Visualising the optimal cat rotation ingame - Page 15

coldbear · 02/12/10, 1:11 PM

Really enjoyed a few iterations ago how you had TF and Berserk grey out if the energy conditions weren't right. On extremely busy fights like PP and Blood Queen I've been using Ovale to a great extent when I'm too busy to spare a glance at my energy or timer bars - especially since the energy bar for ferals in ICC gear tends to go up and down like a yo-yo very quickly compared to previous tiers.

Was there a reason why this functionality was removed?

Great work, keep it up. And good riddance on the NG recommendation - I doubt anyone found that useful.

tagosa · 02/12/10, 2:48 PM

Originally Posted by coldbear

Great work, keep it up. And good riddance on the NG recommendation - I doubt anyone found that useful.

I'm confused why you would think nobody would find that feature useful. It's an easy way to get an OoC proc, and is worth it if you're waiting on energy to regen. I don't know about you, but I enjoy the OoC procs from using it, and I think it's silly to just dismiss people using it because you personally didn't.

With that said, I've yet to try the latest script, and was wondering what type of gains people are having from using it over some of the other user supplied scripts? I realize alot of it can be attributed to rng, but just some general dps gains would be nice to compare them to each other.

Furion · 02/12/10, 3:18 PM

While it's true that NG may be useful under certain circumstances I'm not sure this belongs in the optimal rotation until I get to test this in a sim and see what it actually does to your dps.
One of the problems with it is that you "waste" a lot of gcds for a chance on OOC and during those gcds your autoattack might cause an OOC proc as as well which you can't use until the gcd is over. In a worst case scenario this might even cause you to loose OOC procs especially when you are slow with following the suggestions. So even a gcd when there is apparently nothing to do is not entirely free of opportunity cost. And unless this is a dps gain of more than say... 10 I'm probably just too lazy to do it and don't want it in my suggestions. Realistically I'm pretty sure adding NG to the standard rotation would cause the majority of users to actually loose dps as I estimate the gains to be minimal even in the best case scenario (sim) which never applies to a human.
I need to make a decision in some cases. I could even add using haste pots out of combat to the suggestion list but I think this would be more annoying than helpful ingame.

Barbed · 02/12/10, 6:17 PM

the haste pot suggestion would be a cool idea

tagosa · 02/12/10, 6:43 PM

Originally Posted by Barbed

the haste pot suggestion would be a cool idea

I don't see how that would really be helpful. Sure if you were targetting the boss before the pull you would see that pot icon, but in reality anyone who pre-pots right before the pull now doesn't need a reminder to do so.

I can see your point about the NG Furion, and I've actually questioned it a few times whether the OoC proc I recieved was caused by the NG, or just pure coincidence. But I have a feeling from the number of times it has procced when using NG is just too large to be coincidental. I've mostly used it when I've emptied my energy bar (after a berserk, for instance), and have a a spare gcd while waiting for some regen.

ramenchef · 02/12/10, 9:57 PM

NG definitely procs OOC as I spent awhile just hitting NG without meleeing something and got multiple procs of OOC over that period of time.

Duilliath · 02/13/10, 5:30 AM

The issue isn't whether or not NG procs OOC. It's simply whether complicating the Feral cycle even further is worth the minimal DPS gain from at best one OOC proc per minute. I can well understand Furion in their comment that trying to follow this might actually cause a DPS loss for the slower folk.

The 10-20 DPS gain estimate feels fairly accurate to me as a rough figure, with one non-guaranteed OOC essentially being a free Shred (my average Shred on an ICC night being 11.5k) with 1.70ish combo point gain from each Shred.

WereTaco · 02/14/10, 4:47 PM

I've added some code to support bear tanking. Just in case some of you cats tank too

not sure how to prioritize Lacerate over everything else when it's about to fall off tho.

#Project: Visualising the optimal cat rotation ingame
#Script Version: 2.8
Define(FFF 16857)
Define(FF 770)
Define(RIP 1079)
Define(MANGLE 33876)
Define(SHRED 5221)
Define(TIGER 5217)
Define(RAKE 59886)
Define(ROAR 52610)
Define(BITE 22568)
Define(BERSERK 50334)
Define(OMEN 16870)
Define(TRAUMA 46857)
Define(MANGLEB 33878)
Define(GRASP 16689)
Define(LUST 2825)
Define(LACERATE 33745)
Define(SWIPE 779)
SpellInfo(RIP resetcounter=ripshreds)
SpellInfo(SHRED inccounter=ripshreds)
AddCheckBox(Manglebot "Manglebot" default)

