This is my first post here so please be gentle. :}

Im trying to figure out if my gear is ready for switch from my frost build to a fireball or even a ffb build. My last try sucked because I have a low crit rate. I am completely stumped at why The World of Warcraft Armory reports one crit% 11,89 (282) and the same % is 27,78 (282) I guess it is because the buffs?

So I removed all buffs in Rawr 2.1.8. Then the crit% became 27,60 (282). I just do not understand.

Once I get a grip of how much crit% I have can I get a DPS raise going Fireball or Frostfire bolt with my gear? I have some bits to swap out for more crit. Also my list of loot I aim for is taken from rawr when I use talentoption 0/53/18.

I realy need some help to understand the dramatic difference in crit% based on the same crit rate.

Thanks in advance. BTW my off hand is not equiped after I fished. It is the Ward of the Violet Citadel which only have bearing on the fire build not the difference between the crit%.

This is also my first post on these boards.

I have been using Rawr for a couple of days now, trying to get the hang of reading the information it's giving me.

Firstly on spec, I'm obviously not getting what I should out of it, selecting the 13/0/58 frostbolt spec places it on the top of the TalentSpecs chart, but this also happens when I select FFB or arcane as the selected spec.

Secondly, gems, I'm seeing 20AP 10HR gems coming up as a suggestion for gemming, obviously the HR is going to be the biggest boon to my DPS until I cap but then why doesn’t it suggest a pure HR gem? I tried fiddling with the Tools > Set default gemming, but I just don't understand how to work this feature.

I also see multiple occurrences of the same item in the charts with different gemming, is this simply due to the fact while a 16HR gem may be a good upgrade over the current item or stats, it may not be the best choice when optimised? As a score is given based on the upgrade if nothing was equipped in the slot?

Finally, is there any way to stop it enforcing socket colours for the bonus? Because logic would dictate that a hit rating gem is a bigger DPS boost than socketing a red gem for +6 int.

I apologise if these questions have been answered elsewhere, I looked through this thread and saw an answer about the gemming issue, but I don't understand how to use the Set default gemming tool to fix it.

Thanks in advance to anyone that can help.

I'm guessing you are mistaking the item comparison lists for some kind of suggestions that Rawr is making? It doesn't work like this. Those lists are just showing what the dps value of an item is if you happened by chance to use that gemming. So if the list has some odd gem combinations, just remember that this application is made for other classes as well for mages and that's why it might have a starting selection with strength and expertise in your cloth helm.

So if you only play a mage then you can just do away with all the gemmings that are automatically created for the other models. Open Rawr, don't load your character yet but just choose the Mage model. Open 'set default gemming'. Check the box marked 'delete other gemmings'. Put runed scarlet rubies in the 3 colour slots, and then the chaotic skyflare meta gem in the last slot. Now click 'regem' or whatever the button is called, then close and restart Rawr. Now what you have is an 'item cache' where all the items are gemmed for high dps as default. This way, what you'll see in the item comparison lists is a fair reflection of an item's potential.

But now the question: how do you use Rawr to find the optimal gemming for your gear? There is a tool called optimiser which can help you do this. Use the comparison lists to mark (green diamond) which gear and enchants are available to you. Also select which gems you are considering (in general mark the highest red, yellow and blue gems from the list, plus the veiled gem in case you need a hit adjustment). Make sure to click 'enforce meta gem' so that it knows to give you 2 blue gems for your CSD. Not forgetting of course to fill in all the buffs and fight options correctly. And now run the optimiser and it will make a gear set with gems for your highest dps.

After you've done this I do recommend switching off 'enforce metagem' again. Because what happens after the optimiser regems for you... well it adds the blue gemmed items to your comparison lists and since those are needed to activate your meta gem, those items will always seem better than the red-gemmed default items you made earlier. So by switching off 'enforce metagem' Rawr will again look at items without considering the metagem.

Unless i'm mistaken it look likes the Elemental precision talent is still with % hit to frost and fire spell instead of all spells...

I've created a fake gem with 3% hit for my gloves to accurately model my hit score with arcane spells

Hi Kavan, I have a couple of questions regarding 2.1.8:

1. Do AMABar latency combos assume clipping the 5th pulse of AM?

2. Has JoW been updated to be consistent with 3.0.8 mechanics? The tooltip still reads .25% Max Mana/Hit.

Thanks in advance.

Are you sure you are using the latest version? I can't reproduce what you're saying. For me as soon as I add Precision it increases hit% for arcane.

How is the ABarr bug currently modeled? It doesn't seem to make extreme DPS changes like I'd expect when changing success rate from 0 to 1 (3666 dps VS 3884, I would expect a bigger loss if all ABarrs don't get 60% damage buff).

Forcing scorch seems to only use rotations that end with scorch, even during AP, yet at the end up only casting 11.3 scorches which is fine for the 6 min fight duration. Unless I'm missing something, this is rather wierd since you can re-scorch at pretty much any time that isn't the AM-ABarr part of the rotation.

Does PoM ever get used? I'd like to see if PoM pyro is actually worth it with a scorch spec over PoM AB (which I'd guess would be worth it), however I don't see any PoM being used at all in the rotations.

Why is it using a fraction of a mana potion but full evocation and not the other way around? It seems to be combining the mana potion with a speed potion...

If you model clipping, should probably leave it to the user to be able to set a chance to hit ABarr at any of the reasonable times that it could happen (for example, I'd probably set mine to be something like average between 0.5s b4 AM is over to getting it spot on, and 0.3s before to 0.2s after on procs, unless I can get a better way to do it while keeping errors in mind). Maybe even give a "ABarr miss time" which would mean how accurately you can time your ABarr and actually find the optimal time for hitting ABarr taking the random error into account. And of course telling me at exactly what time I should actually hit my ABarr to get optimal DPS with my error.


