Reforging by hand
I installed Reforgelite to try it out, and found it pretty good for a while. Then I tried to play with not-quite-hit-cap while still meeting my haste threshold, and Reforgelite proceeded to disappoint me on taking 4-5 attempts per new piece of gear to find a set of stats I was pleased with. That’s easily 500-800 gold of reforging, and I just didn’t want to waste more on it. I tried my old by hand method with pen, paper, and basic arithmetic, and netted myself a better simulated DPS with that reforging than anything Reforgelite had ever given me with that gearset.
So I’m back to my old method of reforging, which is doing the math by hand. It’s not even difficult math. It’s mostly a logic problem. I referred to it on Twitter as being like the old logic problem where you have a two-seater boat and you must get a wolf, a goat, and a cabbage across a river. The wolf will eat the goat and the goat will eat the cabbage, so you have to pick which you leave alone wisely.
So let me show you how I do it. You might think it’s just a bunch of work, but what works for you works for you and what works for me works for me.
Since I do this a lot, it doesn’t take me very long. 5 minutes? It’s not the seconds an addon might take, but I don’t mind it.
Stat weights, gear upgrades, and reforging
One stat weight priority does not, in fact, rule them all.
The stat weights that come out of SimulationCraft are meant for additions to the gear that has already been simmed. You may judge enchants, gems, and upgrades in ilevel with these, since you are doing nothing to the original gear. Reforging, however, is not an addition, but a transfer. Different reforges will yield slightly different simulated DPS and stat weights.
Different gemming schemes will also simulate differently, since you are replacing gems not adding gems. Upgrades in ilevel will generally increase your DPS, as they’ll have more of a stat (primarily, more intellect) on them than before, but a change in secondary stats may put a minor ilevel upgrade a little below your current piece, reforging depending. With a good understanding of how a gear change will affect the simulation, you can guess without re-simming.
But it is safest to re-simulate for every gear change you make, no matter how small.
However, you can use the stat weights as a general guideline for what reforging strategy is the best. Ideally, your secondary stats — mastery, haste, and crit — will align to the same value when you’ve found the perfect reforge. That is, you have the Goldilocks amount of each stat on your gear.
On hit: Similarly, because the stat weights address additions to your gear, the weight for hit will drop significantly when simulated with a hit-capped gearset, as any additional hit will be useless. This is not evidence that the hit cap is insignificant. Meeting the hit cap helps you deal with your personal lag at realizing when your spells have missed a target. If you don’t want to worry about missing, then cap your hit as close as possible to cap.
I don’t like missing. It’s just fucking annoying. Thus I try to cap my hit within a certain margin.
My personal margin for hit used to be within a hybrid gem’s worth of hit. As my current gear will reforge for more than that, but not quite double on some pieces, I’ve now accepted 4900 (14.41%) as the lowest I’ll go, though I prefer 5000-5100 (14.70-15.00%).
I used my most recent simulation stat weights for the gem calculations. My stat weights have changed only a little (a margin of 0.10 points, maybe) over about 4 gear piece changes, so I feel confident using recent stat weights on a 1-piece change. In this case, especially so, since the stats on that piece were the same (mastery & crit –> mastery & crit). My most recent simulation yielded these stat weights:
- 3.81 intellect
- 3.13 spellpower
- 1.94 mastery
- 2.05 haste
- 1.64 hit
- 1.35 crit
- 0.06 error margin
In the past, I’ve found that mastery rises significantly after I’ve reached a haste threshold of 11-12% (4677-5102 rating) on my gear. This is not surprising considering the major T14N haste DoT breakpoint is 4717 (Corruption +2).
Previously, mastery had been greater than haste, but this simulation, my gear was severely lacking in haste pieces. Even with reforging mostly to haste, I barely hit the 4717 breakpoint I wanted.
This time I was only a few reforges off the haste breakpoint, so I felt confident going for the “haste (4717) > mastery > haste” gearing/reforging strategy.
You can’t reforge to another stat already present on that piece of gear. The exception is that expertise counts like hit for spellcasters, so reforging to expertise on a piece of gear that already has hit is effectively adding hit without breaking the rule.
Other than that rule, reforging is about knowing when to stop (caps and breakpoints). The more you know about specific amounts of a stat affecting your gear, the better you can reforge without an optimizer.
