Rawr (v2.3.23 released on 9/12) - Page 47

• Wizeowel · 05/26/08, 12:42 PM

Just for further example, I've been trying to figure out how to play this sequence. (From a six-minute fight.)

Icy Veins+Arcane Power+Serpent-Coil Braid+Icon of the Silver Crescent+Arcane Blast: 25.00 sec (3627.02 dps, 306.96 mps, 1492.88 tps)
Serpent-Coil Braid+Arcane Blast: 5.00 sec (2138.15 dps, 183.82 mps, 880.06 tps)
Icon of the Silver Crescent+Arcane Blast: 15.00 sec (2085.25 dps, 183.82 mps, 858.29 tps)
Heroism+Arcane Power+Serpent-Coil Braid+Icon of the Silver Crescent+Arcane Blast: 5.00 sec (3713.93 dps, 318.47 mps, 1528.65 tps)
Heroism+Icy Veins+Serpent-Coil Braid+Icon of the Silver Crescent+ABFrB3FrB: 10.00 sec (2552.46 dps, -0.83 mps, 1412.03 tps)
Heroism+Icy Veins+ABFrB3FrB: 20.00 sec (2207.76 dps, -0.83 mps, 1220.07 tps)
Heroism+Icy Veins+Icon of the Silver Crescent+ABFrB3FrB: 5.00 sec (2348.36 dps, -0.83 mps, 1298.37 tps)
Arcane Blast: 191.43 sec (1968.13 dps, 183.82 mps, 810.08 tps)
ABFrB3FrBSc: 73.61 sec (1397.17 dps, -58.84 mps, 758.49 tps)

My actual plan for the fight goes like this:

at 25 seconds: IV+AP+SCB+Icon+AB
at 150 seconds: Coldsnap+IV+SCB+Icon+AB
around 240 seconds: evocate
at 300 seconds: Heroism+AP+SCB+Icon+AB
at 340 seconds: IV+AB for 20 seconds

This seems a sensible way to do it, but doubts start to form in my mind that my way would be 100dps less than the way Rawr is trying to tell me to do it.

Kavan · 05/26/08, 1:35 PM

Originally Posted by Wizeowel

This seems a sensible way to do it, but doubts start to form in my mind that my way would be 100dps less than the way Rawr is trying to tell me to do it.

I think those cycles are overestimated. Have you tried checking what you get with SMP? The normal solver is good enough for making item comparisons, I've rarely seen it make an error of more than 1%, most usually it's in the 0.1% range. But for actual planning of how to do things in game I strongly recommend enabling SMP and checking what result it gives you. Doing that you will also see how much of an error the faster base solver is making if you mouse over the * next to dps.

rhodin · 05/26/08, 7:19 PM

Can't wait till you guys finish the hunter model, keep the good work up! Im loving both the druid models <3

• Wizeowel · 05/27/08, 6:58 AM

Apologies Kavan, I should have said those cycles are with SMP selected.

Kavan · 05/27/08, 7:21 AM

Originally Posted by Wizeowel

Apologies Kavan, I should have said those cycles are with SMP selected.

Can you send me the xml where that happens?

• Wizeowel · 05/27/08, 8:06 AM

I've posted a new work item to the codeplex site.

gcbirzan · 05/28/08, 5:36 AM

Rawr.Mage is giving me an... interesting answer.

With my current gear, I turn on most buffs, no flame caps, no totem of wrath, no moonkin aura, replace my neck with the MH trash one, my OH with the badges fire one and look at the item budget values. If I modify any of the above, spell haste is just barely better than spell damage (10vs11.7), but with that setup, rawr says it's almost twice as better as spell damage. Boggle?

You can check out the xml I use to get that at http://208.78.101.115/~gcbirzan/poons-spellhaste.xml

kidvid · 05/28/08, 10:43 AM

Is there a way I can configure my RAWR healdin module to better value things in a way that would make more sense in a 5v5 Arena. For one, it would have to ignore consumables, and as far as what I could tweak, I guess I could turn off down-ranking, and shorten the fight length to a couple of minutes... Any ideas on what settings I could change?

• Wizeowel · 05/28/08, 11:14 AM

Originally Posted by gcbirzan

If I modify any of the above, spell haste is just barely better than spell damage (10vs11.7), but with that setup, rawr says it's almost twice as better as spell damage. Boggle?

Indeed, I could reproduce this effect in 14.1 using your save file. However, using a more recent build of the software, it doesn't happen. So I guess it's a bug that is fixed.

Elrinnas · 05/28/08, 2:50 PM

Originally Posted by Binkenstein

I'm still hunting down something to teach myself C#. As it is, looks like I'll need to get something from Amazon (and pay just as much in shipping to NZ)

If you find a good reference guide to learn C#, let me know and I can throw my name into the ring for the Hunter model. As long as one of the other hunters supply the theorycraft.

Astrylian · 05/28/08, 3:11 PM

Originally Posted by Elrinnas

If you find a good reference guide to learn C#, let me know and I can throw my name into the ring for the Hunter model. As long as one of the other hunters supply the theorycraft.

Visual C# Developer Center

That's a good starting point.

Carnivean · 05/28/08, 8:14 PM

First I want to thank Kavan & Astryllian for all the hard work you guys out into this.

Now on to my problem.

I am working on figuring out the right Sequence of using my CDs at Brutallus (ok it's not that hard), but Rawr is the only tool avaible, that gives me am accurate DPS result.

I have Global Optimization, Incremental Optimization, Reconstruct Sequence & SMP all activated, but I get some funny results for the spell sequence:

ImageShack - Hosting :: rawrspellsequencehc6.jpg

Is there anyway, that I could improve my result?

Thanks for any help.

Ivorthemage · 05/29/08, 2:33 PM

Pretty sure I am missing something obvious here, but I was trying to model the dps impact of an innervate last night using Rawr. I entered in one innervate, and nothing changed. Same dps, same recommended spell rotations (which involved rotations). On a hunch, I tried the same thing with mana tide, and had the same problem.

Is this a bug, or am I messing up somewhere?

Awesome, awesome app.

Docjowles · 05/29/08, 2:44 PM

Is it possible the situation you modeled simply didn't have you getting low enough on mana that the innervate mattered? Unless this was an AB-spamming arcane build, I could easily see an innervate making no difference. Try increasing the fight length or decreasing the shadow priest MP5, or give us some more details. It certainly could be a bug, but you need to rule other factors out first

Ivorthemage · 05/29/08, 2:51 PM

It is an AB spamming arc build, and the spell sequence was showing more mana efficient spell rotations being used. Presumably getting 10k more mana would shift that balance to more AB spam. As such, it looks to me like the innervate simply wasn't registering. I kept looking for a "recalc" option...