Somehow I think the stacking trinkets are broken.

A toggle for whether or not to recommend epic gems might be nice. Realistically most of us aren't going to be running all epic gems for a while here, so the ability to optimize our rare gems in the meantime would be useful.

when i have my T1 daggers equipped shadowcraft is showing me at 41345.7 dps but when i switch out my mainhand to morningstar of heroic will, shadowcraft says my dps goes to 41747.4 which means that the T1 daggers are not actually better than a 397 combat weapon in the mainhand unless something is not being modelled correctly.

after spending a good 10 minutes with each mainhand doing a complete rotation at the training dummy running recount I was actually finding similar results as well with the difference in dps between the 2 mainhands to actually be further apart, around 1400 dps different (i can assume since the mainhand makes ss and rs hit a flat 45% harder that the better your gear / the more raid buffs you have will make the 2 closer in value). That being said would this mean that until a certain gear level it would be better to use a 397 combat mainhand over the dagger until t2 of the daggers can be obtained or until you have obtained the level of gear that would make the daggers better?

Everything that I've tested/seen/worked out has shown that the daggers won't be best in slot until T3. Poison damage is lower, as expected, and for some reason main gauche damage is lower as well - not just on a per-hit basis but overall.

Main Gauche damage will be lower - while you're getting more procs from autoattacks (making up for the lower damage), the same is not true for your instant attacks - you have the same number of instant attacks as before, and each such proc does less damage, so...

This is to some extent mitigated by the higher number of autoattack procs giving you more energy regen, so the overall effect on your damage done is not as large as it appears, but its definitely true that the portion of your damage done by Main Gauche is lower than it was.

This is, I should note, related to the reason why your poison damage goes down - you have the same number of special attacks, but each has a lower chance to proc poison, so you have fewer poison procs.

I also have a sneaking suspicion that the daggers might not be as compelling an option for Combat until the final stage, because my sense is that the proc rate of the set bonus is balanced with the legendary proc in mind rather than the straight stacking agility of the lower-stage daggers. The free finishers of the legendary have an additional benefit for Combat due to Restless Blades, so its plausible that the legendary proc needs to happen less often to keep them balanced across specs; this would explain why the proc rate for Combat is so much lower than for the other specs, and in that event its plausible that the lower-tier daggers aren't as good for Combat as the other specs.

To be clear: I'm not saying this is absolutely the case - I haven't had time to fully analyze the procs and weapons across specs - but from what I've heard about the intended balance and what we're seeing so far, its at least plausible that the early-stage daggers might be less advantageous for Combat than they are for the other specs.

Thanks for the info - I had not even considered instant attacks when I noticed the lower MG damage.

On the set bonus: In the discussions I had with a fellow rogue on my server, we agreed that the proc rate on the T1 daggers (and, I had figured, most likely the T2 as well) was indeed balanced around the T3 proc rather than just the agility stacking of the T1/T2 daggers. He suggested, given this, that combat might not want the T1 dagger even in the case that the weapon they replace it with was the same item level, and, likely of No'kaled as well even given its lack of a socket and secondary stats.

Additionally, he suggested that a heroic Morningstar of Heroic Will would be better than the T2 mainhand. I disagreed given that the T2 set bonus is quite a lot better.

Given, again, that no one has even seen these items yet let alone done any sort of testing on them, I am hesitant to say with any amount of certainty which might be a better option for combat in terms of T2 daggers vs other heroic DS options. It seems, given the fairly low amount of agility and quite low stack rate of the T1 proc that a comparable non-heroic option might be better than Fear, but at this point all I can do is wait for more intelligent people than I to figure out these details.

I was wondering if we could start using something like DPS equivalent points instead of Attack Power EP's. I was thinking about it and I think you get a lot more out of knowing how much DPS each stat is worth. So instead of knowing that 1 agi is worth 2.68 ap, we could see that 1 agi is worth like 2.34 dps or something. This would be so much more useful.

If we did this then using the EP's we could just figure out ourselves how much of an increase 50 haste is worth or whatever. You could also compare the values between specs and it would be meaningful, unlike right now because each spec gets a different amount of benefit from AP so it's impossible to tell just by looking at the values. I would love to know how much differently stats like crit and whatnot affect each spec's dps. It would give insight into scaling as well.

Also it's kind of hard to explain EP's to a new player. If I say, "1 agi is worth 2.68 ap!" They would be like, "What?" But, if I say "1 agi is worth 2.34 dps!" They would be like, "Wow..." lol funny example.

I was just wondering what you guys thought about this. If nothing else maybe add the conversion somewhere for attack power and then I could do the conversions to DPS myself.

That is a better solution, but the way the app's currently set up, it wouldn't be trivial to do. I've been playing with backbone.js for work and have the itch to rewrite the whole thing in it, but I don't think that'd be a good idea just yet.

Regarding the talent caches, would it be okay if I just did the cache reset before/after doing talent comparisons?

Debatable. I actually discussed this recently here - the short version is that by normalizing gear comparisons to EP we gain some nice properties in terms of how values respond to gear scaling and DPS buffs, and aren't really any less intuitive as a unit. Its important to keep in mind that when we're talking about theoretical DPS (as reported by ShadowCraft and its ilk) its just as artificial a unit as EP - it doesn't correlate to actual DPS on any real fight. In some ways I wish we didn't need to report a top-level DPS estimate in ShadowCraft, as I think it tends to make people focus too much on the actual DPS estimate and not the gear comparison - which is what the tool is actually designed to do. So while it might seem more intuitive to talk in terms of DPS estimates rather than EP, in practice I don't think it has any real advantages and some very important disadvantages to doing so.

The biggest reason I wanted to have the EP's in terms of DPS was so I could see the differences in value of stats between specs. But, as it is now AP for combat is not worth the same as AP for assassination or subtlety, so it's useless to compare EP's between specs at the moment, and there is no obvious way to compare the value of stats between specs right now that I know of.

Sounds reasonable, and it may solve the problem, I'll check it as soon as possible. The issue is minor though: right after the killing spree talent is computed (is taken out of the spec and put back on, that is), the dps_breakdown (and consequently the dps and ep methods) would report a different output in the order of a 10th of dps; I don't think I'd ever encountered it if I weren't working on my java rewritten engine. Disabling the cache solved the issue and I forgot a bit about it; so, it's worth knowing that something funky is going on (and it's on my todo list to investigate), but I'd say it's safe to be published on the web-app.


Regarding t13 set bonuses: they were hastly coded a while ago (right when they went public). Chances are that they'll get rewritten: if not by me (I'm having a bit of trouble with my computer) by someone else.