Mechanic Testing: Energy for FB w/Omen and Berserk

[raffy Profile]

aggixx brought up the point that I might not be implementing FB extra energy correctly during Berserk, and sure enough, after some dummy testing, it appears like Berserk energy saving apply to FB energy conversion as well.

I guess this was some ninja change during MoP, because afaik, since Vanilla, FB still consumed up to ~37 energy during Berserk.

While doing this testing, which I could only do every 3 minutes, I started testing how much energy FB took while Clearcasting was up. My methodology was: 1. get 5 cp, 2. get to 100 energy, 3. wait for clearcast, 4. cast FB and watch energy deficit. What I noticed is that, it was only consuming about 15-19 energy; I've never seen my energy drop below 80. Sometimes, it wouldn't even consume 10 energy.

I was wondering if anyone else could test this, as it could be addon or latency related.

Edit:

If instead, I cast SR at full energy, I definitely drop below 80 energy.

[Leafkiller Profile]

aggixx brought up the point that I might not be implementing FB extra energy correctly during Berserk, and sure enough, after some dummy testing, it appears like Berserk energy saving apply to FB energy conversion as well.

I guess this was some ninja change during MoP, because afaik, since Vanilla, FB still consumed up to ~37 energy during Berserk.

While doing this testing, which I could only do every 3 minutes, I started testing how much energy FB took while Clearcasting was up. My methodology was: 1. get 5 cp, 2. get to 100 energy, 3. wait for clearcast, 4. cast FB and watch energy deficit. What I noticed is that, it was only consuming about 15-19 energy; I've never seen my energy drop below 80. Sometimes, it wouldn't even consume 10 energy.

I was wondering if anyone else could test this, as it could be addon or latency related.

Edit:

If instead, I cast SR at full energy, I definitely drop below 80 energy.

This is due to latency in updating our energy. We reported this some time ago and the devs assured us it is working correctly.

[raffy Profile]

Still seems weird that clearcast FB is different from SR (both 25), but maybe FB is server-side and SR can be done client-side.

The other thing I'm trying to verify is that energy is quantized as an integer or, if actually behind the wow API, energy is fractional.

Simc implements energy as a float (double) but I have the suspicious feeling that this was a side effect of how they implement their regen mechanics (there is no tick, and instead it just adds energy using a single composite haste lookup for however much time the simulator advanced).

I implement energy as a tick, every 100 ms / composite haste, so I can always gain +1.

If it is actually quantized, then I need to figure out how it works under Berserk. Just looking at the tooltip energy, I figured out the rounding (which later found out to be banker's rounding algorithm). For example, Rake 35 goes to 18 (according to tooltips) whereas FB 25 goes to 12.

[raffy Profile]

If you edit any Ace-like addon, you can add the following event hook to monitor energy costs (this is only valid at 100 energy since the UNIT_POWER event doesn't fire every tick when there are no actions.)

Code: Select all

function addon:UNIT_POWER(event, unit)
   if unit ~= "player" then
      return
   end
   local cur = UnitPower("player")
   if self.last then
      local diff = self.last - cur
      if diff > 0 then
         print("Cost: ", diff)
      end
   end
   self.last = cur
end

[aggixx Profile]

You could probably prove one way or another by activating Berserk and then hitting Rake as many times as you could in a row, and see how the energy value you ended up at compares to the expected value.