds, or 9 seconds. If not it can be renewed after only a 1.5 second downtime. So there are 2 chances not to drop it.
(Most of the math is the same as in block 2)
So if it drops is will be down for:
19.5 seconds 0.8% of the time
15 seconds 1.4% of the time
10.5 seconds 3.5% of the time.
6 sec 9.1% of the time.
1.5 sec 23.7% of the time.
Average downtime calculation (see block 2)
Average downtime = 1.70 seconds IF it is down. Chance on that is 14.8% so average downtime with 12 seconds is 1.70 * 14.8% = 0.25 seconds per 6 seconds = 4.2% downtime.
Is it better to get improved mangle or master shapeshifter? With Idol of the corruptor; I think so. iMangle here I come!
Today I just thought of this when I saw the following item on wowhead. Idol of the Corruptor.
As if it was it's destiny, today we got General Vezax down. And guess what he dropped. And guess who got the item. Yep yours truly has the idol now. Now I don't have the time to do this at some point, but I have to do it now, tomorrow Ulduar is waiting again. (Btw more on the Vezax fight later. It is a really nice druid cat fight).
I tested it very shortly, just before we went Yogg-Saron. I did it on cat, and it has no cooldown and a very high proc rate. I guess it is a copy of Idol of Terror, but then with way higher stats. Also there is no cooldown what soever. I think the Idol of terror used to have one. The duration is also 12 seconds and not 10.
Ok so we know it is nice. The thing is, is it worth it taking iMangle for it?
The proc rate for me calculating miss is 61.5%. (see math block 1 below).
With this we can calculate the uptime on a 6 sec cycle and on a 4.5 sec cycle with a 12 seconds duration. This is a really complex thing. I did an attempt to it (see math block 2 and 3 below). If you have real math send it my way!
6 second cycle:
9.1% downtime.
4.5 second cycle:
4.2% downtime.
The difference is 4.9%
Now simplified; (It is late):
What does it give: 2.44% dodge and 1.98% crit for me.
4.9% extra downtime comes to 0.12% dodge and 0.10% crit.
In bossfights you have to consider the fact that you will be not able to dps some time. Meaning the buff drops often. The sooner it can be up the better. It is a stab in the dark but I would double those values. Then it is pretty significant.
For iMangle we would sacrifice Shapshifter and Master Shapeshifter. You dont want to drop anything else really.
2 points in Master Shapeshifter give 4% damage done = 4% threat.
iMangle gives threat back in a dps cycle. Taking a quick look at an old post by Kalon we can see that iMangle does 2.8% more threat.
I don't now how valid the stats currently are. But I want to believe that what he posted then is in general true.
In total we would loose 1.2% threat and gain dodge and a bit more SD.
Also we can go back to 5/5 furor which is really nice since you can't always choose the moment the pull is gonna be. Having a 100% certain 10 rage can avoid some nasty stuff.
As I have seldom threat issues I am willing to trade the threat for some extra dodge, especially when progressing in Ulduar. This will be my spec as of tomorrow: Link
Below you will find the math I used to calculate the values as mentioned in Block 1, 2 and 3.
Math Block 1
First of all the proc rate. Idol of Terror is said to have 68% proc rate (wowhead.com) I did a quick test on a lvl 83 target dummy. For me I got 52 procs in 51 hits, 28 crits and 6 misses. On hit/crit for me that comes down to 52 / (51 + 28) ~= 66%. So I'll take the 68% as a fact for now.
I also, with current gear, miss 2.33% percent of the time. I get parried (15% cap) 7.25% of the time. So that is a 100 - 2.33 - 7.25 = 90.42% hit chance.
Total percentage of proc = 68% * 90.42% ~= 61.5%. The chance to miss is 38.5%.
Math Block 2
When it procs it has to be renewed within 6 seconds or 12 seconds. So there are 2 chances not to drop it. The chance it drops is when they both miss = 38.5% ^ 2 = 14.8%
Every time mangle is off cd it can be renewed at 38.5% chance.
