Frost Cataclysm Discussion [4.0 thread] - Page 21

Altiris · 11/24/10, 3:56 PM

Originally Posted by Manbis

I did look on the numbers and frostfirebolt does more dps then frostbolt. And if i take ignite into account should it not be better to cast frostfirebolt and nomore frostbolt? Both are proccing FoF... Frost only has the 5% crit but that doesnt seem so much compared to the better dps from frostfirebolt and his glyphed dot which should be constantly on the boss. Or iam wrong?

Frostfire Bolt isn't capable of proccing Brain Freeze, which is intended to sway the equation back in favor of Frostbolt. I believe Blizzard wants Frost to use Frostbolt, so even if Frostfire Bolt were better right now, they would change it so it wasn't.

Roywyn · 11/25/10, 5:51 PM

Originally Posted by Lhivera

One of the things I was doing in my sim was that if the test increase to a crit bumped it up past 34%, it would automatically replace the MA glyph to ensure that the crit rating had value.

How about Glyph of Molten Armor - Item - World of Warcraft - "Your Molten Armour increases your spell haste by 2.8%"? Or "Your Molten Armour causes your spells to deal 5% more damage against frozen targets"? Crit above the soft cap can always be reforged into hit, haste and mastery, so we'd have to check whether it's beneficial to swap the glyph or to keep them while reforging the excess crit.

Lhivera · 11/25/10, 9:42 PM

Testing on Beta had indicated that the latency between Waterbolt casts was gone, and we could drop /cast Waterbolt from our macros -- it looks like that was just something happening on beta. Waterbolt latency is very much in effect on the live servers. I'm seeing an average time between casts of over 3.1 seconds.

Taelons · 11/26/10, 10:11 AM

@Lhivera
Maybe I'm misinterpreting your post. But what i noticed during my testing is that the pet sometimes does really long casts of Waterbolt but it corrects itself by having the next Waterbolt being cast faster.

I did some 2 minute tests on beta a while back with latency and Waterbolt.

1e: 200ms No macro: 50x Waterbolts cast
2e: 200ms Macro: 50x Waterbolts cast
3e: 600ms No Macro: 49 Waterbolts cast

Now since it's only 2 minute tests, I might have had a half a second difference in stopping the Stopwatch which could have resulted in that 1 Waterbolt difference. I also have these tests frapsed so I could upload themt o youtube on request.

Lhivera · 11/27/10, 4:44 PM

This should not be considered complete, but a ton of progress has been made on the Mage module of SimulationCraft over the past couple days, and I thought I'd post how the specs are shaping up relative to each other.

The sets are not well-optimized, and there is certainly some improvement that can be made to the priority lists for Fire and Arcane. Also note that as the sim doesn't yet support DOTs on adds, Impact DOT-spreading isn't included in the Fire results for the runs with adds.

5-minute encounters:
Vanilla: Simulationcraft Results
5 secs movement per 30 secs: Simulationcraft Results
3 adds near boss for 20 secs per 60 secs: Simulationcraft Results
Movement and adds: Simulationcraft Results

7-minute encounters:
Vanilla: Simulationcraft Results
5 secs movement per 30 secs: Simulationcraft Results
3 adds near boss for 20 secs per 60 secs: Simulationcraft Results
Movement and adds: Simulationcraft Results

f1reburn · 11/27/10, 6:17 PM

Interesting results, glad that Frost appears to do better than in Wotlk. Speaking of simulators, I just plugged your simcraft stats into Magegraf and got these stat weights for Frost:

SP 	Intellect	Crit 	Hit 	Haste 	Mastery
1 	1.48 	        1.24 	0.01 	2.75  	0

These are drastically different from the results I got about a month ago and would make me spec out of Leatherworking, which was the best profession at the old weights where Intellect rules all. Could you give me a rough idea of the stat weights in Simcraft?

Vontre · 11/27/10, 6:19 PM

Originally Posted by f1reburn

Interesting results, glad that Frost appears to do better than in Wotlk. Speaking of simulators, I just plugged your simcraft stats into Magegraf and got these stat weights for Frost:

SP 	Intellect	Crit 	Hit 	Haste 	Mastery
1 	1.48 	        1.24 	0.01 	2.75  	0

These are drastically different from the results I got about a month ago and would make me spec out of Leatherworking, which was the best profession at the old weights where Intellect rules all. Could you give me a rough idea of the stat weights in Simcraft?

Magegraf gave you that? Can you link your result set?

f1reburn · 11/27/10, 6:24 PM

Turns out I was using raid-buffed Intellect and Spellpower values instead of unbuffed values, d'oh.