AddIcon help=main
{
if Stance(3) # cat
    {
    if ComboPoints(more 0) and BuffExpires(ROAR 0) Spell(ROAR priority=4)
    if ComboPoints(more 2) and BuffExpires(ROAR 8) and TargetDebuffPresent(RIP mine=1) and at most 3s from TargetDebuffExpires(Rip mine=1) until BuffExpires(ROAR) and TargetDeadIn(more 8) Spell(ROAR priority=4)
    if ComboPoints(more 3) and TargetDeadIn(less 2) Spell(BITE)
    if TargetDebuffPresent(RIP mine=1) and TargetDebuffExpires(RIP 4 mine=1) and Counter(ripshreds less 3) Spell(SHRED)
    if ComboPoints(more 4)
        {
        if TargetDeadIn(less 6) Spell(BITE)
        unless TargetDebuffPresent(RIP mine=1) Spell(RIP priority=4)
        unless TargetDebuffExpires(RIP 8 mine=1) or BuffExpires(ROAR 11) Spell(BITE)
        }
    if CheckBoxOn(Manglebot) and TargetDebuffExpires (MANGLE 1) Spell(MANGLE)
    unless {TargetDebuffPresent(MANGLE) or TargetDebuffPresent(MANGLEB) or TargetDebuffPresent(TRAUMA)} Spell(MANGLE)
    unless TargetDebuffPresent(RAKE mine=1) or TargetDeadIn(less 10) Spell(RAKE)
    if TargetDeadIn(less 10) Spell(SHRED)
    }
if Stance(1) # bear
    {
        Spell(FFF)
    Spell(MANGLEB)
    if TargetDebuffExpires(LACERATE 4 stacks=5) Spell(LACERATE) 
    Spell (SWIPE)
    }

}

AddIcon help=main size=small nocd=3
{
if Stance(3) # cat
    {
    if {TargetDebuffPresent(MANGLE 0.6) or TargetDebuffPresent(MANGLEB 0.6) or TargetDebuffPresent(TRAUMA 0.6)} and TargetDebuffPresent(RAKE 0.6 mine=1) and {Mana(more 79) or BuffPresent(OMEN) or {TargetDebuffExpires(RIP 3 mine=1) and ComboPoints(less 5)} or BuffPresent(BERSERK) or 2s before Spell(TIGER) or {ComboPoints(less 1) and BuffExpires(ROAR 2)}} Spell(SHRED)
    } 
    unless {BuffPresent(BERSERK) or 4s before Spell(TIGER) or Mana(more 32)}
        {
        unless {TargetDebuffPresent(FFF) or TargetDebuffPresent(FF)} Spell(FFF priority=2)
        }
}


AddIcon help=cd size=small nocd=3
{
if Stance(3) # cat
    {
    Spell(TIGER priority=4)
    }
}

AddIcon help=cd size=small nocd=5
{
if Stance(3) # cat
    {
    Spell(BERSERK)
    }
}

The 2.8 Version of the script works pretty good, ty for this.

Furion · 02/18/10, 9:44 AM

I will probably quit wow and working on this script, but I thought if anyone wants to pick up this project it might be a good idea to add my current thoughts on the script.
I think the way shred is displayed can be optimized further. My idea is that the filler shreds are treated like reactionary abilities (because the high priority moves of cats actually form a rather steady rotation with the odd FB from time to time).

Playing a soundfile for example (in addition or as replacement for the 2nd icon) would be an idea but it thould only play (and show when there really is nothing to do in the main rotation so when the sound play you want to be hammering your shred key every time without having to think about it. If the 2nd icon with shred is only shown when there is nothing to do in the main icon it should probably have its size increased, too.

Also once simulators can handle natures grasp coldbear had the idea that the best time to use it would typically be after using FB and I think this might be worth testing.

Umah · 02/18/10, 10:15 AM

If you want to do us one last favour, please comment the more complex lines of your code, maybe split the conditions into multiple lines and comment each of those.

p.s. I like Coldbears idea, will try it next time.

Furion · 02/18/10, 11:42 AM

Not sure how I should do that... but feel free to ask specific questions about lines you don't understand and I'll answer them as good as I can but remember that you can look up pretty much everything in the ovale documentation

Umah · 02/18/10, 12:44 PM

Originally Posted by Furion

Not sure how I should do that... but feel free to ask specific questions about lines you don't understand and I'll answer them as good as I can but remember that you can look up pretty much everything in the ovale documentation

Questions as comments # foo - always above the line in question

AddIcon help=main
{
if Stance(3) # cat
	{

	# makes sense

	if ComboPoints(more 0) and BuffExpires(ROAR 0) Spell(ROAR priority=4) 

	# Why the rip conditions here "and TargetDebuffPresent(RIP mine=1) and at most 3s from TargetDebuffExpires(Rip mine=1)"

	if ComboPoints(more 2) and BuffExpires(ROAR 8) and TargetDebuffPresent(RIP mine=1) and at most 3s from TargetDebuffExpires(Rip mine=1) until BuffExpires(ROAR) and TargetDeadIn(more 8) Spell(ROAR priority=4)

	# makes sense, though: does it do more damage than shred with 4 CP?

	if ComboPoints(more 3) and TargetDeadIn(less 2) Spell(BITE)
	
