Look! I'm going to do this all scientific like! Maybe if my thesis advisor runs across this, she will be proud instead of holding her head in her hands.
INTRODUCTIONAwhile back,
Matticus did a little test in an effort to determine which bubble absorbs damage first,
Power Word: Shield, or
Divine Aegis. If I understand his methodology correctly, he essentially cast
Power Word: Shield on himself as well as
Divine Aegis, committed suicide by way of
Molten Colossus, and examined his Combat Log to see which faded first. He tried this in various combos, proccing
Divine Aegis first, and also using
Power Word: Shield first. His conclusion was a "
first come, first served" method. In other words,
the bubble first cast/procced will be the first absorption used.
There was a comment on his post that suggested this might be inaccurate due to the way the combat log functions. I'm not really an expert on computery things, but I tend to agree with the commenter -
when multiple effects occur at the same time (such as all your buffs fading when you die), the combat log has to have some order in which to list them that do not necessarily make any contextual sense. I.E. I'm not sure if this really determined which absorption works first, because everything drops off at the same time when you die, and
is listed in a somewhat non-sensical way on the Combat Log. You can see this yourself if you ever note that your Combat Log may say you gained
Spirit Tap before listing that you have slain a mob.
So I tried to test it too. I'm not as cool as Matt though, I didn't use
Molten Colossi... I used various mobs around Icecrown while I did some dailies.
METHODOLOGYI tested this by running around screaming wildly at mobs, standing around stoicly as they beat on me, helping them along with
Shadow Word: Death, and tracking the results.
Anytime Divine Aegis dissipated by timing out instead of by being fully absorbed, I threw out that result.
As a side note, self buffed,
Shadow Word: Death is usually enough to make a
Penance-procced
Divine Aegis dissolve.
RESULTSBoth test versions were performed multiple times.
Test 1:
Test 2:
CONCLUSIONSBased on my not-really-very-scientific-either tests, it appears that in every scenario,
Divine Aegis is used first for absorption. However, I don't have any super fancy awesome addons that tell me how much absorption is left on each shield, so
the alternative conclusion is that whatever bubble you have with the least amount of absorption value left is the one used first.Ideally, this would be the best way for the absorption priority to work - the bubble with the least amount of absorption left gets used up first, thereby preventing it from being wasted.
DISCUSSIONThe main reason I suspect this is NOT the case is due to the
inability to cast a new Power Word: Shield on someone that is "less powerful" than the one they are currently wearing, even if the current remaining absorption value is lower. This is in response to a question I posted yesterday - if you've ever received the "a stronger spell is currently in effect" error message when trying to bubble someone, you have run into this scenario. When you cast your first bubble on them,
your spellpower was higher via a proc, or buff, or something like that, and therefore
your bubble bigger. When you go to cast another bubble on them before the current one has expired, but after weakened soul has fallen off, you are
unable to do so because now your spellpower (and therefore your bubble) is lower. It doesn't matter if there is only 10 hit points worth of absorption left on your target - it
uses the original value of the bubble to determine priority.The system is not currently "smart" enough to recognize this, most likely because it
doesn't do a very good job of tracking the remaining absorption value on a shield. Therefore, "
whatever bubble you have with the least amount of absorption value left" is most likely NOT the first one used, because I don't think the system recognizes that.
Feel free to leave me your thoughts, criticisms, suggestions, and results of any tests you perform yourself in the comments. I may try this again with some better addons to validate the potential alternative conclusion. It's a bit tricky, because you have to be very careful with the timing.