I’ve enchanted and gemmed each piece separately. I’m trying out pure hit gems in blue sockets, so I can push as much reforging towards favorable stats. Once I reached about 11-12% haste, my mastery stat weight grew so that hybrid gems (orange for mastery/intellect) started to overtake my pure intellect gems (red), so I gem on a yellow-based scheme:
- Prismatic or yellow sockets = +320 mastery / Fractured Amberjewel
- Red sockets = +80 intellect & +160 mastery / Artful Vermillion Onyx
- Blue sockets (way below hit cap) = +320 hit / Rigid River’s Heart
- Blue sockets (close to cap) = +160 hit & +160 mastery / Sensei’s Wild Jade
I start with all my gear unreforged. It doesn’t cost gold to unreforge, and at worst, I might spend 150-200g reforging a bunch of pieces. But this starts with me what I need to achieve and all my options.
I already have a shorthand I write with; most are obvious, except hit is T since H is already haste. My stat priority is I > SP > T (to cap) > M >= H (4717) > C. I also have reduced the slots to two-letter combinations in my shorthand. I used to have a “wd” for the ranged slot, and a mh/2h designation, but not anymore!
I write down the slots, their secondary stats in priority order, and how much the lesser stat will reforge into. Parentheses indicate I’d rather leave that piece alone, but here’s how much it will reforge to, just in case. Writing down both stats helps me lay out which stats I can or can’t reforge into. The second column is for knowing what pieces I’ve set aside for reforging.
|hd = TC –> 257||? –> ?|
|nk = T(M –> 145)||? –> ?|
|sh = HC –> 205||? –> ?|
|bk = MC –> 166||? –> ?|
|ch = MC –> 299||? –> ?|
|wr = T(H –> 184)||? –> ?|
|ha = T(H –> 260)||? –> ?|
|wa = M(H –> 220)||? –> ?|
|lg = TC –> 303||? –> ?|
|ft = M(H –> 220)||? –> ?|
|f1 = T(M –> 129)||? –> ?|
|f2 = T(H –> 166)||? –> ?|
|t1 = n/a||? –> ?|
|t2 = (H –> 382)||? –> ?|
|mh = MC –> 92||? –> ?|
|oh = T(M –> 150)||? –> ?|
I want to meet a specific cap (5100 hit but no more) and a specific breakpoint (at least 4717 haste). So I write down at the unreforged totals:
- 4965 hit
- 4177 haste
I want a tiny bit more hit (135 rating to cap) and a handful of haste (540 rating).
Immediately, I see from my list that reforging the main hand dagger to hit will net me pretty close to cap. I’ll have 5057 hit rating (14.87%), which is close enough for me.
It’s time to work on haste. I need more haste, so I don’t want to reforge any of it off just yet, so I strike out all the “H –> #” reforges. I also would prefer not to reforge off any mastery, so I temporarily strike out my “M –> #” reforges. I also strike out my shoulders as reforging to mastery, since I obviously cannot reforge the crit on there to haste. Here’s what’s left:
- hd = TC –> 257
- bk = MC –> 166
- ch = MC –> 299
- lg = TC –> 303
I won’t be reforging my back and chest pieces to mastery, since they already have mastery on them. Leaving them alone would be leaving all that unwanted crit on there, so that makes them prime pieces for haste reforging.
166 + 299 = 465, which is 75 off from the desired 540+. I do have some haste pieces without mastery on my reforge list, however, and I could always reforge those off into mastery to hit the correct haste breakpoint. Let’s look at those:
- wr = TH –> 184
- ha = TH –> 260
- f2 = TH –> 166
- t2 = H –> 382
The lowest +haste reforge is the head slot at 257, and the matching -haste reforge that I chose was the second finger slot at 166.
Here’s the math again: 4177 + 166 + 299 + 257 – 166 = 4733. That’s only 16 rating off the haste breakpoint. Pretty tight, eh!
I then let all the remaining mastery-able reforges be mastery, since affliction loves that stat right now. Here’s the end result with 5057 hit (14.87%), 4733 haste (11.14%), and 6204 mastery (55.02%):
|hd = TC –> 257||C –> H|
|nk = T(M –> 145)||–|
|sh = HC –> 205||C –> M|
|bk = MC –> 166||C –> H|
|ch = MC –> 299||C –> H|
|wr = T(H –> 184)||–|
|ha = T(H –> 260)||–|
|wa = M(H –> 220)||–|
|lg = TC –> 303||C –> M|
|ft = M(H –> 220)||–|
|f1 = T(M –> 129)||–|
|f2 = T(H –> 166)||H –> M|
|t1 = n/a||–|
|t2 = (H –> 382)||–|
|mh = MC –> 92||C –> T|
|oh = T(M –> 150)||–|
And really…that’s it. That’s all there is to reforging by hand. No complicated math; just adding and subtracting goats and cabbages.