Excuse me in advance if this is the wrong way to ask it, but since I haven't found anything in the forum archives and have never heard anything about it, I'll just ask here:
Does anyone know at what point in time a sheep's duration is determined?
I see two possibilites:
1) Sheep duration is determined to be 1..50 seconds when casting the polymorph spell. If this were so, it would be helpful to resheep every few seconds. For example, if WoW's algorithm decides the next sheep should last only 15 seconds, you could prevent the sheep from breaking early by resheeping every 10s. The sheep would then only ever break early if WoW decides it should last less than 10 seconds, reducing the overall number of times a sheep breaks.
2) WoW decides every few seconds whether the sheep is going to break now or not. In this case, resheeping often and early would be useless, since the chance for the sheep to break in the next few seconds would be unchanged.
The same applies to other forms of CC that can break early, such as shackle undead, sap, banish, so maybe this is not really a mage question.
One fight this question would be of interest is The Lurker Below - should I sheep the naga often or only when the polymorph is about to run out anyway?
You could probably find out about this empirically - polymorph something every 10seconds for a few hours, and then polymorph it every 50 seconds for a few hours and count the number of times the sheep breaks, but that sounds like a whole lot of boring work.
As far as I know is WoW rolling in the background every few seconds if the sheep brakes or not. You can affect this with your +hit-gear, that means the more +hit you have the more the CC lasts its duration. Imho it is the same procedure as with damagespells. That means with no +hit on your gear or +hit talents the sheep will have 17% chance to brake every second.
2) WoW decides every few seconds whether the sheep is going to break now or not. In this case, resheeping often and early would be useless, since the chance for the sheep to break in the next few seconds would be unchanged.
This is called the "tick resistance check". Each tick, 2 seconds, the resistance of the mob to the polymorph is checked for both level-based resistance and school resistance, in this case arcane.
The mob has a minimum of 1% to resist the polymorph, and this is checked each tick. Thus, for a 50 second polymorph, the mob has a 25% chance to not make the entire duration.
This is called the "tick resistance check". Each tick, 2 seconds, the resistance of the mob to the polymorph is checked for both level-based resistance and school resistance, in this case arcane.
The mob has a minimum of 1% to resist the polymorph, and this is checked each tick. Thus, for a 50 second polymorph, the mob has a 25% chance to not make the entire duration.
Hmm, 25% chance to not to remain polymorphed the entire duration? I doesn't really seem like that ingame. Typically all polymorphs last full duration and very rarely they break early.
If there's one piece of honor or arena gear you're in a position to get, i'd recommend Gladiator's War Staff or Merciless Gladiator's War Staff . If of course you can get to a 1850 PR, the S3 dagger would probably be your best bet.
All this time I had thought that the duration of polymorph was reduced by the usual resistance percentages - 25, 50, 75, and heartbeat (if it broke on its own early at all). I've never timed early breaks, I guess it was just a placebo effect of me feeling like it was a heartbeat resist (~2 seconds), or that it self-broke consistently at around 13 seconds, 25 seconds, or 38 seconds. Was this ignorance or has the tick-resist also not been tested?
As far as I know is WoW rolling in the background every few seconds if the sheep brakes or not. You can affect this with your +hit-gear, that means the more +hit you have the more the CC lasts its duration. Imho it is the same procedure as with damagespells. That means with no +hit on your gear or +hit talents the sheep will have 17% chance to brake every second.
Athros
That is utter hogwash. Checking a 17% per second would mean you have a 0.83^50 chance to see a polymorph last it's full duration. In case you're not a fan of calculators, that's about 0.08993% it'll last 50sec. It also means there's a merely 15.5% it'll last 10sec to boot.
While I have no evidence to back I'm tempted to believe poly rolls a possible partial resist when it's cast. Crude observation indicates that the vast majority of Poly lasts it's full length, rarely it is resisted on cast and equally rarely (though perhaps slightly more frequent) it seems to break early. As far as I can determine, the spread of the breaking is rather linear; that means, it's not more likely to break later than it is to break earlier.
On a hunch here, I'll go out on a limb and even say that it's on the same mechanic as Partials, meaning it's either going to break at 25%, 50% or 75% duration. The reason I say this is I've never had trash break poly before 25% of the duration is over (ignoring mobs that got dispelled of course).
I re-poly a whisker before half-duration thus covering the (theoretical) possibility of 2/3 of partial resists.
Edit: Do not sign your posts please. This is EJ, the Forum Rules are at the top and they are there so you read them.
Originally Posted by Doroteasenjk
This is called the "tick resistance check". Each tick, 2 seconds, the resistance of the mob to the polymorph is checked for both level-based resistance and school resistance, in this case arcane.
