Feral talents/abilities - WotLK discussion - Page 190

Valerian · 10/08/08, 6:43 PM

Whats the difference between a pure rip and pure FB finisher cycle, as gear changes? The numbers nightcrowler showed was ~80 dps. Any idea if this will get worse or better with better gear? Frankly I'll sacrifice 80 dps for a far simpler cycle mainly due to the fact that I imagine messing up the "better" cycle will cost even more DPS. I was having an annoying enough time on the PTR juggling the 9 second rake with 12 second rips and mangles (didn't have the glyphs). The only concern with this sacrifice is if that dps loss becomes more substantial as gear improves.

Diameter · 10/08/08, 7:14 PM

I hope I'm not the only person who got lost on your latest dps calcs nightcrowler.

What is a "free-SR" or "half free SR" that you reference in your different cycles? I'm trying to keep up with what will be the best dps rotation but if you could explain that a little better I'd appreciate it.

Biarch · 10/08/08, 8:01 PM

I pretty much have my lvl 70 tank spec sorted but i am really unsure whether to get OoC or berserking with that last point. Any Ideas?

cana · 10/08/08, 8:05 PM

Originally Posted by Diameter

I hope I'm not the only person who got lost on your latest dps calcs nightcrowler.

What is a "free-SR" or "half free SR" that you reference in your different cycles? I'm trying to keep up with what will be the best dps rotation but if you could explain that a little better I'd appreciate it.

AFAIK "free-SR" means "SR as soon as it wears off, regardless of your current CPs", while "half free SR" means "after SR faded, build 4 CPs and do SR"

Davaeorn · 10/08/08, 8:10 PM

Originally Posted by Biarch

I pretty much have my lvl 70 tank spec sorted but i am really unsure whether to get OoC or berserking with that last point. Any Ideas?

Berserking is probably going to give you more bang for your buck with 1-3 mangles on every GCD for 15 seconds every 3 minutes.

Soultrigger · 10/08/08, 11:03 PM

I understand that so far we are comparing skills and when to use them, but can you please try to see if this rotation would be ideal or a dps loss, the math behind it is just that all the rotation has to support itself. (before rake buffs, rakes were shreds, and rip lasted enough for FB to get the 50% crit bonus). All I did is to make sure in every situation the buffs were up, rake will fall for 1 or 2 seconds at most (because of energy regen), the only issues are two: 1) One shred will always go off without the SR up (btw, I am thinking this rotation with 5 point SR, 5 point Rip, 5 point FB), and 2) the other is that between the last rake in the rotation and the first one, there will be a 10 seconds without rake debuff, because we need to allow the energy to recharge to restart the rotation. No idea if it is any dps gain actually, because I am not that math guy. Oh, another detail, for every 3 skills that generate combo points, I am counting 2 crits for 5 points, or hoping that the OoC procs make up for non crits streaks. Also I am considering mangle and rip glyphs, and 2 pieces T7.

(wait for 74 energy when it is the second time you start the rotation, btw 3.9s means 3.9 seconds to reflect energy regen time)

Mangle (3.9s)+Rake (3.5s)+Shred(4,2s)+Savage Roar(2.5s)

(rake will drop for 1.2 secs)

+Rake(3.5s)+Shred(4.2s)+Tiger´s Fury(-6.0s)+Mangle(3.9s)+Rip(3.0s)

(On this part, wait for 9 energy to use TF, simply because you will save the energy regen time to do double rip+mangle to avoid ticks without mangle debuff, btw you are renewing MG at 17.9s, just 0.1 secs before it expires)

+Rake(3.5s)+Shred(4.2s)+Shred(4.2s)+Ferocious Bite(3.5s)

(the rake on this rotation, because of TF, will be at 8.6 sec, so you might want to let the dot tick off before reapplying it. When this rake finishes, it will take about 10 second before you can have a rake up again, this is because of the energy cost of ferocious bite, mangle and rake).

This is the best I could come up with as a rotation to use the new buffed feral skills, but maybe it is not the best dps rotation, but I felt like presenting it for evaluation.

tangedyn · 10/08/08, 11:51 PM

Originally Posted by Taudark

I'm getting pretty good results with 2 SR/5 rip.

On the simulator? How does it compare against the baseline strategy?

