TBC Shadow Priest 101: How to Melt Faces Effectively - Page 20

Lazare · 01/01/08, 9:18 AM

You're wondering why more priests don't use the Bracing Earthstorm? Um...because the "best" passive meta is worthless compared to the actual best meta? What's to wonder about? They don't use it because it is a suboptimal. See also: Why more priests don't strip naked for boss fights, gem with crit gems, or get agi enchants.

This whole "well, the theorycraft says it's not as good, but it lets me focus on the stuff that's more important" is nuts. What's more important than doing more damage? Now, if the issue here is that as a practical matter you find that you personally net more damage using the Bracing gem, then that's a completely different issue. But if the issue is "why don't other priests use a crappy gem?", I think you've already answered your own question.

• Snowy · 01/01/08, 2:32 PM

Originally Posted by Redux

It's the best passive Meta Gem for a shadow priest. Why do people only want the half cast one? I've been running with Bracing for several months now, and have had no regrets. Just wondering why more priests don't do the same.

Okay, since you asked for it... because it's a crappy gem? +9 damage and a largely irrelevant -threat at that point in the game, along with an ugly gem requirement. More yellow than blue? When the T6 helm has a blue socket?

"More priests" don't have too much of a problem making use of the half-cast gem, that's why, and it's hands down superior to the best passive meta.

DeeNogger · 01/01/08, 10:51 PM

Honestly, the only fight its worthwhile on is Void Reaver. Now I *did* use that gem when doing the beginning parts of T5 content, and therefore it was helpful on Void Reaver. However, that was mostly because of my ignorance, not because the gem was any good.

Amonra · 01/02/08, 6:46 AM

Originally Posted by LucidityAxel

I am not anti-MSD, or even anti-proc; it's just another tool in the toolbox. I think it's important to understand that the effectiveness of this particular tool can be greatly curtailed whenever I am frequently forced to move or am otherwise interrupted a lot by boss mechanics, and starts to really shine whenever I can stay in my full-burn spell cycle for large, uninterrupted chunks of time.

Actually I would suspect that the MSD actually gets better (compared to other meta gems) as you start spending time unable to cast. It is only wasted in the rare circumstance that the interruption happens after the MSD procs but before you can make use of the proc. If the interruption happens at any other time then the internal cooldown on the MSD will keep running, effectively increasing the proc rate based on actual cast time.

Don't forget that the CSD is completely useless while not casting - no casting means no chance to crit. Although I guess if the interruption is very short then it would help with MB cooldowns (but if the interruption is this short, the MSD proc duration will probably be long enough to still cast a double speed MF).

The BED is also almost completely useless while not casting - the only impact it will have is that dot's will continue to tick at a slightly higher amount.

Balkoth · 01/02/08, 1:30 PM

Originally Posted by DeeNogger

Honestly, the only fight its worthwhile on is Void Reaver. Now I *did* use that gem when doing the beginning parts of T5 content, and therefore it was helpful on Void Reaver. However, that was mostly because of my ignorance, not because the gem was any good.

I think it could be useful on Bloodboil and RotL phase 2/3, though. Perhaps Illidan's Flames as well, but you're talking maybe three fights out of 13 at the t6 level. Not really worth it.

Redux · 01/03/08, 7:43 AM

Originally Posted by Snowy

Okay, since you asked for it... because it's a crappy gem? +9 damage and a largely irrelevant -threat at that point in the game, along with an ugly gem requirement. More yellow than blue? When the T6 helm has a blue socket?

"More priests" don't have too much of a problem making use of the half-cast gem, that's why, and it's hands down superior to the best passive meta.

I don't disagree with you, but people keep talking about the +crit gem. Isn't this meta hands down better than the crit one? It comes in a clear second place from what I can tell.

Taurmindo · 01/03/08, 10:51 AM

I was thinking about the recommended enchant for the cloak posted by the OP.

Why is the -2% threat recommended? Well, I do see that most cloak enchants don't make any major differences, but this one is making an exceptional minor one. 2% in itself is not much, and when added up with the shadowpriest talent to reduce threat it gets even smaller.

