I am not sure about fire, but for arcane the usual cast time for my char with optimized gear (14xx haste, couldn't get any lower) is about 1.3s for AB, which puts 2AB very close to the optimal time for VPLC to proc (2.5s ICD). Maybe this is more efficient than for fire.

Edit: I did some quick SimC runs with my char in arcane and fire spec using nonheroic VPLC.

arcane:
VPLC: dps:1296, count:18.8, interval:23.42s, dpe:31281

fire (only respec, no gear change):
VPLC: dps:1177, count:21.1, interval:20.85s, dpe:25221

Although it procs more often, it does less dps, this is mainly due to the lower dpe.

I also simmed a well-known fire mage from our server, Shieki, who has about 7.5% more crit (same iLvl as me and mainly the same pieces, but different reforges of course)

fire:
VPLC: dps:1323, count:22.7, interval:19.42s, dpe:26347

It procs more often due to the higher crit chance (pretty obvious) and thus does more dps.

It remains why it does more dps per charge for the arcane spec. I read on mmo-champion (link) that it scales with mana adept. I have about 20 Mastery, i.e. 30% more damage at full mana. 31281/25221= 1.24 = 1+(0.3*0.8). An average of 80% mana is a little bit higher than what SimC is showing, but in the right order of magnitude.

To conclude, one of the reasons VPLC is THAT good for arcane is that its damage is modified by mana adept. Fire can compensate by having a higher crit chance (also increasing the chance of lightning bolt crits) and dots, which may trigger charges as well.

What optimizers are people using until this model gets picked up again for Rawr? Using it recently decreased my dps significantly because the enchants weren't correct for fire spec so I'm staying away from it for now.

Use Simcraft to find your stat weights then plug them into askmrrobot.

Simcraft for stat weight, Mr.robot for gear optimizer (with correct stat weight editing) and reforgelite in game for reforging optimization.

I'm so sad to hear that Kavan won't be updating this. I'm lost! Guy gets a pair of boots named after him and look what happens.

Just kidding Kavan, thanks for all the hard work on past tiers.

In any case I just checked in here to report that the stats for Inscribed Jade Fan are wrong and that the enchant Off-hand Major Intellect is not even in the cache at all.

Enchant Off-Hand - Major Intellect - Item - World of Warcraft
Inscribed Jade Fan - Item - World of Warcraft

I also found this issue

I did both wowhead item update and a manual update on the Inscribed Jade Fan and this resolved the issue


However the trinket from Elegon normal:

[Light of the Cosmos] i have updating from wowhead and says the item is unavailable - and manually updating does not seem to be possible

So sorry to see Kavan go - just so grateful for the work hes done! Also RAWR documentation references v5.0.3
Rawr - Documentation
but I cant see where to download it. Hopefullly the new version has all the new operation shieldwall items

You can do that. But realistically, your relative stat values won't change that much when you change reforge/gem unless you hit significant breakpoints.

As Spacedonkey says, in most circumstances the relative position of your stat weights will be the same, unless you change/reforge large amounts of stats.

If you really are in a situation, where the marginal dps gain of two stats is about the same, SimC offers you reforge plots to find a balance ( "sweet spot" ) between the two. But Simc really loses out compared to Rawr when it comes to analysis of marginal values, because the cost of reducing statistical noise to a low enough level is just too high ( and is reduced only by the square root of the number of iterations ). Rawr doesn't have to go down the path of linear approximation ( scale factors ) at all, but can directly calculate the dps of all possible item permutations for you.
It would be great if SimC could offer the same GUI item selection features ( with all relative dps values ) as Rawr, but that requires a speedup of factor 1000 or more, which just isn't possible.

The problem with using relative stat values to evaluate whether an item is an upgrade or not is that RSV completely ignores the effect of removing one item and equipping another.

Rawr computed DPS with old item dps vs new item dps and the difference was displayed. Using RSV the RSV of your gear with old item equipped is often not the same as the RSV with new item equipped, as Thot indicated.

To get an accurate evaluation with RSV what you need to do is to evaluate your equipment with the old item equipped and get a set of RSV for the old gear set. Multiply up this old RSV with the stats of the old item to get the RSV value of that item. Then evaluate your equipment with the new item equipped and get the RSV of the new gear set. Then multiply up this new RSV with the stats of the new item to get the RSV value of the new item. Compare old RSV value with new RSV value to see if item is a upgrade or not.

Even with this hideously onerous (ie: no one is likely to bother) process, you are not getting a DPS comparison only a rough guide. The major problem with the RSV approach is when caps are hit, so you CANNOT use RSV in any manner to compare when you remove an item with +hit on it and replace it with an item without +hit as the relative value of hit will dramatically changed if you are no longer capped.

Similarly RSV will not cope with hit becoming worth zero when you are capped. eg: Item has +350 hit on it this takes you to the cap calculating RSV would give zero for hit as you are capped and more is worthless. However if you wanting to see if its worth replacing that item the 350 hit is valued at zero due to cap. However without that item you aren't capped so actually hit from that item does have a value. This is my point about the procedure above to use RSV properly you MUST consider the effect of removing the item under consideration for an upgrade.

Rawr didn't need to do any of this as it worked entirely on calculating DPS not the shifting sands that is RSV. It's a little known fact but Rawr did not use RSV in any way at all.

To answer Thot's question you break the loop by comparing pre and post situations. Get RSV values for what you have and then RSV values for the fully iterated final setup. Be prepared to run SimCraft numerous times to get the iteration to settle down.

I updated Rawr.Mage with the following changes, and afaik a new release (Rawr5.0.3?) is planned for this week or next.

- Critical Mass nerf to 1.30x
- Combustion nerf - removal of pyro contribution
- Arcane Charge buff to 25%dmg / 75%mana