The mob has a minimum of 1% to resist the polymorph, and this is checked each tick. Thus, for a 50 second polymorph, the mob has a 25% chance to not make the entire duration.
You are showing a bad understanding of statistics: A 1% to break poly each tic for 50sec does not result in 25*1%=25%.
Under your assumption, which I am willing to accept for PvP context but I believe to be grossely off for PvE the chance a mob lasts the whole polymorph is 0.99^25 which is in fact 77.78%.
The reason I say I think you're off is because with a lesser hit rate (let's say, 3% miss) we result in 46.6% the mob will last it's entire duration. Somehow the possibility that with less than 16% hit rate one needs to chain-poly to maintain his CC seems rather far-fethced.
While I have no evidence to back I'm tempted to believe poly rolls a possible partial resist when it's cast. Crude observation indicates that the vast majority of Poly lasts it's full length, rarely it is resisted on cast and equally rarely (though perhaps slightly more frequent) it seems to break early. As far as I can determine, the spread of the breaking is rather linear; that means, it's not more likely to break later than it is to break earlier.
On a hunch here, I'll go out on a limb and even say that it's on the same mechanic as Partials, meaning it's either going to break at 25%, 50% or 75% duration. The reason I say this is I've never had trash break poly before 25% of the duration is over (ignoring mobs that got dispelled of course).
That seems about right, but obviously all it would take to disprove that hunch and lend credence to a "tick" or "heartbeat" based system is one verified sheep breaking at a point other than 25%, 50%, or 75%.
And there's still the question of why it seems 71 and 72 level mobs seem to break sheep at a higher rate than the level based partial resist can account for. Maybe its just nerves from the importance of keeping an eye on your sheeped naga or what have you, but it does seem that there are more poly casts that eventually break on thier own than there are partial resists for any given raid elite i polymorph.
Um... what level 73 mobs do you poly???
I think he was just being glib, obviously athros's math is way off and wouldn't hold up for 71 or 72 mobs either.
Heartbeat resist was implemented at the same time Diminishing Returns was, to prevent chain-CC in PvP both in BG and in world-PvP. While I can't recall the notion that it was only active on players, I'm tempted to think so. Either way, Heartbeat resists seem to check only for a very short length of time after initial application: You may get one 0.5-1.5sec after a cast, but not later.
While I don't recall one on raids, that might be because sub-consciously I classify a heart-beat under the same banner as "noob melee tabbed wrong mob" or "noob tank Thunderclapped". I can with plenty of certainty however, state that I don't remember any RE-poly ever resulting in a heartbeat.
I was thinking about an alternative to 2/48/11 namely 5/45/11 like Talent Calculator - World of Warcraft
I doubt that the lost points in Incineration would matter much, but if there is sheeping/spellstealing to be done
the points in arc focus might be worth taking...
Pretty sure the tick based resistance checks are how the system works for polymorph. I think its the same system they incorporated for DoT spells actually. Where the spell does a binary check on application (all or nothing) and then does a resistance check for each individual tick. In the case of sheep the secondary check would also be binary since the target is either sheeped or it is not sheeped.
Guess the easiest way to check would be to take off all hit gear (and have no talents specced in Arcane Focus) and find a level 72 mob to polymorph. Keep it polymorphed and see how often the polymorph is resisted and how often it breaks early and the duration if it does break early.
Edited to respond to the post above:
Originally Posted by Baruk
I was thinking about an alternative to 2/48/11 namely 5/45/11 like Talent Calculator - World of Warcraft
I doubt that the lost points in Incineration would matter much, but if there is sheeping/spellstealing to be done
the points in arc focus might be worth taking...
Note the glibbness of the response above you. There aren't any level 73 mobs to polymorph currently which means there is no need to keep 16% hit in arcane. If you are hit capped for Fire/Frost for a boss then you are capped for hit for polying trash.
As for Spellstealing - it's important for Krosh but once you leave Gruul's lair the other "Spellsteal" encounter is Council and since Dampen lasts 2 minutes and is reapplyed by the target during that duration there are no real issues with having a spellsteal be resisted.
I'm about 60% sure that resist checks on CC are done at the one-eighth mark, the one-quarter mark, and the halfway mark. In other words, on a 60 second crowd control spell there would be checks at 7.5 seconds, 15 seconds, and 30 seconds. If it goes past halfway, the crowd control will last the complete duration.
I know that there was a post about this in the EJ forums somewhere, but I can't dredge it up with search. I do know that I've never seen a sheep break late, it was always early- then again, sheeps don't always go the distance for the full CC duration.
As for Spellstealing - it's important for Krosh but once you leave Gruul's lair the other "Spellsteal" encounter is Council and since Dampen lasts 2 minutes and is reapplyed by the target during that duration there are no real issues with having a spellsteal be resisted.