	# makes sense (prolongs rip by the shred glyph)

	if TargetDebuffPresent(RIP mine=1) and TargetDebuffExpires(RIP 4 mine=1) and Counter(ripshreds less 3) Spell(SHRED)
	
	if ComboPoints(more 4)
		{

		# makes sense but isnt it a duplication of the rule above "if ComboPoints(more 3) and TargetDeadIn(less 2) Spell(BITE)"

		if TargetDeadIn(less 6) Spell(BITE)

		# two lines below make sense

		unless TargetDebuffPresent(RIP mine=1) Spell(RIP priority=4)
		unless TargetDebuffExpires(RIP 8 mine=1) or BuffExpires(ROAR 11) Spell(BITE)
		}

	# makes sense, maybe RAKE should be higher in priority? Furthermore I would reduce less 10 to less 9 as the time to die is en extrapolation (it might go up again)? good/bad idea?

	if CheckBoxOn(Manglebot) and TargetDebuffExpires (MANGLE 1) Spell(MANGLE)
	unless {TargetDebuffPresent(MANGLE) or TargetDebuffPresent(MANGLEB) or TargetDebuffPresent(TRAUMA)} Spell(MANGLE)
	unless TargetDebuffPresent(RAKE mine=1) or TargetDeadIn(less 10) Spell(RAKE)
	if TargetDeadIn(less 10) Spell(SHRED)
	}
}

AddIcon help=main size=small nocd=3
{
if Stance(3) # cat
	{
		{
			# why mana more 79. Why pool energy for 79?

			# why exactly 2 sec before tigers fury is ready?

			if {TargetDebuffPresent(MANGLE 0.6) or TargetDebuffPresent(MANGLEB 0.6) or TargetDebuffPresent(TRAUMA 0.6)} and TargetDebuffPresent(RAKE 0.6 mine=1) and {Mana(more 79) or BuffPresent(OMEN) or {TargetDebuffExpires(RIP 3 mine=1) and ComboPoints(less 5)} or BuffPresent(BERSERK) or 2s before Spell(TIGER) or {ComboPoints(less 1) and BuffExpires(ROAR 2)}} Spell(SHRED)
		}

		# are you sure that it is a personal dps increase to have FFF that far below in priority?

		unless {BuffPresent(BERSERK) or 4s before Spell(TIGER) or Mana(more 32)}
		{
			unless {TargetDebuffPresent(FFF) or TargetDebuffPresent(FF)} Spell(FFF priority=2)
		}
	}
}

Furion · 02/18/10, 1:39 PM

Most of these questions aren't really about understanding the script... it's you questioning the theorycrafting behind it. The answer to most of those questions is simply: Because it was the best dps in simcraft so refer to the best possible dps in simcraft thread for that. My view is: As long as I can back this up with simcraft you have the burden of proof.

if ComboPoints(more 2) and BuffExpires(ROAR 8) and TargetDebuffPresent(RIP mine=1) and at most 3s from TargetDebuffExpires(Rip mine=1) until BuffExpires(ROAR) and TargetDeadIn(more 8) Spell(ROAR priority=4)

I can understand if this line is giving people headache but again this isn't rocketscience its just what worked best in simcraft so far. Feel free to come up with something better. This line refreshes SR early to desynchronise rip and roar if they expire nearly at the same time. It doesn't make any sense to refresh savage roar early if you have no rip on the target as there is nothing to desynchronise!

FFF priority isnt tested (although I'm quite confident about it) but for most good players irrelevant as they have a moonkin.

One last thing about "if Mana(more 79)": I can understand people arguing that pooling to 80 energy is too high for actual human players so feel free to lower that threshhold to a point you feel comfortable working with. I don't have any means to find out what works best for the majority of people using this script as it depends on skill (=how fast you hammer your buttons), latency etc so some choices I had to make are irrational by necessity. So don't treat any of the lines in this script as gospel, but be aware that none should be totally crazy, either since I really put a lot of effort into this script and testing it with simcraft and ingame which makes me confident to the point where I believe that there won't be any huge discoveries until cataclysm. But who knows.

sakkie69 · 02/20/10, 4:15 AM

Originally Posted by Furion

One last thing about "if Mana(more 79)": I can understand people arguing that pooling to 80 energy is too high for actual human players so feel free to lower that threshhold to a point you feel comfortable working with. I don't have any means to find out what works best for the majority of people using this script as it depends on skill (=how fast you hammer your buttons), latency etc .

I was actually going to ask about that - is it possible to make some adjustments based on latency as well? Some of us do not have the luxury of < 100ms ping, but have to play with 200-500. So it would be nice of the timers could be offset a little bit. Is that possible, or is it something we can set in Ovale?

Umah · 02/20/10, 7:19 AM

I wondered about target distance. If target distance > 10 -> fff. =)