help with simcraft results

[diablo85]

just simmed my gear for shadow spec as i picked up the shoulders off of baleroc last night and it came up with the results of crit being better than mastery...should i follow this and plug the stat weights into simcraft? im wondering because i thought mastery was supposed to be a much better stat than crit for us even at higher gear levels?

armory link



and with those stat weights plugged into wowreforge

[zakaluka]

[QUOTE=diablo85;13916515]came up with the results of crit being better than mastery...should i follow this and plug the stat weights into simcraft?[/URL]

I think you meant to ask: should I reforge with these stat weights? (you asked if you should sim against them) First: I was somewhat surprised to see that simcraft rates MST so much lower than CRIT. Trinket choice plays into this strongly: NF will lower MST's scale factor, DMC will lower crit's scale factor, and VPLC might raise the value of crit substantially (but I'm not entirely sure).

The answer to your question is, probably not. Scale factors can answer the question 'which stat is the most valuable for me to add from this point?' The question you're asking, 'How should I reforge?' is a little more complicated.

Here's why - when calculating scale factors, simcraft simply re-runs the whole simulation at +300/-300, and projects the value of that stat inside that given range, with all other stats held constant. For instance, if you have 1100 mastery and simcraft gives MST a scale factor of 1.5, you could reliably assume that reaching 1400 MST (without changing anything else) would add 450 dps.

What you cannot do with this is anything more complicated, like a complete reforge based on scale factors at your gearpoint. Why: Suppose in your situation you forge off 500 MST and forge on 500 CRIT. (1) scale factors are based on a delta of 300. Since you're not close to that delta, the scale factors are at best a vague approximation. (2) the scale factor for CRIT is extremely dependent on your HST/MST ratings (and cast behavior/which trinkets you're wearing). I.E. scale factors don't always do what you expect them to do, especially if you try to use them to guess what will happen after a change in multiple stats (like a big reforge).

shorter answer: no, you can't. Forging off mastery will raise the value of mastery, and lower the value of crit. Forging on crit will lower the value of crit, and raise the value of mastery. Since crit/mst have pretty similar values, stacking one will devalue it until they're even.

The best you can do is ask wowreforge to calculate a reforge based on these new stat weights, then manually enter that reforge into simcraft and run the sim again. The results might be not quite what you would expect, and will definitely give you a different set of scale factors. If you repeat this process several times there's a good chance you'll reach an optimal solution. This so-called 'optimal solution' could be recognized by: (1) haste above (or equal to) your highest achievable plateau (2) a: scale factors for mst/crit being very close to equal, b: OR the higher of the two still has a significantly higher scale factor (I'd expect a to happen first but I could be wrong).

[Yvaelle]

That kind of shift in crit/mastery is normal, and Zakaluka already outlined the problems and how to respond very well so I'll skip that (but endorse his response). What is more shocking to me is how low your haste is worth - that seems odd to me - very odd. How many simulations did you run? I'm guessing 100-300, and would recommend not making decisions on anything less than 1000 (and if your computer can handle it or you are going away from your computer for a bit, 10000). If you did only do 100-300 simulations, that would also alter your crit/mastery obviously. Alternately you could just have a setup that is really decreasing haste value to be comparable to crit/mastery, but that doesnt seem right to me because while I remember discussing that point a few months back on shadowpriest.com, the normalized weights were much higher than the 0.45-0.5 range you are seeing them at.

[eErike]

That kind of shift in crit/mastery is normal, and Zakaluka already outlined the problems and how to respond very well so I'll skip that (but endorse his response). What is more shocking to me is how low your haste is worth - that seems odd to me - very odd. How many simulations did you run? I'm guessing 100-300, and would recommend not making decisions on anything less than 1000 (and if your computer can handle it or you are going away from your computer for a bit, 10000). If you did only do 100-300 simulations, that would also alter your crit/mastery obviously. Alternately you could just have a setup that is really decreasing haste value to be comparable to crit/mastery, but that doesnt seem right to me because while I remember discussing that point a few months back on shadowpriest.com, the normalized weights were much higher than the 0.45-0.5 range you are seeing them at.

This, though I would say that you never should gear after anything but sims with 25k + litterations

[diablo85]

odd...i nearly forgot about this thread. never happens. @ eErike/Skill, i never run simcraft with any less than 10k iterations.

@zakaluka, my trinket choice is normal NF, and DMC:V as i dont yet have access to the VPLC as i'm only used as shadow for a few fights and our warlock has yet to get the rag trinket. also, i have access to the rag gloves for shadow (not using them for disc as it would break 4 piece)

i've also just been using the standard simcraft profile that simcraft generates for you when importing my character in from the armory. is there one that i should use as a better alternative?


