virtus753 wrote: »

PudraSekeri wrote: »

I guess it acts like a damage over time ability, but deals direct damage over 1.2 seconds.

It is multiple instances of direct damage done over the course of one cast. In that respect it is the same as Jabs. Jabs is channeled, while Bound Armaments is not, but that shouldn't be (and doesn't appear to be) relevant for status effect purposes.

If each hit of Bound Armaments had a 10% chance to apply its status effect, by the end of the cast (assuming all four hits reached the target), you'd have only a 65.61% chance of not proccing that status effect at all. That means a 34.39% chance of proccing it at least once: more specifically, you'd have a 29.16% chance of proccing it exactly once, a 4.86% chance of exactly twice, a 0.36% chance of three times, and 0.01% chance of all four. That would result in the status effect being applied on average 40 times per 100 casts instead of 10, as suggested by the 10% chance being rolled four times per cast.

So you'd have a 3.439x greater chance of applying the status effect at least once per cast compared to one-hit-per-cast abilities, but a 4x greater rate of proccing the status effect in total.

For greater parity with one-hit-per-cast abilities, there are two options:

1) We could aim for an average of 10 procs per 100 casts: just cut the 10% chance per hit to 2.5% per hit.

This would result in a 90.34% chance not to get a proc at all in a given cast, meaning a 9.63% chance to proc it at least once. The expected average would still be 10 procs per 100 casts because that 9.63% comprises a 9.27% chance of proccing it exactly once, a 0.36% chance of proccing it exactly twice, a 0.0061% chance of proccing it exactly three times, and a 0.000039% chance of proccing it on all four hits.

The downside to this is a slightly lower chance to get the status effect proc at least once per cast. The upside is it could hit multiple times, since status effects have no cooldown. Since Sundered has two 4-second long effects (the weapon/spell damage boost and the Minor Breach), in addition to its direct damage, it would be in the sorc's interest to have a higher chance of getting the effect at least once per cast, as having Sundered go off multiple times within a GCD (all four hits land within 1 second*) wastes a great deal of the buff and debuff.

2) So alternatively we could aim for an overall chance of 10% to proc the status effect at least once per cast. This would result in a slightly higher average of procs per cast, since multiple procs could and would occur, albeit relatively rarely. But it would mean a more equal chance of getting the proc at all in a given cast.

To do that, find the chance per hit such that (1 - CPH) to the fourth power is 0.9 (90%). Chance per hit would be 0.025996 (2.6%). That would result in an expected average of 10.4 procs per 100 casts, which may be an acceptable trade-off to achieve the same 10% chance of getting the status effect at least once per cast.

It may be impractical to figure out whether ZOS used 2.5% or 2.6% (or some other number) from testing. But either one of those would arguably represent greater parity with once-per-hit abilities, whereas using a 10% per hit chance would result in a 3.44x greater chance of applying the status effect at least once per cast and a 4x greater average of procs per cast in total.

*Bound Armaments should normally be doing all of its damage within 1 second. The difference between the first hit and the fourth is nominally only 0.9 seconds: if the first hit lands at 0.0s, then the second lands at 0.3s, the third at 0.6s, and the fourth at 0.9s.

Here is an example from live:



It did take about 0.7s for the initial projectile to land (the cast activated at 2.488s), but the four hits all came within less than 0.9s of each other (3.183, 3.374, 3.741, 4.029). Since status effects proc upon the hit landing, not upon cast, that's four chances to proc the related status effect within 0.9s. Since its damage fits within a single GCD, I don't see a need to adjust for a figure of 1.2s or even the 1.5s it took from the time my cast was registered to the time the last hit landed.