1 time: 38.5% ^ 1 = 38.5%
2 time: 38.5% ^ 1 = 14.8%
3 time: 38.5% ^ 1 = 5.7%
4 time: 38.5% ^ 1 = 2.2%
5 time: 38.5% ^ 1 = 0.8%
Now there is a 0.8% is will go to 5 misses.
This means that there is a 2.2% - 0.8% chance = 1.4% chance it will only go to 4 misses. This has been implemented below:
So if it drops is will be down for:
30 seconds 0.8% of the time
24 seconds 1.4% of the time
18 seconds 3.5% of the time.
12 sec 9.1% of the time.
6 sec 23.7% of the time.
Average downtime calculation:
time * percentage
-> 6 * 23.7% + 12 * 9.1% + etc.
-> 6 * 0.237 + 12 * 0.091 + etc.
The result is the average downtime.
Average downtime = 3.72 seconds IF it is down. Chance on that is 14.8% so average downtime with 12 seconds is 3.72 * 14.8% = 0.55 seconds per 6 seconds = 9.1% downtime.
Math Block 3
When it procs it has to be renewed within 4.5 seconds, or 9 seconds. If not it can be renewed after only a 1.5 second downtime. So there are 2 chances not to drop it.
(Most of the math is the same as in block 2)
So if it drops is will be down for:
19.5 seconds 0.8% of the time
15 seconds 1.4% of the time
10.5 seconds 3.5% of the time.
6 sec 9.1% of the time.
1.5 sec 23.7% of the time.
Average downtime calculation (see block 2)
Average downtime = 1.70 seconds IF it is down. Chance on that is 14.8% so average downtime with 12 seconds is 1.70 * 14.8% = 0.25 seconds per 6 seconds = 4.2% downtime.
Friday, March 13, 2009
New PTR Patch
Synopsis
New things on the PTR's latest patch; t8 bonus calculations and enchant.
Warning if you are weak on math, don't break your head over this, just assume the numbers are true!
As posted by Karthis there is a new PTR patch available at the moment (mmo champion)
Before joining in on the tier 8 bonus I'd like to make 2 remarks on other things in the patch.
However, depending on the exact workings of this enchant it could be really good when you have 2 weapons!
Start of the fight you equip the Blood Draining (BD) and you get the buff. Once you have it press a macro and switch to the Mongoose Weapon. As soon as the buff is gone (this can either be that you dropped below 35% or the buff timer is done) re-equip BD again till you have to buff.
2 ways this would not work, would maybe work, but would be too much of a hassle to do.
On a side note, I am also still unable to research new glyphs. I now even get a message I have no target (probably a placeholder message to not waste research materials when you research while nothing is available).
So, now to the tier 8 bonus sets as they are currently (everything can change on PTR) available.
Now Cat get a huge boost from the 2t8 bonus. No question about that. The real thing we have to look at here is: does 4t8 outweigh 2t7 buff.
With the new shred glyph Rip would last 22 seconds instead of 26. It has to be up at 5CP always (preferably). 100% uptime means no loss in DPS, it only means that you have 4 seconds less to gather 5 CP to renew Rip.
As can be seen from the calculations below (**) you can see to need 5 CP you need an average of 3.5625 CP generating abilitities. I take an average of energy needed for one ability of 40 (some are a bit above, others are below).
Thus to do Rip you need 3.6525 * 40 + 30 (Rip itself) = 172.5 energy.
Loosing 4 seconds on Rip we would need now 172.5 * (26/22) = 203.86 energy to keep Rip up the same time.
That is an energy loss per second of (203.86 - 172.5) / 26 = 1.21 energy lost per second.
Now the 4t8 bonus would give 8 seconds on SR. At the moment most people use a 4 CP SR (see below *) rotation. A 1 CP SR rotation would last to short to safely do a Rip. With the 8 seconds longer SR I'll take a look at both.
A few things up front: (wee below **).