25k iterations with the 378 shoulders equipped.

[Zuziza]

Another question: which fight setting are you using? Patchwerk, HS, one of the new ones, etc...

[diablo85]

Patchwerk.

[zakaluka]

Haste can show as low value if:

1) your haste is already really high (above VT+2 and out of range of DP+2). I see 2360 haste on your armory, so it's not entirely out of the realm of possibility. See Kilee's guide for a comparison, although I'm still rather curious how she generated those graphs. Simcraft's plots look a lot different than hers, and I like the way kilee's graphs read a lot better than the way mine do ><
2) your action list is wrong. Have you made any changes to the default? As far as I know the default shadow actionlist is pretty great in simc. Only a few minor arguments can be made over it, like whether to wait for orbs on MB. If I recall correctly, the changes you CAN make have very little effect. But, if you tweaked the actionlist incorrectly you could get really strange results. Post a link to the whole sim if you can?

Reading your last response a little closer, looks like you're just using the default. Maybe trust what it says . I'm pretty sure the last time Kilee updated her graphs, her reference gear set was a bit lower level than yours.

[hikamiro]

Since nobody else is asking... Are you getting DI or FM?

Crit has always been significantly better then people give it credit for, it helps procs shadow orbs and lowers your shadowfiend cooldown, on top of doubling damage. I know for me personally FM accounts for around 1k dps, but also devalues the rest of my crit fairly heavily. So if you are not getting FM that might be a reason for it being valued so high. I do not see any reason why crit would be valued higher then haste though, unless your crit is EXTREMELY low. (It would have to be low enough to the point that you can not consistently get shadow orbs).

As others have noted, when you do reforge based on this data, make sure you do it slowly in small chunks at a time, and then re-sim; You want to stop moving mastery into crit when they become equal.

[diablo85]

Since nobody else is asking... Are you getting DI or FM?

Crit has always been significantly better then people give it credit for, it helps procs shadow orbs and lowers your shadowfiend cooldown, on top of doubling damage. I know for me personally FM accounts for around 1k dps, but also devalues the rest of my crit fairly heavily. So if you are not getting FM that might be a reason for it being valued so high. I do not see any reason why crit would be valued higher then haste though, unless your crit is EXTREMELY low. (It would have to be low enough to the point that you can not consistently get shadow orbs).

As others have noted, when you do reforge based on this data, make sure you do it slowly in small chunks at a time, and then re-sim; You want to stop moving mastery into crit when they become equal.

For fights that i do go shadow, i do get DI but no fm.
@ zaka, i ran a couple other sims last night (with the rag hands, bael shoulders) with and without dark intent toggled on and found that only with di toggled off was i getting normalized stat weights of roughly the same as found in kilee's guide. In fact, i think that in one of the
sims i ran, crit even out weighed haste (cant post pics as im at work right now, will post when i get home)

[Yvaelle]

For fights that i do go shadow, i do get DI but no fm.
@ zaka, i ran a couple other sims last night (with the rag hands, bael shoulders) with and without dark intent toggled on and found that only with di toggled off was i getting normalized stat weights of roughly the same as found in kilee's guide. In fact, i think that in one of the
sims i ran, crit even out weighed haste (cant post pics as im at work right now, will post when i get home)

I'd like to see those results when you get a chance!

Edit: Oh also! I wouldn't say crit is bad - I even advocated balancing crit/haste in Tier 11 for spriests with mana problems as any use of Shadow Word: Death for mana except while moving is a much bigger dps loss than what you gain out of the extra haste - I'd be very happy to see a rise in crit's relative value again and haven't played with simcraft much since the start of the tier so I don't know if it's happening and I just didn't notice!

[diablo85]

I'd like to see those results when you get a chance!

Edit: Oh also! I wouldn't say crit is bad - I even advocated balancing crit/haste in Tier 11 for spriests with mana problems as any use of Shadow Word: Death for mana except while moving is a much bigger dps loss than what you gain out of the extra haste - I'd be very happy to see a rise in crit's relative value again and haven't played with simcraft much since the start of the tier so I don't know if it's happening and I just didn't notice!

running the sims now since i shut down my computer last night without posting them. should have them up shortly

---------- Post added 2011-10-30 at 08:50 PM ----------


with DI

no DI

with DI, no FM

no DI, no FM

[Worshaka]

Since nobody else is asking... Are you getting DI or FM?

Crit has always been significantly better then people give it credit for, it helps procs shadow orbs and lowers your shadowfiend cooldown, on top of doubling damage.