75*.98 = .735, e.g. down to only 1.5%

Additionally, if using the blessing of salvation, which most spriests would do in raids I suppose, an additional 0.7 multiplier lowers the effect of the enchant even further:

without enchant: (.7*.75 = .525) 47.5%
with enchant: (.7*.75*.98 = .5145) 48.55%

Thus the effect of the cloak is reduced to only 1.05% threat reduction in most raid enviroments. Is that truly the best enchant for spriests?

vpolden · 01/03/08, 11:06 AM

Well, the alternatives to Subtlety aren't exactly awesome. Spell penetration won't make a difference in PvE, while 120 armor or 12 agi / dodge rating won't be very useful unless you're planning to tank stuff.

tedv · 01/03/08, 12:04 PM

Originally Posted by Taurmindo

I was thinking about the recommended enchant for the cloak posted by the OP.

Why is the -2% threat recommended? Well, I do see that most cloak enchants don't make any major differences, but this one is making an exceptional minor one. 2% in itself is not much, and when added up with the shadowpriest talent to reduce threat it gets even smaller.

75*.98 = .735, e.g. down to only 1.5%

Additionally, if using the blessing of salvation, which most spriests would do in raids I suppose, an additional 0.7 multiplier lowers the effect of the enchant even further:

without enchant: (.7*.75 = .525) 47.5%
with enchant: (.7*.75*.98 = .5145) 48.55%

Thus the effect of the cloak is reduced to only 1.05% threat reduction in most raid enviroments. Is that truly the best enchant for spriests?

This is a common misconception about how threat works. Often people assume that because the final factor is reduced, they are getting less of a benefit from it. In reality, multiplying threat factors maintains the proportion of extra threat that must be generated to pull aggro.

For example, consider the effect of an additional 50% reduction in threat. This ability will let you generate twice as much threat before pulling aggro, no matter what other modifiers there are. Suppose the threat ceiling is 10,000 with no threat reduction. Look at the effect of a 50% reduction applied to a variety of starting threat modifiers. (The actual ceiling is 10000 / modifier.)

Start:           Half:   
100% (10,000) => 50% (20,000)
 75% (13,333) => 37% (16,666)
 20% (50,000) => 10% (100,000)

A 2% reduction in threat will ALWAYS be a 2% increase in damage you can deal before pulling aggro. This is why Tranquil air totem is good, by the way, even though many people think it's bad for the same reason. ("It's only a 9% reduction!")

Galanna · 01/03/08, 12:06 PM

Originally Posted by Taurmindo

Thus the effect of the cloak is reduced to only 1.05% threat reduction in most raid enviroments. Is that truly the best enchant for spriests?

It's 1.05% of your 'base threat', but this means nothing. It's still 2% of your 'effective threat'.

Kalman · 01/03/08, 12:19 PM

Originally Posted by Redux

I don't disagree with you, but people keep talking about the +crit gem. Isn't this meta hands down better than the crit one? It comes in a clear second place from what I can tell.

Yeah, I'd put it as solidly better than the CSD. I actually do intend to get one for my T5 helm (which has a gemming requirement that supports its use)... once I get a T6 helm. Basically I'd only use it on Bloodboil, though, because it's incredibly easy to threatcap on that fight. On anything else, the MSD is solidly better.

Caligula · 01/03/08, 12:49 PM

Is anyone even threat capped on any fight except Gurtogg anymore? I stopped using 2% on cloak awhile back because I stopped seeing any benefit save that single fight. At least spell penetration helps on trash. I might go back when we see the new encounters in Sunwell but until then I can't really justify enchanting my gear around a single encounter, especially when farming.

Balkoth · 01/03/08, 1:04 PM

Well, that depends on whether I have a Tranquil Air totem. I can still threat cap myself on Naj'entus, Bloodboil, EoS phase 2/3, and Illidan Elementals extremely easily with Wrath of Air. I can typically do the same for Kaz'rogal and Gorefiend as well.

I figured that since trash isn't an issue, I'd prefer the help from less threat overall, even if I don't always use it.

Caligula · 01/03/08, 1:15 PM

Originally Posted by Balkoth

Well, that depends on whether I have a Tranquil Air totem. I can still threat cap myself on Naj'entus, Bloodboil, EoS phase 2/3, and Illidan Elementals extremely easily with Wrath of Air. I can typically do the same for Kaz'rogal and Gorefiend as well.

I figured that since trash isn't an issue, I'd prefer the help from less threat overall, even if I don't always use it.

I suppose the difference there is that you have Imp VE and I dropped that talent the first time I saw Naj

Diff'rent strokes.

ObservingLife · 01/03/08, 3:44 PM

Originally Posted by Thezilch

While it's true that the MB and SW:P cooldowns aren't head-to-head often, SW:P will always do more DPS per global, if it runs the full duration (bosses).