Hm ok, I'm not at Council yet. If the duration of the buff is longer than the interval after which it is
reapplied by the mob one is probably better off with Incineration, even tough I still think that Inc.
is only a extremly marginal buff of single target dps. But alas there are no real alternatives :/
Keep in mind that WWS under reports the dps of Fire due to the Fireball dot.
Is the amount or percentage (or whatever) of this under-reporting quantified anywhere? In other words, is there a way to estimate the "real" output of a Fire Mage by adjusting for this factor on or through WWS?
Apologies for asking this again if it's been answered before; I searched all through the forums here and either it hasn't been mentioned or my ability to select the correct terms to find it is lacking.
Is the amount or percentage (or whatever) of this under-reporting quantified anywhere? In other words, is there a way to estimate the "real" output of a Fire Mage by adjusting for this factor on or through WWS?
No, this depends on the encounter. For example, if you had an encounter where you could only cast fireball that hits for 2500 non-crit and then had 8 seconds of movement, your DPS calculation would look like this:
2500 dmg in 3 seconds + 150 dmg in 8 seconds = 2650 dmg in 11 seconds dps time = 240 dps.
If you fired 3 arcane missiles that each hit for 833 instead of one fireball, you'd have:
2500 dmg in 3 seconds + 8 seconds of non-DPS-time = 2500 dmg in 3 seconds dps time = 833 dps.
As for Spellstealing - it's important for Krosh but once you leave Gruul's lair the other "Spellsteal" encounter is Council and since Dampen lasts 2 minutes and is reapplyed by the target during that duration there are no real issues with having a spellsteal be resisted.
I have to admit I have never tanked council myself, but we have seen a number of wipes because the first spellsteal was resisted. On Krosh firehand you can survive one 9,5 k fireball, but one of zerevors 15 k arcane bolts is usually game over.
I don't believe it was mentioned so far, but there is also the issue of Malande resisting counterspells which will cause me to put at least 2 points in arcane focus
well you wont wipe to missing a heal on malande, merely prolong the fight.
<Eej> YOU"RE GONNA PULL
<Eej> IF YOU SQUEEZE OFF ANOTHER ARCANE BLAST
<Spectear> You've obviously never played with Manly.
<Spectear> That's hardly a reason to stop DPS. Very Manly Staff
Unless it's been altered substantially, the check on polymorph includes +hit in some way. Way back pre-1.11 when Arcane Focus could eliminate the 1% minimum resist chance, having enough points in it to pass the hit cap on the 61/62 mobs that you'd be sheeping ensured that polymorphs always lasted the full duration.
I don't have the permissions to follow that link, but I was able to google archive it... and get two people contradicting each other. That is where I remember hearing the 1/8, 1/4, and 1/2 resist checks. It should be easy enough to test with a brand new mage getting Polymorph at level 8. I'll see if I can rope someone into tanking that tonight, if no one does it over today.
OK, so that's 3 theories being posited now on how polymorph resists work:
1) a binary resist check on each tick
2) the partial resist on cast reducing duration by 25%, 50%, 75%, or 100% (single-tick) (12.5, 25, 37.5 seconds)
3) binary resist check at 1/8, 1/4, and 1/2 duration (6.25, 12.5, and 25 seconds (corrected from erroneous 60-second duration)
I always thought it was #2.
Well that was a very boring 10 minutes but somewhat productive.
Went up to netherstorm and rank 1 polyed a level 70 etherial over and over and over again. First resist was 15-16 seconds in so thought maybe it could be the partial resist theory. Then had a break at 12 seconds in and the final one (before I was just too bored to keep polying the thing - really need to do this on a level 72+ so there is a decent chance of failure) occured with 2 seconds to go on the polymorph.
Not a very scientific study of the effect but it probably is the binary resist check every 1-2 seconds.
No, this depends on the encounter. For example, if you had an encounter where you could only cast fireball that hits for 2500 non-crit and then had 8 seconds of movement, your DPS calculation would look like this:
2500 dmg in 3 seconds + 150 dmg in 8 seconds = 2650 dmg in 11 seconds dps time = 240 dps.
If you fired 3 arcane missiles that each hit for 833 instead of one fireball, you'd have:
2500 dmg in 3 seconds + 8 seconds of non-DPS-time = 2500 dmg in 3 seconds dps time = 833 dps.
In fact, WWS ends "dps mode" after 5 seconds where no damage is dealt.
Therefore, in such a scenario, you'd get 2650 dmg in 16 seconds for Fire (165,6 dps) and 2500 dmg in 8 seconds for Arcane (312,5 dps).
12/12/07, 10:02 AM
Similar Threads