This is the link with the proper values, not sure if a link like this works: Vontre's Magegraf. The combat ratings are still higher in value than I expected in comparison to Intellect. This is for a 5-minute fight.

Vontre · 11/27/10, 6:31 PM

Originally Posted by f1reburn

Turns out I was using raid-buffed Intellect and Spellpower values instead of unbuffed values, d'oh.

This is the link with the proper values, not sure if a link like this works: Vontre's Magegraf. The combat ratings are still higher in value than I expected in comparison to Intellect. This is for a 5-minute fight.

Oh, old Magegraf. That's on the WLK rules. I suppose it's about time to update that version, huh?

Updated now! You should be seeing realistic numbers there.

TyrianeKdV · 11/27/10, 9:53 PM

Originally Posted by Lhivera

5-minute encounters:
Vanilla: Simulationcraft Results

Is it possible, that the Frostfire Bolt-Glyph-dot is not included in the calculations? At least it is not listed as damage source, and Frostfire's dps gain per mastery point is less than Fire's, which seems a bit strange.

Brainpower · 11/27/10, 10:02 PM

Originally Posted by TyrianeKdV

Is it possible, that the Frostfire Bolt-Glyph-dot is not included in the calculations? At least it is not listed as damage source, and Frostfire's dps gain per mastery point is less than Fire's, which seems a bit strange.

It could possibly come from the fb glyph that adds 5% crit so fb produces more ignites than ffb.

Lhivera · 11/27/10, 11:36 PM

Originally Posted by TyrianeKdV

Is it possible, that the Frostfire Bolt-Glyph-dot is not included in the calculations? At least it is not listed as damage source, and Frostfire's dps gain per mastery point is less than Fire's, which seems a bit strange.

It's not listed separately. If you look in the rightmost 4-5 columns of each damage source, you'll see various columns related to DOT ticks; on the FFB line, that's from the glyph.

Gain per mastery point could have something to do with the gearing or the priority list, both of which have had only the most cursory work done toward optimizing them. Or it could be that Fire having more Pyros and crits just plain gets more out of Mastery (though that would be counter to the intended design, I think).

Really, I'm pretty impressed with how close the specs seem to have come out. At first glance, Frost looks like it's lagging 10% back, but as soon as you start throwing in movement or making the fight longer, it gets up into a much more comparable place. It never seems to beat Arcane or Fire, but it should be plenty usable for anyone who isn't chasing first-kill bragging rights, and depending on how useful its survivability is, maybe even then.

Unfortunately, it does look like at least one, possibly two points need to go into Clearcasting; putting those points into Netherwind Presence instead results in a loss of a few hundred DPS after about 4.5 minutes. This surprises me, as my own sim had Frost's endurance seeming much higher, but so far I haven't found an error in SimulationCraft that would lead me to think it's doing something wrong.

ETA: I've corrected a significant Arcane bug, improved its priority list, and I've slightly improved Frost's gear set. I've also added 10-minute versions of the Mage-only output for a better view of how the specs degrade over long fights. Easiest is probably just to browse the results, but if you look at the raid results, bear in mind that many of those classes may not have received as much attention as they should, so the results are iffy so far. (Hell, ours are still a little iffy.)

Index of /misc/sc403

Lhivera · 11/28/10, 12:23 PM

(double post, please delete)

Nathyiel · 11/28/10, 1:17 PM

I have a few reflexion on your results:

Your Evocation's count is low for Fire and Frost (1.9 on 599s fight) but not with Arcane.
Your Mana Time's line is a bite mysterious. It show mana problem for fire and mana management for arcane, but I don't comprehend why mana shut down at the end of all fight.
Your Mastery weight seem low. Is it because of crit below the soft cap?

I think that in frost, Evocation can be use differently. Evoc began with 15% of mana max on use, with other tick channeled. We can use it for the first 15% each time we go under 80% (by exemple). And we can/should use it to go full at the end of 2nd IV (or hero/blood).

Last question, what about spec into 1pt Clearcasting and 1pt Netherwind Presence.

Jaerel · 11/28/10, 2:39 PM

I'm not great at reading these outputs, but it looks like maybe the sim for fire & ffb is taking a poor approach to its mana management such that it's needing to drop LB (and the ffb dot) from its rotation for significant portions of the fight, and that probably isn't intended. (either LB is worth the mana to cast relative to FB or FFB so it should be near 100% uptime, or it's not so it should be 0%, as is it looks like it's ~65%). Similarly, for the FFB profile, some behavior that triggers conservation earlier so as to preserve the 3 stacked dot near 100% uptime would be a stronger approach (I would think) vs. spamming the mana into consecutive ffb's early.