Originally Posted by Valerian

Whats the difference between a pure rip and pure FB finisher cycle, as gear changes? The numbers nightcrowler showed was ~80 dps. Any idea if this will get worse or better with better gear? Frankly I'll sacrifice 80 dps for a far simpler cycle mainly due to the fact that I imagine messing up the "better" cycle will cost even more DPS. I was having an annoying enough time on the PTR juggling the 9 second rake with 12 second rips and mangles (didn't have the glyphs). The only concern with this sacrifice is if that dps loss becomes more substantial as gear improves.

You can try this out on Wowcatsim yourself, but here's what I've found:
Using the default stats on Wowcatsim:
Baseline strategy: 2275 dps
Baseline strategy with FB: 2090 dps

800 AP, 50% crit, 8% hit, 5% expertise, 30% haste:
Baseline strategy: 3450 dps
Baseline strategy with FB: 3338 dps

Considering the amount of talent points in FA it takes to buff up FB, I think it's probably better to spend talent points elsewhere, concentrate on using Rip, and use FB only as an energy sink when SR/Rip are all up.

nightcrowler · 10/09/08, 2:37 AM

ERROR FIX
I made an error in my simulation about FB, I've computed Savage Roar 2 times. The program is now fixed. The new dps value are:

Free-SR/RIP+FB: 4991
Free-SR/RIP: 4838

About FB vs RIP question:
I don't know how tange simulator works and ability/buffs implemented, but let's do some math:

suppose 50% crit, hit and expertise capped (like in my simulation):

Base FB: 0.35*AP+1640 +extra_energy*(9.4+AP/410)
Now suppose that you have PI and crit meta gem and feral aggression and naturalist.

FB damage = Base FB * 2*1.1*1.03*1.15*1.1 =1.00*AP+4701+extra_energy*(27+0.007*AP)

Base RIP: (534+0.05*AP) x tick
Suppose you have mangle up and naturalist:
RIP damage = (764+0.0715*AP) x tick

Suppose that on average we will use FB at 40 energy (from my simulation you usually use it from 36 to 42 energy).

FB DPE = (0.025875*AP+121)
RIP DPE =(0.002383*AP + 25.5) x Tick

Let's assume a 25% damage reduction on the boss:

FB DPE = (0.019406*AP+91)

So we can find the number of tick to make them equal:

Tick = (0.019406*AP+91) / (0.002383*AP + 25.5)

AP                 APbuff         Ticks
0         	     0	         4
1000	         1540	         4
2000	         3080	         5
3000	         4620	         5
4000	         6160	         5
5000	         7700	         6
6000	         9240	         6
7000	         10780	         6
8000	         12320	         6
9000	         13860	         6
10000	         15400	         6
11000	         16940	         6
12000	         18480	         6
13000	         20020	         7
14000	         21560	         7
15000	         23100	         7

AP Buff = AP * 1.4 * 1.1 (Savage Roar + Unleash rage, thinking about a fix...)

Conclusion:
Considering that base RIP is 6 ticks and with glyph is 8 ticks and with 2T7 is 9 ticks (10 if they fix it), it seems that as ppl stated above RIP is always better than FB. BUT IT'S NOT TRUE. The reason is simple: RIP should ticks to be better. Suppose you have 8k AP. So from the table above you know that you need 6 ticks to make RIP better than FB, you also have RIP glyph. So your RIP ticks 8 times for 16 seconds. IF you have time to let your RIP expire obviosly RIP is always better, but usually that's not the case. Usually you take about 10 seconds to reach 5 CP. So suppose that you have used RIP but only 10 seconds are lapsed (5 ticks), what do you do? Ferocious Bite.

Other notes:
The table I wrote above is determinated by a simplified scheme. In my simulation I use a formula to find the energy value at wich FB is better than RIP for each tick, buff, debuff, etc..