* Crit doesn't affect shadow orb procs, if anything haste increases shadow orb procs by increasing the amount of SW:P & MF ticks you get per second. Whether those ticks crit or not has no bearing on orb procs.

* Yes it lowers your fiend CD, but no more than haste will. 1% crit means 10 sec lower fiend CD per 100 casts, this is the same rate that haste gives you.

* Crit is inherritly weaker than haste in terms of increasing your dps. For starters you need about 179 crit rating for 1% compared to 128 for haste. Secondly crit rates are additive, so if you have 5% crit from some buff and 10% from gear you get 15% crit. Haste is derived multiplicatively, so if you had a 5% haste buff and 10% from gear you actually get 15.5% haste total.

Just to show the power of that, assuming DI you get 1.03 * 1.05 * 1.03 * haste rating. Those figures correspond to DI, darkness & shadow form. That means for each 1% of haste rating you add to your gear you are really getting 1.1139% haste. If you want to discount DI the figure is 1.0815% haste.

If you normalise crit rating and compare it to the true effect of haste rating you get:

* 1.559462 crit rating = 1 haste rating with DI.
* 1.514102 crit rating = 1 haste rating without DI.

That means you need over 50% more crit rating to have it behave the same as haste.

Keep in mind that 1% haste has the exact same effect as 1% crit, all be it in different ways. If you cast a spell that always hit for 1 damage 100 times you would get 100 damage. Now add 1% crit you now get 101 damage.

1% haste means that in 100 casts you can now cast 101 of them in the same time. Casting 101 spells now means you do 101 damage.

---------- Post added 2011-10-31 at 10:35 AM ----------

I've also left off BL/Heroism but you should factor in some of it's uptime but that uptime is variable per encounter so is hard to do definitively. Suffice to say the % haste gain from BL/Heroism is multiplied on top of your other haste values.

[hikamiro]

* Crit doesn't affect shadow orb procs, if anything haste increases shadow orb procs by increasing the amount of SW:P & MF ticks you get per second. Whether those ticks crit or not has no bearing on orb procs.

* Yes it lowers your fiend CD, but no more than haste will. 1% crit means 10 sec lower fiend CD per 100 casts, this is the same rate that haste gives you.

* Crit is inherritly weaker than haste in terms of increasing your dps. For starters you need about 179 crit rating for 1% compared to 128 for haste. Secondly crit rates are additive, so if you have 5% crit from some buff and 10% from gear you get 15% crit. Haste is derived multiplicatively, so if you had a 5% haste buff and 10% from gear you actually get 15.5% haste total.

Just to show the power of that, assuming DI you get 1.03 * 1.05 * 1.03 * haste rating. Those figures correspond to DI, darkness & shadow form. That means for each 1% of haste rating you add to your gear you are really getting 1.1139% haste. If you want to discount DI the figure is 1.0815% haste.

If you normalise crit rating and compare it to the true effect of haste rating you get:

* 1.559462 crit rating = 1 haste rating with DI.
* 1.514102 crit rating = 1 haste rating without DI.

That means you need over 50% more crit rating to have it behave the same as haste.

Keep in mind that 1% haste has the exact same effect as 1% crit, all be it in different ways. If you cast a spell that always hit for 1 damage 100 times you would get 100 damage. Now add 1% crit you now get 101 damage.

1% haste means that in 100 casts you can now cast 101 of them in the same time. Casting 101 spells now means you do 101 damage.

---------- Post added 2011-10-31 at 10:35 AM ----------

I've also left off BL/Heroism but you should factor in some of it's uptime but that uptime is variable per encounter so is hard to do definitively. Suffice to say the % haste gain from BL/Heroism is multiplied on top of your other haste values.

I was never comparing crit and haste, i was comparing crit and mastery. Also apparently i read the Harnessed Shadows talent wrong, i assumed it meant when you critically hit, but it means when you are critically hit

[Zuziza]

See Kilee's guide for a comparison, although I'm still rather curious how she generated those graphs. Simcraft's plots look a lot different than hers, and I like the way kilee's graphs read a lot better than the way mine do ><

Kilee makes the list by doing an average of multiple the Simcraft settings. The reasoning is that there is no pure Patchwerk, HS, or any of the other settings as real raid bosses in Cata. It reflects as if a fight was going to have equal parts of each setting, which is also unrealistic, but it's still a truer estimation than looking at one alone or only averaging Patchwerk and HS.