It would be nice if maintaining scorch would also have an option for how many times or how often you have to completely restack scorch in a fight (switching targets, running around for too long etc).

The difference is less than you expected most likely because it selects a different cycle in each case. It will most likely prefer AB3AMABar at 100% reliability and AB3ABarC at 0% reliability.

The scorch cycles don't imply that scorch is a regular part of the cycle. For example ABAMABarSc is a normal ABAMABar cycle with only casting scorch when needed to refresh the debuff. Right now it is not supported to model scorch timing explicitly. As a result all used cycles average in expected number of scorches needed to maintain the debuff, even during cooldowns. While this is not completely accurate, the additional computational cost needed to model this accurately is not worth it.

PoM is not modeled. I might include it at some point in the future, but it is currently in the same situation as the above, the added complexity is not worth it. This is something that would be very easy to fit in a simulation type tool, but is a lot more involving in an optimizer.

The default optimizer has the ability to use partial consumables. This usually provides better results for purposes of gear comparison. If you would like it not to use fractions there is an advanced option for integral mana consumables that you can use, but I would advise you to read about it first.

How to accurately model the clipping is a topic for discussion. I've proposed some open questions in the arcane thread. Right now I think I will only provide options to enable clipping on AM or MBAM, but not any timing controls.

I have considered adding means to model restacking scorch, but since the cycles usually operate on steady state assumptions it is not a good fit for modeling. That does not mean that it would be impossible, but I think there are other fight mechanics like movement that would have higher priority on the list if I wanted to increase the complexity of modeling. It would be nice to have a completely accurate model of in game fights, but everything has its cost in development time.

Another wierd thing I've noticed, is that it seems to often (but not always) prefer running me oom than using a different rotation. Also does is AB-ABarr better efficiency cycle than AB2-ABarr or is the latter simply not modeled? Thinking of which, it would be nice to be able to see a list of all supported rotations, including the ones that weren't used (unless it's there and I'm missing it).

How come PoM is so hard to model? It's just 1 more cooldown, isn't it?

For proc trinkets, is there anything that takes into account the fact that spell damage is calculated on cast finish while haste is calculated at cast start? For short duration procs this can make a noticeable difference, so it would be good to know if they're being over/under-valued.

A quick question regarding Rawr that I am a little unsure of.. if I choose blood elf as my race, will Rawr automatically calculate the extra mana gained from Arcane Torrent every 2 minutes (off the GCD)?

It's not huge, but the 6% translates to some 1300 mana when you have 22k mana raid buffed for free.

If not, could it be added for a future version?

You can see the list of all supported cycles in the spell info section. Efficiency of cycles can't be evaluated just by looking at a single cycle, but must be viewed in the context of the complete problem. Both of those cycles are modeled, again check the spell info section (it's on the Stats tab if you scroll down).

Pom is basically just one more cooldown yes. But each additional cooldown basically doubles the number of combinations that have to be examined. In addition PoM is a bit unique since it operates on a single spell and not on a cycle, so it would need additional support to fit in the framework.

For haste procs that is taken into account and the number of affected spells will depend on whether a spell is a cycle. This should be accurate for single spell spam, but the modeling of such procs on cycles is currently done by averaging, so it might not be completely accurate.

Considering arcane torrent does not consume a GCD (and the silence is generally useless in raiding), there's absolutely no reason not to use it whenever it's up for free 6% mana refill.

Sorry for the bump.

Howcome the crit rateing gives different crit% in WoW/Armory and Rawr 2.1.8?

Crit rate 282 gives ~27% in Rawr but ~12% in WoW/Armory.

Are you sure you're taking account of all the buffs? Scorch/Winter's Chill depending on spec? Is armory crit% the same as character crit% in game? As far as I know no one else is having this problem so we'll need the xml file you're using to see what you're doing wrong.

mmmmhh it seems i was dreaming, can't reproduce it myself too but the tooltip is still saying fire and frot anyway.

Sorry for the mistake.

This is my xml-file. Currently I have some buffs on it. While you look at it would it be feasable to go ffb or arcane?

http://web.comhem.se/~u43490656/Moder.xml
http://web.comhem.se/~u43490656/Moderffb.xml
http://web.comhem.se/~u43490656/Moderarcane.xml

I turned off those buffs and my Rawr 2.1.8 says 28,34% (307). Armory says 12,7% (307). How can that be?

Anyways, is it ok to switch to a fireball build or arcane?

When I disable all buffs I get 12.73% crit. Several things to look for that you probably overlooked. In a frost spec with Winter's Chill the buff is automatically applied even when not selected in buff list. When you have Focus Magic talent by default you will gain crit based on Focus Magic Rate setting. Additionaly when Automatic Armor is enabled it is very likely that Molten Armor is selected as optimal, which will also increase crit.

That figures. I will have to toy around more with Rawr. Many thanks guys.

[Illustration of the Dragon Soul] wont show up in the item/gear select list even though it's in the Item Editor. But if I manually mess with its stats it does. Saw the post saying this was fixed, so is it just me or?

It broke again in 2.1.8, due to the changed wording in wow 3.0.8. We'll have it fixed for the next version.

I understand Rawr encompasses several classes, but since no cloth class would ever gem for strength and expertise, perhaps a feature could be added to Rawr to limit the gems it presents in the gear list based on what class you've chosen to model. Right now, as soon as I install any new version of Rawr, I just go through each equipment slot and delete every occurence of a piece of loot using inappropriate gems. This would just make life easier by automating that.