The average CP generating abilities needed to do 1 CP SR = 1. With a proc chance of 50% it will eaither be 1 CP SR or 2 CP SR. It will thus last an average time of (0.5 * 14 + 0.5 * 19 =) 16.5 seconds (not counting 4t8).
The average CP generating abilities needed todo 4 CP SP = 2.875. With a proc chance of 50% it will be either 4 CP SR (68.75%) or 5 CP SR (31.25%). It will thus last an average of time of (0.6875 * 29 + 0.3125 * 34 =) 30.56 seconds (not counting 4t8).
Using no 4t8, below we see that 1 CP SR is more energy efficient then 4 CP SR.
1 CP SR = 16.5 sec / (40 * 1 + 25 energy) = 16.5/65 = 0.254 sec uptime per energy.
4 CP SR = 30.56 sec / (40 * 2.875 + 25 energy) = 30.56 / 140 = 0.218 sec uptime per energy.
The ratio is 0.254 / 0.218 = 1.164 = 16.4% more efficient.
However as said 1 CP Sr can't be used reliably.
Using new 4t8 we can add 8 seconds to the duration.
1 CP SR = 16.5 + 8 sec / 65 energy 24.5/65 = 0.377 sec uptime per energy.
4 CP SR = 30.56 + 8 sec / 140 energy = 38.56 / 140 = 0.275 sec uptime per energy.
The ratio is 0.377 / 0.275 = 1.369 = 36.9% more efficient.
Now not having unreliable 1 CP SR uptime and a efficientce of nearly 37% vs 4 CP SR you would be stupid not stay using 4 CP SR.
But now what does that mean for engergy save vs the energy lost.
4 CP SR costs 140 energy (see above). 4t8 gives a 38.56 sec duration vs 30.56 duration.
Without 4t8 it would cost 140 / 30.56 = 4.58 energy per second.
With 4t8 it would cost 140 / 38.56 = 3.63 energy per second.
4.58 - 3.63 = 0.95 energy per second gain.
This does not weigh up to the 1.21 energy per second lost on rip.
Since we are gonna use 1 CP SR let's see how that does;
Without 4t8 it would cost 65 / 16.5 = 3.94 energy per second.
With 4t8 it would cost 65 / 24.5 = 2.65 energy per second.
3.94 - 2.65 = 1.28 energy per second gain.
This does weigh up to the 1.21 energy per second lost on rip.
So this proves the 4t8 bonus next to the better stats is already better for people who were using (for me) unreliable 1 CP SR.
But if you were using 4 CP SR the bonus is huge!
4.58 - 2.65 = 1.93 energy per second gain.
This completely obliterates the 1.21 energy per second lost on rip.
In conclusion this makes me say definitely go for 4t8! It makes the rotation so much easier ( I suspect it does, it will be months before it is thoroughly tested) and more energy efficient!
Any input on this is welcome.
* Why 4 CP and not 5? If you go for a 5CP SR you have the chance of critting while you are at 4 CP. Meaning it is a wasted crit! You only get 1 CP anyway. Therefore when you are at 4 CP, use SR, don't go for a 5th. If you are at 3 and you crit you might be at 5 CP anyway, but at least you did not waste a crit!
** The ways to get to 4 CP are:
Situation => Chance on situation => Sum of moves
Hit + hit + hit + (hit/crit) => (50%^3) = 12.5% => 4
Hit + hit + crit => (50%^3) = 12.5% => 3
Hit + crit + (hit/crit) => (50%^2) = 25% => 3
Crit + hit + (hit/crit) => (50%^2) = 25% => 3
Crit + crit => (50%^2) = 25% => 2
Sum percentages = 100%
Average finishing moves needed = 2.875
Percentage of 5 combopoints: (12.5% + 25% + 25%) / 2 = 31.25%
The ways to get to 5 CP are:
Situation => Chance on situation => Sum of moves
Hit + hit + hit + hit + (hit/crit) => (50%^4) = 6.25% => 5
Hit + hit + hit + crit => (50%^4) = 6.25% => 4
Hit + hit + crit + (hit/crit) => (50%^3) = 12.5% => 4
Hit + crit + hit + (hit/crit) => (50%^3) = 12.5% => 4
Crit + hit + hit + (hit/crit) => (50%^3) = 12.5% => 4
Crit + hit + crit => (50%^3) = 12.5% => 3
Hit + crit + crit => (50%^3) = 12.5% => 3
Crit + crit + (hit/crit) => (50%^2) = 25% => 3
Sum percentages = 100%
Average finishing moves needed = 3.5625
New things on the PTR's latest patch; t8 bonus calculations and enchant.