the formula is the following:

energy_threasold=
fb.energy+((((rip.hit*(ap*0.39*(1+src)+4165.2)*(rem*2.0)/(12.0)))) + ((1.0*(fb.energy-rip.energy))*((1-red)*(((ap/14.0)*2.25)*(1+src)+(weapon*2.25+742.5))*(1+ret_m1*bdebuff)*1.3*(shred.hit+shred.crit*2*(1+pistinct_m)*(1+META*0.03)))/(1.0*shred.energy)) - ((1-red)*(1+ferocity_m)*(0.35*(1+src)*ap+ 1640)*(fb.hit+fb.crit*2*(1+pistinct_m)*(1+META*0.03)))) /(((1-red)*(1+ferocity_m)*(ap*(1+src)*0.002439+9.4)*(fb.hit+fb.crit*2*(1+pistinct_m)*(1+META*0.03)) - (((1-red)*(((ap/14.0)*2.25)*(1+src)+(weapon*2.25+742.5))*(1+ret_m1*bdebuff)*1.3*(shred.hit+shred.crit*2*(1+pistinct_m)*(1+META*0.03)))/(1.0*shred.energy))));

I hope the formulas talk by himself: ret = Rend and Tear, Meta is the +3% crit meta gem, src = 0.4 (savage roar), red is the damage reduction, weapon is the weapon damage (55.4 + 14 from stones) and so on.

About scaling:
I'm worried about our scaling in t8 and t9 raids, at t8 level we will have expertise and hit capped, and FB-crit capped (crit% > 50%).

Here is an example report that make clear the FB vs RIP question:

Ferocious Bite vs RIP, mob IS BLEEDING

Use Ferocious Bite for any energy value, Rip ticks: 1, Rip debuff counter: 17
Use Ferocious Bite for any energy value, Rip ticks: 2, Rip debuff counter: 15
Use Ferocious Bite for any energy value, Rip ticks: 3, Rip debuff counter: 13
Use Ferocious Bite for any energy value, Rip ticks: 4, Rip debuff counter: 11
Use Ferocious Bite if you have less than: 97 energy, Rip ticks: 5, Rip debuff counter: 9
Use Ferocious Bite if you have less than: 68 energy, Rip ticks: 6, Rip debuff counter: 7
Use Ferocious Bite if you have less than: 40 energy, Rip ticks: 7, Rip debuff counter: 5
Use RIP, ticks: 8, Rip debuff counter: 3
Use RIP, ticks: 9, Rip debuff counter: 1


Ferocious Bite vs RIP, mob is NOT BLEEDING

Use Ferocious Bite for any energy value, Rip ticks: 1, Rip debuff counter: 17
Use Ferocious Bite for any energy value, Rip ticks: 2, Rip debuff counter: 15
Use Ferocious Bite for any energy value, Rip ticks: 3, Rip debuff counter: 13
Use Ferocious Bite for any energy value, Rip ticks: 4, Rip debuff counter: 11
Use Ferocious Bite for any energy value, Rip ticks: 5, Rip debuff counter: 9
Use Ferocious Bite if you have less than: 95 energy, Rip ticks: 6, Rip debuff counter: 7
Use Ferocious Bite if you have less than: 47 energy, Rip ticks: 7, Rip debuff counter: 5
Use RIP, ticks: 8, Rip debuff counter: 3
Use RIP, ticks: 9, Rip debuff counter: 1

RIP: 12.4479% of total damage (SR: 15.8804%, Avg dmg: 15764.5, #/cycle: 0.965188)
Ferocious Bite: 7.6872% of total damage (SR: 29.6877%, Avg dmg: 14241.5, #/cycle: 0.659797, Average Energy Usage: 38)

N.B. the cycle starts-ends when you use SR.

@Tange: have you implemented King of the jungle and Berserk? Wich buff do you use? Haste? Boss armor? Armor penetration? If you give me talent/buff/stats info you use in your simulation I can try to figure out the difference with mine.

tangedyn · 10/09/08, 3:54 AM

If you apply Rip immediately as long as Savage Fury is up and Combo Points = 5, then you will be wasting Rip ticks.
If you notice, my Strategies don't do that.

For my Baseline strategy, it will not cast Rip until the previous Rip has expired:

    // If Savage Roar is up ap CP = 5, cast Rip
    if (status.getComboPoints() >= 5) {
        if (status.getRipTimer() == 0 && (status.getEnergy() >= 30 || status.isOocUp()))
            return Action.RIP;
        else
            return Action.NONE;
    }

However, sometimes Savage Fury will expire before Rip expires. In this case, Savage Fury will be casted instead of Rip. Occasionally, we may end up wasting energy when the energy bar reaches 100 before Rip expires. For my implementation of your strategy, I calculate whether my energy will hit 90 before Rip expires to decide if I should Bite immediately or wait for Rip to expire.

