Sorry if this has been covered and I've missed it. This is something I feel I should know but don't. Has there been much testing to see exactly how these "break early"? Does the type of damage matter, do Dots break it more or less often than direct damage, is it the number of attacks or the damage involved?

I'm curious about this too. I think it's just probably a random chance built into the ability, like the chance for an engineering item to backfire

but then again, +hit gear is supposed to dramatically reduce the chance for mind control to break early, so maybe it's some kind of resist check

They recently changed the fear mechanic for warlocks where DD is significantly more apt to break a fear, where dots are a little less apt.

I believe either roots and frost nova work the same way - a certain level of DD will break them. That being said, I've been rooted mid air and taken from 100-0 before, without being able to retaliate. Not fun.

Also, how does the rank of Frost Nova come into play?

I've always had a feeling that it's easier to break nova with damage from the mage who casted it, compared to damage from other people.

Ice Lance eats Frozen Effects when it crits, just so you know. ;D

From what I have seen over time that any damage from anyone has a base chance of breaking. I have stood there rooted a warrior and doted him and starfire and wrathed him for half his hp and it still doesn't break on the otherhand I have rooted a player and 1 wrath has broken it.

A reason why theres a lot of "Roots breaks all the time" threads on the druid forums is because it has a chance to break on its own damage aswell so this combined with your damage and dots makes and I assume a high chance to break means it lasts very short.

that and all crowd controls are limited to a maximum of 10 seconds in the arenas (although this mainly effects fear and polymorph), but aye i've seen entangling roots break early fairly often on my druid.

Roots make my warrior cry. >_<

Nothing is more annoying then getting rooted, and watch as 5 ranged classes nuke you down from 100% to 0% with impunity. (And yet still not break the root!)

Its just easy to assume that roots can and will break when you need them to stay most.

My hypothesis is that roots suffers from a double problem:

1. Periodic Resist Check
AND
2. Damage (On Tick) Check

without testing of my own, the best information I have for all of these effects is that there is a periodic resist check. in the case of entangeling roots I believe a resist check is either passed or failed, if passed the damage is applied and the root is 'refreshed', so that the roots own damage no longer causes it to break.

My own anectdotal evidence is that a root would last 1 starfire, 2 starfires max.. but I could wrath almost with impunity.

I wanted to just believe this was my perceptions playing tricks on me, so I got owned in the face many times trying to starfire rooted people before giving up.

as with frostbolt/fireball, I'm not sure if roots are more durable through wrath because it hits for less or because it's nature damage instead of arcane.

the next time I go rootbombing maybe I'll try rank1 starfire and see if it behaves differently.

May I humbly advance the theory that crowd control resist checks are not, in fact, binary? I have no research, no graphs, nothing, so feel free to throw this away quickly, but..

On a low level warlock, I switched from Eagle to Shadow gear, and I noticed two (and let me underline how ancedotal and poorly sampled this is) neat changes: A) I could "cook" (er, DoT) mobs longer while feared, and B) fear seemed to last "better" just in general.

So may I suggest that a control spell (frost nova, entangling roots, fear) will have a spellpower coefficent applied towards a hidden "this amount of damage is what I'll break under" value. Partial resists apply to this number as well, complicating sampling the value - if, with 0 spellpower, you land a 1000 damage break fear, but the mob resists partially resists, it'll break under 500 damage, for example.

But with 400 spellpower (and, let's presume, a 75% spellpower coefficent since I'm lazy), the damage break value becomes 1300.

Anyway, it's a thought. Feel free to tear me a new one for all the observed data that contradicts this theory. This would explain DDs breaking control faster than DoTs, and such. Another possibility is that the relative threatvalue of the damage done is factored, so frost with that -threat talent is going to require huge numbers to break roots.

Anyways stopped with 200 Roots. Std Dev off of 3 seconds intervals rose to ~0.334 though my latency rose to around 200 ms towards the end of the testing.

So approximately 60% of roots without any extra damage last full duration.

While the sample size is small, it seems to indicate an almost equal chance of breaking at any point. Does this trend continue as you collect more data?

Also to make this data a little more thorough, do you have any hit gear or spell damage equiped?

For the tests I was wearing no hit or +dmg gear. Afraid I don't even have Hit Gear on my druid. One fluctation though was how many towers were taken in EPL but the damage variation was from ticks of 30 to 32 so nothing too bad.

Awesome, look forward to the results of damage

The gathering of data is simple enough. I created an addon that times from application to removal of entangling roots, the largest issue is how long this how takes spent around 2 hours on it about before I got bored and went on to something for the meantime. Once I add damage tracking in will upload the addon though its nothing really complex.

PvPing and working on my other main addon drew my attention.

Anyways heres the addon I've developed tracking this information should also work with gathering data on Fear, Frost Nova, and Polymorph as well.

http://acm.jhu.edu/~cryect/cctimer.zip

The data gets stored in cctimer.lua in your savedvariables folder for the account you are using it on. Tracks data by Spell used, target name, target's level, duration lasted, then stores each type of damage you did with the time from CC applied to the damage taken with the spell name, damage done, and damage type. Make notice its setup so it only tracks the damage you do and not the damage other players do to help you try to reduce variability (otherwise players in the same area attacking mobs named the same would be counted also). No way to disable the addon ingame by itself so disable it at the addon screen or via one of the ingame addon enabler/disablers.

Resetting data you will have to do manually by /script CCTimerData={}

BTW my current hypothesis on how Entangling Roots works is as follow.

Every time damage is dealt a check is done that is potentially influenced by the amount of damage done and type of damage done.

This probability works out for the DOT tick from Entangling Roots to 5% to break every time the DOT ticks.

My proposed plan then for determining on other types is deal damage then record how many of each attack doesn't break and how many do break the roots. Any damage attacks done after 25 seconds will be ignored since begins to get in the questionable area did the damage break it or just running out at 27 seconds.

Did some playing around with the mod, works well and is simple to export data to excel.

Now I just need a good mob to collect sample data from. The optimal mob would be humanoid/beast so I could pretty much nova, sheep, nova, sheep ....
thus killing two birds with one stone.

I'm fairly certain there is a periodic resistance check and that player level vs. mob level is a factor so that would pretty much mean a level 60 mob is needed.

My question: Anyone know of a level 60 mob with no frost resistance?

Barring that any other ideas on collecting data that would be worth the effort?

Using the addon right now, I have some hit gear (jindo's staff and some random blue's), I'll see if I can find a difference between using the +hit gear and withouth it.
Hypothetical situation:
14:01:00 > mob is affected by roots
14:01:00 > exe launches uber 1k starfire crit
14:01:03 > mob dies
14:01:03 > roots fade from mob

Does this count as a 3 second root breaking from 1K arcane damage? Or does the mod ignores the last 3 seconds?

edit: My avatar isn't running anymore! This is a sad day, a very sad day!

In my experience with frost nova, the amount of damage matters. I definitely saw a pattern than my 200% frostbolt crits seemed far more likely to break novas than hits, back when I was frost spec.

As for the type of damage, I consider it a plain fact that fire damage breaks frost effects faster than other types. I cannot say whether frost has a lower chance of breaking though but I'm not aware of that. I think it's rather a product of most mages only comparing frost to fire in this context.

On Maexxna, you will typically have most mages gathered and ready to nova + AoE the spawns down. You should see a very clear pattern that nova tends to break when you blast wave, making it far more effective to use blast wave at the end of the AoE spam instead of right after nova, before using AEs. Blast wave does like 50% more damage of course but I'd estimate that nova is about as likely to break in one blast wave as in 3 AEs.