Warning if you are weak on math, don't break your head over this, just assume the numbers are true!
As posted by Karthis there is a new PTR patch available at the moment (mmo champion)
Before joining in on the tier 8 bonus I'd like to make 2 remarks on other things in the patch.
- Faerie Fire (Feral) now deals [15% of AP + 1] damage. (Up from [5% of AP + 1])
- Enchant Weapon - Blood Draining *New* - Permanently enchants your weapon to sometimes grant Blood Reserve when striking an enemy or inflicting damage with bleed attacks. When you fall below 35% health, Blood Reserve restores 1750 to 2250 health. This enchantment requires the wielder is at least level 75.
However, depending on the exact workings of this enchant it could be really good when you have 2 weapons!
Start of the fight you equip the Blood Draining (BD) and you get the buff. Once you have it press a macro and switch to the Mongoose Weapon. As soon as the buff is gone (this can either be that you dropped below 35% or the buff timer is done) re-equip BD again till you have to buff.
2 ways this would not work, would maybe work, but would be too much of a hassle to do.
- The buff is removed when you have no weapon with BD equipped.
- The buff only lasts a couple of seconds (10 or so). This would mean too much switching.
On a side note, I am also still unable to research new glyphs. I now even get a message I have no target (probably a placeholder message to not waste research materials when you research while nothing is available).
So, now to the tier 8 bonus sets as they are currently (everything can change on PTR) available.
- 2t8: The periodic damage dealt by your Rake, Rip, and Lacerate abilities has a chance to cause you to enter a Clearcasting state.
- 4t8: Increases the duration of Savage Roar and Survival Instinct by 8 sec.
Now Cat get a huge boost from the 2t8 bonus. No question about that. The real thing we have to look at here is: does 4t8 outweigh 2t7 buff.
With the new shred glyph Rip would last 22 seconds instead of 26. It has to be up at 5CP always (preferably). 100% uptime means no loss in DPS, it only means that you have 4 seconds less to gather 5 CP to renew Rip.
As can be seen from the calculations below (**) you can see to need 5 CP you need an average of 3.5625 CP generating abilitities. I take an average of energy needed for one ability of 40 (some are a bit above, others are below).
Thus to do Rip you need 3.6525 * 40 + 30 (Rip itself) = 172.5 energy.
Loosing 4 seconds on Rip we would need now 172.5 * (26/22) = 203.86 energy to keep Rip up the same time.
That is an energy loss per second of (203.86 - 172.5) / 26 = 1.21 energy lost per second.
Now the 4t8 bonus would give 8 seconds on SR. At the moment most people use a 4 CP SR (see below *) rotation. A 1 CP SR rotation would last to short to safely do a Rip. With the 8 seconds longer SR I'll take a look at both.
A few things up front: (wee below **).
The average CP generating abilities needed to do 1 CP SR = 1. With a proc chance of 50% it will eaither be 1 CP SR or 2 CP SR. It will thus last an average time of (0.5 * 14 + 0.5 * 19 =) 16.5 seconds (not counting 4t8).
The average CP generating abilities needed todo 4 CP SP = 2.875. With a proc chance of 50% it will be either 4 CP SR (68.75%) or 5 CP SR (31.25%). It will thus last an average of time of (0.6875 * 29 + 0.3125 * 34 =) 30.56 seconds (not counting 4t8).
Using no 4t8, below we see that 1 CP SR is more energy efficient then 4 CP SR.