Depending on your role for Beth, you're AOE'ing or multidotting or going up top for interrupted single target, then the second half is Patchwerk. For Shannox (reg) you're multidotting, and for Shannox H you're more like HS than Patchwerk due to movement, but neither is perfectly applicable. For Ryolith you're multidotting legs while burning adds. If you're flying for Alysrazor, it's a 3D Patchwerk with short interrupts, but if you're on the ground, it's running around a lot while nuking down adds as they spawn. I could go on, but the point is just this: no single type of sim is going to accurately show how you should gear for all of Firelands.

Sure, there will be some people who have the option to have a gearset and minmax for every single fight, but that's not who the guide is for. Those are the types who usually don't need a guide to reference how to play in the current patch.

All that said, if you want to figure out a good Sim for a particular fight, game the weights of each type to reflect how the fight goes. However, Kilee's strategy for overall gives you a good idea of how to approach Firelands in general. Patchwerk only Sims are highly misleading because they artificially raise the value of mastery. Any movement reduces the number of MF ticks, so your orb procs go down, and unless you can stop and cast MB on CD every time, your MB damage is much higher in the sim than it will be live.

[Worshaka]

I was never comparing crit and haste, i was comparing crit and mastery. Also apparently i read the Harnessed Shadows talent wrong, i assumed it meant when you critically hit, but it means when you are critically hit

You were making a point that, "Crit has always been significantly better then people give it credit for." I was pointing out that crit isn't magically better than what people think it is. Unless crit has some secondary benefit (like increasing a resource like it does for warriors) it will always be a very average stat. The secondary effect for spriests is lower fiend CD, however fiend doesn't contribute nearly enough dps, nor does 1% crit affect fiend up time by any significant amounts.

[hikamiro]

You were making a point that, "Crit has always been significantly better then people give it credit for." I was pointing out that crit isn't magically better than what people think it is. Unless crit has some secondary benefit (like increasing a resource like it does for warriors) it will always be a very average stat. The secondary effect for spriests is lower fiend CD, however fiend doesn't contribute nearly enough dps, nor does 1% crit affect fiend up time by any significant amounts.

Crit IS significantly better then what people give it credit for. People seem to think that mastery is miles above crit on every fight, and this just isn't true especially if you have a lightning gen~

[Worshaka]

Crit IS significantly better then what people give it credit for. People seem to think that mastery is miles above crit on every fight, and this just isn't true especially if you have a lightning gen~

Can't say I agree but your definition is relative to mastery and there is no doubt crit and mastery are closer together than crit and haste. However crit is always subject to rng, just because you increase your crit rate by 1% doesn't guarantee that you crit 1% more on a given encounter. It just means on average over all encounters you will crit 1% more. Whereas mastery is a guranteed benefit, you will do x% more damage without argument.

Mastery also increases our damage by 1.45% per 179 rating. It doesn't affect SW and you are likely to have the occasionaly MF not benefit from it when bad luck with orbs. Still mastery is going to increase our damage by well over 1%, 1.2% is probably a conservative estimate. Crit only ever increases our damage by 1% when you move from 0% crit to 1% crit... as you get more crit the percentage of damage increase declines. Moving from 20% to 21% crit increases damage by 0.833%, so even with the trinket and factoring in more fiend uptime, it's not going to increase that value by 0.4%.

[hikamiro]

Can't say I agree but your definition is relative to mastery and there is no doubt crit and mastery are closer together than crit and haste. However crit is always subject to rng, just because you increase your crit rate by 1% doesn't guarantee that you crit 1% more on a given encounter. It just means on average over all encounters you will crit 1% more. Whereas mastery is a guranteed benefit, you will do x% more damage without argument.

Mastery also increases our damage by 1.45% per 179 rating. It doesn't affect SW and you are likely to have the occasionaly MF not benefit from it when bad luck with orbs. Still mastery is going to increase our damage by well over 1%, 1.2% is probably a conservative estimate. Crit only ever increases our damage by 1% when you move from 0% crit to 1% crit... as you get more crit the percentage of damage increase declines. Moving from 20% to 21% crit increases damage by 0.833%, so even with the trinket and factoring in more fiend uptime, it's not going to increase that value by 0.4%.

I disagree with crit being RNG for dps; i'd actually say the opposite, crit DECREASES rng. If you have 1% crit there is still that 1 in a billion chance that you crit with every spell. It makes your dps range very wide. If you gain more crit then that margin gets significantly less wide with each additional point.

Crit is also significantly better in any sort of movement fight where you cannot keep up ES at all times because of lack of mindflays to proc shadow orbs, and lack of mindblast casts while moving.

[Worshaka]

I've yet to experience an encounter that is so movement intensive that my ES uptime is significantly affected. What you say is correct in theory but I don't believe it applies to any practical situation in game.