    // If Savage Roar is up ap CP = 5, cast Rip
    if (status.getComboPoints() >= 5) {
        if (status.getRipTimer() == 0 && (status.getEnergy() >= 30 || status.isOocUp()))
            return Action.RIP;
        // If I will hit 90 energy before rip tick is over, FB now
        else if (status.getRipTimer() > (90 - status.getEnergy()) && (status.getEnergy() >= 35 || status.isOocUp()))
            return Action.BITE;
        else
            return Action.NONE;
    }

If your simulator is RIPing immediately and canceling previous ticks, then naturally it will diminish the DPE of Rip.

nightcrowler · 10/09/08, 3:58 AM

Originally Posted by tangedyn

For my Baseline strategy, it will not cast Rip until the previous Rip has expired:

That's make all clear, obviosly if you wait for rip to expire, RIP will always beats FB, I'll try to implement that strategy in my simulation and see if it's better than RIP+FB without waiting.

P.S.
You should edit your post, I think you mean Savage roar not savage fury.

EDIT: I've noticed that you use OOC for all ability, In my scheme I always Shred with OOC, it ends up with higher dps (at least for me).

I've implemented that strategy but it doesn't work well (at least for me):

Free-SR + RIP (wating for expire) = 4721 dps. Also this way it increase the dps component due to RIP to about 18% of total dps.

This is the piece of code where I choce from RIP or FB:

ripvsfb=fb.energy+((((rip.hit*(ap*0.39*(1+src)+4165.2)*(rem*2.0)/(12.0)))) + ((1.0*(fb.energy-rip.energy))*((1-red)*(((ap/14.0)*2.25)*(1+src)+(weapon*2.25+742.5))*(1+ret_m1*bdebuff)*1.3*(shred.hit+shred.crit*2*(1+pistinct_m)*(1+META*0.03)))/(1.0*shred.energy)) - ((1-red)*(1+ferocity_m)*(0.35*(1+src)*ap+ 1640)*(fb.hit+fb.crit*2*(1+pistinct_m)*(1+META*0.03)))) /(((1-red)*(1+ferocity_m)*(ap*(1+src)*0.002439+9.4)*(fb.hit+fb.crit*2*(1+pistinct_m)*(1+META*0.03)) - (((1-red)*(((ap/14.0)*2.25)*(1+src)+(weapon*2.25+742.5))*(1+ret_m1*bdebuff)*1.3*(shred.hit+shred.crit*2*(1+pistinct_m)*(1+META*0.03)))/(1.0*shred.energy))));

						
						// savage roar?
						
						if((free_sr==1)&&(sr==0)&&(en>=sr_energy)&&(cp>=sr_t)){
							////cout << "\n" <<"Energy: " << en << ", Time: " << (1.0*timer)/100.0 << ", CP: " << cp <<"\n";
							//cout << "Time: " << timer/100.0 << ", Energy: " << en <<"\n";
							//cout << "SR\n";
							sr_timer=(9+cp*5)*100;
							en-=sr_energy;
							cp=0;
							sr=1;
							gcd=100+lag;
							cycle_time+=1;
							cycle=2;
							//	//cout << "mana: " << mana <<"\n";
							s_count=0;
							goto fine;
						}
						
						//ferocious bite?
						
						if((en>=fb.energy)&&(cp>=fb_t)&&(((bdebuff==1)&&(en<=(ripvsfb-(int)(lag/10))))||(rip_t>5))){
							
							att=randnum();
							//	////cout << "Attack roll: " << att << "\n";
							
