Guild Wars 2

Please don’t post pure speculation in this thread, I’ve found enough conflicting accounts already. Instead, I ask my fellow players to go out and do some actual testing, as well as report findings from previous tests.

If anyone has already worked out all the details please point me at their contribution, because what I’ve found so far is rather confusing.

The problem I noticed is with the staff autoattack: burns from Winds of Chaos show inconsistent behavior if one or more bleeds from previous attacks are still ticking.

First of all the testing conditions I used: The basic layout consists of a PvP Staff of Smoldering (+10% burn duration); basic PvP gear with runes of Flame Legion, Balthasar and Baelfire (2 each); and 30 points in Domination, without actually picking any traits. The gear was removed for low-duration tests (i.e. attacking naked instead of overwriting the runes), the staff was kept at all times.

For testing I picked an uncontested (usually Heavy) golem in the dps test corner and manually cast #1 with the autoattack feature disabled. I then counted the damage floaters above the golem as the number of times a condition “ticked” for damage.

As long as only 1 strike is made the “standard lore” holds: bleeding and burning tick every second, and only full additional seconds cause additional ticks. Assuming nothing is bugged my full test spec adds up to +85% burn duration, and so far no burn ticked twice. My sample size is too small to claim this is always the case (less than an hour of testing), but at least nothing seemed out of place.

Now here’s the odd part: If I land a burn proc while the target is still bleeding from a previous attack the burn almost always ticks twice. In fact, out of about 20 times I obtained this constellation every single one had the burn tick twice! Attacking naked (should be +40% burn duration) didn’t have constant double-ticks, but after 5 or so attacks I got one. I gave up after that as taking a sufficiently large sample size to use statistic evaluation is extremely boring and the core method has too many additional influences (for example, I got several burn procs without having lingering bleeds on the target, these seemed to tick only once but might have “flushed out” remaining burn durations <1s)

Notably Vulnerability doesn’t cause the odd double-tick. I didn’t test other conditions as I have no reliable way of applying them (maybe Sigil of Doom could be used for Poison)

Does anyone have additional/conflicting data?

As far as I know, that’s the normal behaviour of all conditions that tick their damage per second; the most prominent one probably being fear.

Such conditions always have their dmg-calculation at the same time. That means a burn of 1.8 seconds will only tick once if it’s the only (damaging? according to your tests) condition on the target. But if there’s already a bleed and you add this burn between 0.2 and 0.9 seconds after the bleed ticked, you will get two damage-ticks from it (since it lasts long enough to be present during two damage-calculations). Of course you could still fail to get the second tick with a 1.9s burn, but the longer it lasts, the higher the chance that you’ll get a second tick out of it, since the time frame to apply it becomes larger.

The explanation for this is that all conditions tick on the same instant, regardless of when they were applied. So if you have a burn with 1.5 second duration, and you just apply it by itself, it’ll only tick once. However, say you have a bleed already on which is about to tick in 0.2 seconds and you apply the same burn, it’ll tick once (after 0.2 seconds), and then again (after 1.2 seconds).

This works out so that if you assume condition application time is independent of your previous stacks, “on average” a 1.5 duration condition will tick once half the time, and twice half the time. Not really sure how viable that assumption is, and also the second tick will never happen if it’s the only condition you’re applying, but still additional condition duration isn’t a complete waste even if you don’t cross the next second threshold.

I presume iElasticity was not used? Also, what are you trying to gain from this investigation? Overall, burn is a rather low priority condi for us. If it is indeed a defect, (omg… speculation) then possibly it was caused by code churn from the recent bounce priority changes.

Helios.3598:

I presume iElasticity was not used? Also, what are you trying to gain from this investigation? Overall, burn is a rather low priority condi for us. If it is indeed a defect, (omg… speculation) then possibly it was caused by code churn from the recent bounce priority changes.

No other traits, no bounces back to the original target at all.

I was trying to figure out how condition duration works at all, when I tried to read up on the subject a month ago I found unbacked claims of anything from “0.5s duration is always rounded up” to “any fraction of a second is completely wasted”. Some people also claimed bonus durations are rounded to the nearest quarter of a second, which is only true for tooltips.

Thanks alot for your tests, photoloss. This at least answers the basic concept of ticking conditions. Out of the two (three) theories, your short tests “proofed” that the second is implemented. The floating numbers of bleeds are just there to confuse us.

Theory:
Once a ticking-based condition (boon?) is applied, a general tick counter starts on the respective entity (player, npc, mob). Every second (maybe half-second for torment, but I think it’s still second) that tick counter checks for active conditions and the applies their tick-based effect.

Result:
Assuming you continuously attack (randomized tick start), any fraction is turned into a random tick with the probability of the fraction. 4.2s burn turns into 4.2 burn ticks (4 garantueed + 1 tick at 20% probability). So any fractions are viable.

Result2:
Assuming you test regeneration in a laboratory (no real environment), all fractions will be cut, because there’s no other tick-based boon. So testing regeneration needs to be done with conditions (assuming they share the counter).

Unsolved questions:

• What happens if you duration-stack fractions?
• Do different sources get different tick counters on the same entity?
• What happens when two different sources duration-stack fractions?
• Do boons and conditions share the same tick counter?

You guys should check something: If anyone of you got a necro, I beleve he has a Skill that applies just tons of bleeds. If you use this on an enemy, does the life bar vanish smooth or little by little? If first one is the case, then the tics are just a visual effect and the damage goes properly. If it goes (lack of english but i hope you know what i mean >.<) little by little (?) piece by piece (?) then every tic does damage and it’s not just visual.

Xyonon.3987:

You guys should check something: If anyone of you got a necro, I beleve he has a Skill that applies just tons of bleeds. If you use this on an enemy, does the life bar vanish smooth or little by little? If first one is the case, then the tics are just a visual effect and the damage goes properly. If it goes (lack of english but i hope you know what i mean >.<) little by little (?) piece by piece (?) then every tic does damage and it’s not just visual.

Well, the visual things can be bypassed. Like the bleeding floaters when you stack 10+ bleedings. You don’t see them tick once a second but all the time. The client could smooth the life bar.

It could proof a tick-based system, but not proof the opposite. (proof of the devil)

Well if you do an auto attack it doesnt go smooth aswell, does it? Also maybe its better to take ele with burn only

The other day I was on the pvptv stream and they were talking about this subject at length, but focused on burning. They were saying the likelihood of an addition burning tick is dependent on when the first tick occurs. Some skills apply conditions instantly upon cast (such as fear), but most do not (skill execution > successful hit > actual burning occurrence). This may factor into the equation but I was unable to figure out exactly how. The discussion did not include anything regarding other damaging conditions, purely burning and it’s duration.

Also, awhile back a mesmer posted a non-confusion based condition build with a heavy focus on extending burn duration (I believe the person’s name was Palu). They did some heavy testing on this subject and there is some great information in those threads if you want to dig for them.