1 CP SR = 16.5 sec / (40 * 1 + 25 energy) = 16.5/65 = 0.254 sec uptime per energy.
4 CP SR = 30.56 sec / (40 * 2.875 + 25 energy) = 30.56 / 140 = 0.218 sec uptime per energy.
The ratio is 0.254 / 0.218 = 1.164 = 16.4% more efficient.
However as said 1 CP Sr can't be used reliably.
Using new 4t8 we can add 8 seconds to the duration.
1 CP SR = 16.5 + 8 sec / 65 energy 24.5/65 = 0.377 sec uptime per energy.
4 CP SR = 30.56 + 8 sec / 140 energy = 38.56 / 140 = 0.275 sec uptime per energy.
The ratio is 0.377 / 0.275 = 1.369 = 36.9% more efficient.
Now not having unreliable 1 CP SR uptime and a efficientce of nearly 37% vs 4 CP SR you would be stupid not stay using 4 CP SR.
But now what does that mean for engergy save vs the energy lost.
4 CP SR costs 140 energy (see above). 4t8 gives a 38.56 sec duration vs 30.56 duration.
Without 4t8 it would cost 140 / 30.56 = 4.58 energy per second.
With 4t8 it would cost 140 / 38.56 = 3.63 energy per second.
4.58 - 3.63 = 0.95 energy per second gain.
This does not weigh up to the 1.21 energy per second lost on rip.
Since we are gonna use 1 CP SR let's see how that does;
Without 4t8 it would cost 65 / 16.5 = 3.94 energy per second.
With 4t8 it would cost 65 / 24.5 = 2.65 energy per second.
3.94 - 2.65 = 1.28 energy per second gain.
This does weigh up to the 1.21 energy per second lost on rip.
So this proves the 4t8 bonus next to the better stats is already better for people who were using (for me) unreliable 1 CP SR.
But if you were using 4 CP SR the bonus is huge!
4.58 - 2.65 = 1.93 energy per second gain.
This completely obliterates the 1.21 energy per second lost on rip.
In conclusion this makes me say definitely go for 4t8! It makes the rotation so much easier ( I suspect it does, it will be months before it is thoroughly tested) and more energy efficient!
Any input on this is welcome.
* Why 4 CP and not 5? If you go for a 5CP SR you have the chance of critting while you are at 4 CP. Meaning it is a wasted crit! You only get 1 CP anyway. Therefore when you are at 4 CP, use SR, don't go for a 5th. If you are at 3 and you crit you might be at 5 CP anyway, but at least you did not waste a crit!
** The ways to get to 4 CP are:
Situation => Chance on situation => Sum of moves
Hit + hit + hit + (hit/crit) => (50%^3) = 12.5% => 4
Hit + hit + crit => (50%^3) = 12.5% => 3
Hit + crit + (hit/crit) => (50%^2) = 25% => 3
Crit + hit + (hit/crit) => (50%^2) = 25% => 3
Crit + crit => (50%^2) = 25% => 2
Sum percentages = 100%
Average finishing moves needed = 2.875
Percentage of 5 combopoints: (12.5% + 25% + 25%) / 2 = 31.25%
The ways to get to 5 CP are:
Situation => Chance on situation => Sum of moves
Hit + hit + hit + hit + (hit/crit) => (50%^4) = 6.25% => 5
Hit + hit + hit + crit => (50%^4) = 6.25% => 4
Hit + hit + crit + (hit/crit) => (50%^3) = 12.5% => 4
Hit + crit + hit + (hit/crit) => (50%^3) = 12.5% => 4
Crit + hit + hit + (hit/crit) => (50%^3) = 12.5% => 4
Crit + hit + crit => (50%^3) = 12.5% => 3
Hit + crit + crit => (50%^3) = 12.5% => 3
Crit + crit + (hit/crit) => (50%^2) = 25% => 3
Sum percentages = 100%
Average finishing moves needed = 3.5625
Subscribe to:
Posts (Atom)