							if(att<fb.crit){
								//cout << "Time: " << timer/100.0 << ", Energy: " << en <<"\n";
								fb.adamage=(1-red)*((0.07*ap*(1+sr*src))*cp+(ap*(1+sr*src)*0.002439)*(en-fb.energy+(int)(lag/10.0)))*(1+naturalist_m)*(1+ferocity_m);
								fb.acritdamage=fb.adamage*2*(1+pistinct_m)*(1+META*0.03);
								fb.bdamage=(1-red)*(190+290*cp+(9.4)*(en-fb.energy+(int)(lag/10.0)))*(1+naturalist_m)*(1+ferocity_m);
								fb.bcritdamage=fb.bdamage*2*(1+pistinct_m)*(1+META*0.03);
								fb.hit=yhit-fb.crit;
								dmg=(1+RA*0.02)*((fb.acritdamage)+fb.bcritdamage);
								damage+=dmg;
								ydamage+=dmg;						
								fb_damage_tot+=dmg;
								fbnum++;
								avgfbenergy+=en+(int)(lag/10.0);
								fb_damage+=(1+RA*0.02)*fb.acritdamage*(sr*src);
								en=-(int)(lag/10.0);
								cp=0;
								gcd=100+lag;	
								
								if(doublecycle==1){
									switch (doublecycle_counter){
										case 0:
											cycle=2;
											doublecycle_counter=1;
											break;
											
										case 1:
											cycle=4-2*free_sr;
											doublecycle_counter=0;
											break;
									}
								}
								else{
									cycle=4-2*free_sr;
								}
								
								//cout << "FB (crit): "<< dmg << "\n";
								//	////cout << "fb (crit): " << dmg <<", Total damage: " << damage <<"\n";
								
							}
							else if(att<(fb.crit+fb.hit)){
								//cout << "Time: " << timer/100.0 << ", Energy: " << en <<"\n";
								fb.adamage=(1-red)*((0.07*ap*(1+sr*src))*cp+(ap*(1+sr*src)*0.002439)*(en-fb.energy+(int)(lag/10.0)))*(1+naturalist_m)*(1+ferocity_m);
								fb.bdamage=(1-red)*(190+290*cp+(9.4)*(en-fb.energy+(int)(lag/10.0)))*(1+naturalist_m)*(1+ferocity_m);
								dmg=(1+RA*0.02)*(fb.adamage+fb.bdamage);
								damage+=dmg;
								ydamage+=dmg;						
								fb_damage_tot+=dmg;
								fbnum++;
								avgfbenergy+=en+(int)(lag/10.0);
								fb_damage+=(1+RA*0.02)*fb.adamage*(sr*src);
								en=-(int)(lag/10.0);
								cp=0;
								gcd=100+lag;
								if(doublecycle==1){
									switch (doublecycle_counter){
										case 0:
											cycle=2;
											doublecycle_counter=1;
											break;
											
										case 1:
											cycle=4-2*free_sr;
											doublecycle_counter=0;
											break;
									}
								}
								else{
									cycle=4-2*free_sr;
								}
								
								
								//cout << "FB (normal): "<< dmg << "\n";
								//	////cout << "fb (normal): " << dmg <<", Total damage: " << damage <<"\n";
							}
							else{
								//cout << "Time: " << timer/100.0 << ", Energy: " << en <<"\n";
								en=(int)((precision_m2)*en);
								cp=0;
								gcd=100+lag;				
								//cout << "FB (miss)\n";
								//	////cout << "fb (miss)" << ", Total damage: " << damage <<"\n";
							}
							
							goto fine;
							
						}
						
						//rip?
						
						if(((rip_expire==0)&&(((en>=rip.energy)&&(cp>=rip_t)&&(fb_t>5))||((en>=rip.energy)&&(cp>=rip_t)&&(en>(ripvsfb-(int)(lag/10)))&&(doublecycle_counter==1)&&(fb_t<=5))||((en>=(rip.energy+2*shred.energy-20-2*(int)(lag/10)))&&(doublecycle_counter==0)&&(cp>=rip_t)&&(fb_t<=5))||((en>=rip.energy)&&(cp>=rip_t)&&(fb_t<=5)&&(bdebuff==0))))||((rip_expire==1)&&(rip_timer<0)&&(((en>=rip.energy)&&(cp>=rip_t)&&(fb_t>5))||((en>=rip.energy)&&(cp>=rip_t)&&(en>(ripvsfb-(int)(lag/10)))&&(doublecycle_counter==1)&&(fb_t<=5))||((en>=(rip.energy+2*shred.energy-20-2*(int)(lag/10)))&&(doublecycle_counter==0)&&(cp>=rip_t)&&(fb_t<=5))||((en>=rip.energy)&&(cp>=rip_t)&&(fb_t<=5)&&(bdebuff==0))))){
							
							att=randnum();
							////cout << "Attack roll: " << att << "\n";
							