Without 2pc T6, SW:P will tick 8 times. If we take an average tick of 600 (yes, this is quite low to send home the message), we can expect 8 * 600 for 4800 damage over a 1.5 second global (3200 DPS). Again, this is close to the lowest end of SW:P's DPS capability. I can personally expect closer to 8100 damage done (5400 DPS)

Without even going into the math, with 4pc T6 and Imp. SB proc, I'm looking at mid-4800 crits from MB. That is, if the stars are all aligned, MB will _match_ the lowest expectations from SW:P. Averaged over a fight, the MB GCD falls behind.

I was thinking about this, and the logic behind it seemed flawed so I started running through it. The premise in the above math is the full damage per cast of SW:P vs Mind blast. This would be useful if recasting SW:P did all of its damage, but it doesn't, unless it has run its course.

So instead, I propose a differential comparison. The situations where this is relevant is: option to cast pain or mind blast, and options to cast pain or mind blast, with VT ending between 1.5 and 3 seconds from now. This premise operates on attempting maximum VT uptime.

A note: in the quoted text, a 600 pain tick was mentioned as being low, but my calculated pain ticks are comming in at about 570. Taking into account his gear (1300 shadow damage, 29% crit), my pain ticks land at 639. My numbers only account for priest specific talents: 25% increased damage from talents, and 15% increased damage to target.

In the first instance, from time zero, if you cast pain, it will go up immediately, and blast will be cast at 3 seconds. If you cast blast, then pain, blast is cast at 1.5 seconds, and so will pain (while latency is an issue, it will be addressed later). An enginneered spreadsheet came up with these DPS values at 25% crit (assuming talents) and 1000 shadow damage:

DPS:
Flay: 536
Blast: 1251
Death: 1128
Pain: 187

Conceptually, if you cast pain then blast, you are trading 1.5 seconds of blast dps for 1.5 seconds of pain + flay dps, due to the cooldown on blast. As blast DPS is higher than Flay + Pain combined (1251-536-187=528dps, or 792 damage), you will do more damage casting blast and then pain.

In the second situation, you would cast pain, then renew VT, then cast mindblast, or you would cast blast, renew VT, then cast pain. You trade 1.5 seconds of mind blast dps for 1.5 seconds of mind flay dps and 3 seconds of pain dps. An additional 1.5 seconds of pain dps will still result in the loss of 512 damage by casting pain first.

A similar comparison using death instead of blast results in 607 damage lost in the first example, or 327 lost in the VT example.

Additional crit over the model obviously favors blast/death more, additional + damage has a minimal effect on the damage difference (at 1300 damage its 832 damage lost, or 513 in the VT example, compared to 792 and 512). This makes sense because in a pure dps sense, blast gets more from + damage than pain, as pain gets 1.1/18 = 6% +damage per second, while pain gets .4286/1.5=28.6% +damage/second (ignoring crit, which skews it slightly higher), and flay gets .59/3=19.7% +damage per second.

This model assumes mana is a non-issue. The optimization here also exists in the realm of cancelling mind flays to mind blast or shadow word: death, which needs to be executed well to actually provide a benefit. As this situation could only come up an absolute maximum of every 24 seconds (assuming no 2pT6 and talented pain), and is more likely to come up not more than every couple minutes, the dps increase is extremely small (Max end of 34.7 dps increase, much more likely in the range of 3-7 DPS.)

It does, however, imply a change in the "priority" list for spellcasting:

VT
MB
SW:D
SW:P
MF

Base damage has the 25% bonus, and the 15% target bonus, applied to it. I'm not sure in which forms percent bonuses are multiplicative, but it does not change the conclusion in this case.

Dmg Cast Mana +Dmg Coefficient Base Dmg Crit % Avg Dmg DPS
Flay 528 3 230 1000 0.59 1607 0 1607 536
Blast 731.5 1.5 450 1000 0.4286 1668 0.25 1877 1251
Death 618 1.5 309 1000 0.4286 1504 0.25 1692 1128
Pain 1236 18 575 1000 1.1 3358 0 3358 187

DPS difference 1.5 second delay 3 second delay
528 792 511.5
405 607.5 327

I can't post the actual spreadsheet currently, but the math is there...I'd love any corrections, or any additional thoughts on any of the logic or math.

