Moonkin Raiding - Page 105

• Hamlet · 06/18/09, 6:16 PM

I'm not at all familiar with simcraft. How do you model cast time in relation to lag (especially near 1.0s)?

• Adoriele · 06/18/09, 6:22 PM

Your scale factors look a little wonky. Int, Spi, and Spellpower all take unexpected jumps, Haste drops more than I'd expect (though it could really be that bad), and latency takes a nosedive. That last one makes no sense whatsoever, as you're casting more wraths than before. Decreasing your latency should have a more positive impact, not a less positive one.

Starfox · 06/18/09, 6:25 PM

Originally Posted by Arawethion

I'm not at all familiar with simcraft. How do you model cast time in relation to lag (especially near 1.0s)?

Lag increases the time when the player is ready() again, different values for channeled spells, normal spells and after a gcd.
Channel lag: ~0.250s
Queue lag: ~0.075s
GCD lag: ~0.150s

• Hamlet · 06/18/09, 6:29 PM

Ok. So the effective cast time of an instant is max(1.5/HasteFactor,1.0) + GCDLag, and the effective cast time of a Wrath is max(1.5/HasteFactor,1.0) + QueueLag? What happens Wrath hits 1.0, does the cast time jump back up to 1.0 + GCDLag?

Starfox · 06/18/09, 6:35 PM

Originally Posted by Arawethion

Ok. So the effective cast time of an instant is max(1.5/HasteFactor,1.0) + GCDLag, and the effective cast time of a Wrath is max(1.5/HasteFactor,1.0) + QueueLag? What happens Wrath hits 1.0, does the cast time jump back up to 1.0 + GCDLag?

I think, but I have to ask nate about a more detailed explenation on the lag model.

Xerophyte · 06/18/09, 6:37 PM

You're also getting uptimes as

         73.2% : eclipse_starfire              
         66.4% : eclipse_wrath

This seems a bit implausible unless I am entirely misunderstanding what uptime means in this context. They should sum to less than 100%.

anyakaschala · 06/18/09, 6:40 PM

For me, as least my thoughts, DoTs become almost trivial dps, dps to be sure, but if eclipse is up 80%+ of the time, then that will and should be our primary concern. IS and MF's only time to be casted/refreshed would only be in that 20% time or fights during the dps downtime when you are moving.

I don't really have many thoughts about Owlkin Frenzy, in terms of raiding. I don't think it procs enough regardless to be an effective mana option.

• Hamlet · 06/18/09, 6:43 PM

Originally Posted by anyakaschala

For me, as least my thoughts, DoTs become almost trivial dps, dps to be sure, but if eclipse is up 80%+ of the time, then that will and should be our primary concern. IS and MF's only time to be casted/refreshed would only be in that 20% time or fights during the dps downtime when you are moving.

I wouldn't assume that; IS and MF still beat the DPET of Eclipsed spells. But there's the same tricky issue of whether to refresh during Eclipse or sacrifice some uptime to only refresh outside of Eclipse.

• Hamlet · 06/18/09, 6:44 PM

Also, what DoT refreshing rules do you use for the simcraft model?

I'm going to work up some WC stuff when I get home. But I'd like to have more firm understanding of the effective Wrath cast time when the listed cast time is very near 1.0.

• Adoriele · 06/18/09, 6:58 PM

Originally Posted by Xerophyte

You're also getting uptimes as

         73.2% : eclipse_starfire              
         66.4% : eclipse_wrath

This seems a bit implausible unless I am entirely misunderstanding what uptime means in this context. They should sum to less than 100%.

For Simcraft, uptime is defined as the percentage of time where you could have used that buff, and you had it. I.E. you don't care about eclipse when you're casting Moonfire. It's a kind of wonky way of measuring, but these values mean that ~73% of Starfires cast were affected by Eclipse, and ~66% of Wraths.

Lord BEEF · 06/18/09, 7:50 PM

Spirit being better than haste would pretty dramatically effect gearing.

Starfox · 06/18/09, 7:53 PM

Originally Posted by Adoriele

For Simcraft, uptime is defined as the percentage of time where you could have used that buff, and you had it. I.E. you don't care about eclipse when you're casting Moonfire. It's a kind of wonky way of measuring, but these values mean that ~73% of Starfires cast were affected by Eclipse, and ~66% of Wraths.

Yea, but there are plans for an options to switch to, for me somewhat useless, time-based uptimes.

Regarding scale factors:

  Druid_T8_58_00_13_SF       Int=0.51  Spi=0.48  SP=1.40  Hit=2.59  Crit=0.90  Haste=0.90  Lag=4.29
  Druid_T8_58_00_13_W        Int=0.57  Spi=0.47  SP=1.44  Hit=2.67  Crit=1.09  Haste=0.52  Lag=5.50

The only thing that seems weird is that haste drops below the value for a solar rotation.

Xerophyte · 06/18/09, 9:06 PM

Ah. The uptime values make more sense then and I can see how that metric would be useful. It feels a bit strange to use the term uptime for it, I guess.

Originally Posted by Starfox

The only thing that seems weird is that haste drops below the value for a solar rotation.

This isn't necessarily wrong. You're casting proportionally more Wraths than in a current Solar with SF filler rotation and the post-Lunar dot refresh you'll commonly do in 3.2 has a much higher probability of being under NG than the equivalent dot refresh at the end of the Eclipse cooldown in a Solar rotation. It's possible that outweighs the addition of a Lunar eclipse to the proceedings, though I agree that it seems weird.

• Hamlet · 06/19/09, 12:48 AM

Nah, that's far too low to make any sense.

Taelrien · 06/19/09, 3:57 AM

[removed my message because I miss read the eclipse change]