							if(att<(rip.crit+rip.hit)){
								////cout << "Time: " << timer/100.0 << ", Energy: " << en <<"\n";
								////cout << "RIP\n";
								ap_rip=ap*(1+sr*src);
								cp_rip=cp;
								en-=rip.energy;
								ripnum++;
								bdebuff=1;
								if(bdebuff_timer<((12+rip_g*rip_glyph+twoT7*3)*100)){
									bdebuff_timer=(12+rip_g*rip_glyph+twoT7*3)*100;
								}
								rip_timer=(12+rip_g*rip_glyph+twoT7*3)*100;
								gcd=100+lag;
								////cout << "\n" <<"Energy: " << en << ", Time: " << (1.0*timer)/100.0 << ", CP: " << cp <<"Step: " << doublecycle_counter <<"\n";
								//cout << "RIP" << "\n";
								cp=0;
								rem=0;
								if(doublecycle==1){
									switch (doublecycle_counter){
										case 0:
											cycle=2;
											doublecycle_counter=1;
											break;
											
										case 1:
											cycle=4-2*free_sr;
											doublecycle_counter=0;
											break;
									}
								}
								else{
									cycle=4-2*free_sr;
								}							}
							else{
								//cout << "Time: " << timer/100.0 << ", Energy: " << en <<"\n";
								en-=(int)((1-precision_m2)*rip.energy);
								
								gcd=100+lag;
								////cout << "\n" <<"Energy: " << en << ", Time: " << (1.0*timer)/100.0 << ", CP: " << cp <<"\n";
								//cout << "rip (miss)" << "\n";
							}
							
							goto fine;
							
						}

Beace · 10/09/08, 8:20 AM

Got a small question. It's my understanding that with the new glyphs, tiger's fury, t7 set bonus etc, we will easily have more combo points than we can use up if we only do SR and rip. So how exactly do the rotations that do not use bite look? Do you refresh an already ticking rip or what?

Nestario · 10/09/08, 10:26 AM

If you want to use Rip and SR as the only finishers you will end up using shred/rake/mangle with 5CP alot.

You will use a 2SR/5Rip rotation but you will waste a lot of CP since your minimum cycle-time is capped by the duration of rip.

Malazaar · 10/09/08, 10:34 AM

@nightcrowler: while this will work in a simulation, you can't always reapply rip in live situations before it has expired. If the last rip was done with higher AP than you currently have (due to procs, tinkets, drums, etc.) you will get a 'more powerful spell already active' message.

If they changed that already on beta, ignore my post.

Also, the break even point for FB and Rip is highly dependant on the target's armor. With only 20 % damage reduction, FB comes a lot closer to Rip. So if there is a lot of variance in mob armor (like in TBC) your best cycle might actually depend on the encounter itself.

smurph98gt · 10/09/08, 10:42 AM

Originally Posted by tlbj6142

I still do quite a bit of Cat DPS in our 10/25-man T4 runs. Which is what makes a hybrid spec at lvl 70 so difficult. At 80 it is not so tough, but at 70, I don't know what to give up to make a good build.

Aren't most DPS classes getting a boost to their DPS at lvl 70? If so, maybe I'll favor bear over cat in my hybrid build knowing someone else will pick-up my loss in DPS???

Talent Calculator - World of Warcraft is what I tried out last night on the PTR for a quick Kara run. The feel is similar to live except you get to mangle faster and swipe a million mobs at a time. I stayed in Bear for the netherspite kill, and did 650 dps in my bear gear. In Cat form with the same gear I did 700 dps on Prince. In Cat gear and Cat form, I did 750 dps. I was only Mangling to 5 and the Ripping however since untalented shred takes 6 seconds to pop up, it's awful.

It's hard to feel effective in cat form using this build. Shred eats up so much energy, it really does suck. If I was going to be tanking all the time, I probably wouldn't take OoC, it barely popped up. Full on Cat or a Hybrid build, I'd probably pick up OoC as it's worth a few free Shreds a minute.

tlbj6142 · 10/09/08, 11:58 AM

Originally Posted by smurph98gt

Talent Calculator - World of Warcraft is what I tried out last night on the PTR for a quick Kara run.

When you tanked, you didn't have threat issues without RnT? Did IW work on the bosses?