CaseyTheRetard · 01/04/08, 5:51 PM

Originally Posted by ObservingLife

An enginneered spreadsheet came up with these DPS values at 25% crit (assuming talents) and 1000 shadow damage:

DPS:
Flay: 536
Blast: 1251
Death: 1128
Pain: 187

1128 DPS over a 12 second cooldown period implies that SW

is hitting for 13,536 - an impressive number, to say the least. I think your numbers for Blast and Death here are in fact damage per cast time, and not actual damage per second.

ObservingLife · 01/04/08, 6:53 PM

The numbers are obviously not all damage per cast time, as the SW:P would be much higher. The numbers are the DPS that that spell has when you're casting it, as your effective DPS is based on cast time when it comes to straight cast spells as opposed to dots, and this assumes a maximized VT uptime, so the only dot is SW:P.

For instance: if you were to chain cast fireball, and it had a dps value of 1600, you would do 1600 dps. if you're looking at an individual time period of the single cast time of a fireball, your dps is still 1600, and effectively the spell's individual dps is 1600.

While you cannot chain cast the spells being referenced, in the individual timeframe you are making choices of which spells to cast, which affects your ability to use these spells in the future. Thus, a comparison of the DPS done in that individual timeframe by the possible spells that would be used provides information of what spells are "better" for a priority system. Postponing casting spells now that have a cooldown and are reset will effectively lower your dps if those cooldowned spells will do more damage than the non-cooldown spells you are casting instead (this can be seen anecdotally by JUST using VT, SW:P, and MF: the more MB you add, the more your DPS goes up, as MB is a higher DPS spell)

A bit of this premise keeps bugging me, though: while in theory you are "replacing" 1.5 seconds of MB time with additional MF time, you are only technically pushing the MB time 1.5 seconds into the future. As all fights are of finite length, unless you "push off" a MB off the end of the theoretical "fight length", you have not actually lowered your dps, only switched around the position of your spells. It would take a maximum of 4 iterations to "push off" a MB and thus lose the DPS (assuming 4/5 or 5/5 imp MB), and an additional 4 after the first one was lost for each additional DPS loss. I don't feel confident enough in this analysis to "hang my hat on it", but I would think then that the "DPS increase" numbers would be variable, but on average could be divided by 4, for a max increase of 5 DPS and a more likely increase of 1-2 DPS. Any insight on accurate quantification of this would be welcomed and interesting.

Caligula · 01/04/08, 7:39 PM

The issue I have with that model is, if SWP only does 157 DPS, then why cast it at all? MF spam should theoretically ALWAYS lead to higher DPS. Except that it doesn't. While you've calculated the theoretical DPS increase at the beginning of the cycle. You haven't calculated the DPS at the end of the cycle. I'm not saying your wrong, it just seems incomplete. You don't seem to be taking into account that SWP actually ticks for the full 24 seconds from the beginning of the cast. If you wait an extra 1.5-4.5 seconds to cast it, you may be increasing your current DPS, but what about 24 seconds down the road? 48 seconds down the road? If you wait an average of an extra 3 seconds to reapply pain due to placing MB and SWD higher in your priority, you may be gaining 327 or 607 damage initially, but are you gaining or losing that damage in the end by not having SWP ticking for that last amount of time? I mean, what you're really doing is adding an artificial cooldown to SWP in order to lower the cooldown (over the course of 4-5 rotations) of MB.

Good posts though. I'm sure there is someone who has access to an automated script or something that can properly model both priority systems to see which comes out higher. If I feel up to it I might go Dr. Boom for awhile and see which comes out ahead for me. Anecdotal evidence doesn't seem like it will be enough to convince me or anyone else.

edit: After some more thought, I am thinking that you are correct, but there is going to be a point in + spell damage that putting MB+SWD ahead of SWP will be a decrease in DPS due to spell scaling alone. It would be great to know what that point is and if it falls anywhere along a spectrum where we should be concerned (somewhere between 0-2k shadow damage) about which priority system to use .

Xertigo · 01/04/08, 7:50 PM

Recently, I've been approaching the priority list problem more from a 'What is the maximum amount of damage I could possibly do?' approach and it has also caused me to rethink the placing of SW:P over MB in the priority list.

The way I theorycraft things is to start with a fixed fight length, say 3 minutes (180 seconds) against a single target.

In this fight, the maximum number of SW:P ticks one can have in this fight is 60. The fight length also suggests that you will cast SW:P 8 times (I'm pre-T6, so no 2pc bonus for me, for T6 2pc, you only cast SW:P 7 times). The question then becomes, in practice, how many fewer ticks than 60 do I get (this is ultimately the dot uptime question). Clearly, MB and SW:P won't be up at the same time 8 times during a 3 minute fight, but lets imagine it happens 2 times and you delay casting SW:P by 1.5 seconds each time. This would mean, you lost 3 seconds or 1 tick of SW:P damage as a result.

Now, lets look at the reverse. If you delay casting MB by 1.5 seconds twice, that would be 3 seconds of MB damage lost. Assuming 4/5 Imp. Mind Blast (another pre-T6 assumption), the casting + cooldown cycle of MB is 7.5 seconds, so 3 seconds represents 40% of a MB cycle. In my experience, 40% of a MB cycle is greater damage than 1 tick of SW:P.

I don't agree with prioritizing SW

over SW:P though. 3 seconds is only 25% of a SW

cycle, which is less than a SW:P tick. Returning to my 3 minute fight example, you can only fit a max of 15 SW

into a 3 minute fight and it seems more realistic to aim for 13-14. Even eating up 6 seconds at the start, still gives 6 seconds of slack for the rest of the fight.

I've seen similar stuff on shadowpriest.com with the associated comment that a priority list that favors MB is more mana intensive, but higher DPS.

LucidityAxel · 01/05/08, 11:42 PM

Spell prioritization starts to change once the target is low on hit points. The rule of thumb is to prioritze DoTs over direct damage spells as long as you think the DoTs are likely to get close to their full duration. Stated another way, this means you skip DoT refresh once the mob only has a few more seconds to live.

This is something that only experience can teach you, as it depends greatly on the dps of your individual raid -- you won't learn it from a spreadsheet or from other theorycrafting.

Balkoth · 01/06/08, 4:30 AM

Originally Posted by LucidityAxel

Spell prioritization starts to change once the target is low on hit points. The rule of thumb is to prioritze DoTs over direct damage spells as long as you think the DoTs are likely to get close to their full duration. Stated another way, this means you skip DoT refresh once the mob only has a few more seconds to live.

This is something that only experience can teach you, as it depends greatly on the dps of your individual raid -- you won't learn it from a spreadsheet or from other theorycrafting.

No, just...no. If SW:P did 1/3 of its usual damage per tick, you'd use MB/SW:D and possibly MF ahead of it. The question is DPS, not DoT uptime. You CAN learn this from theorycrafting and apply it.

The question was, and will always remain, what do you lose the least amount of damage by postponing? Some of the recent posters are correct, MB IS better than SW:P...at higher gear levels, more points in Improved Mind Blast, and/or 4 piece t6 bonus. However, at lower levels of gear or if you don't abuse Improved Mind Blast, SW:P before MB will result in more DPS. Divide the average amount of Mind Blast damage by its cooldown + 1.5 seconds. Divide a SW:P tick by 3. Use whichever spell results in a larger number first. SW:D can also be calculated by dividing the damage by 12. You can figure out what order those spells should usually be in based upon your gear and talents.

At lower levels, you'll get...
1. SW:P
2. MB
3. SW:D

At higher levels, you'll get...
1. MB
2. SW:P
3. SW:D

Theoretically, SW:D might pull ahead at some extreme level, but I'm not sure what that point is. It technically scales better than SW:P but SW:P has a large base lead.

galzohar · 01/06/08, 1:05 PM

If you know the mob will live X more seconds, you can definitely theorycraft wether dots are worth reapplying or not. What you can't theorycraft is how many more seconds is the mob/boss going to live. And to be honest, any mob that matters will live long enough for even the last 10-20 seconds of the fight to be negligible, or at least the DPS difference between reapplying dots to not reapplying dots on such a small portion of the fight will be negligible.

cwap · 01/06/08, 6:22 PM

Im sorry for asking yet another question in this debate, but I can't hold myself back :P Been searching the web up and down, but couldnt find the info.

Basically, what I want to know is how you guys cope under the influence of heroism? You still use MB / Sw;D? Or just dot + MF? Personally, I have a hunch that the dps difference is very minor, and that the mana you spent will go through the roof, but what do I know :P

My point is that MB hits for maybe 3k average. With heroism MF ticks for maybe 2k in the same time (1.5 seconds). Personally, I dont think that this will improve our dps a lot, while it will improve our mana a lot if we spam MF. More like a mini-innervate than a dps increaser? :P Either way you do it, heroism wont be the big dps booster it is for other classes.

So, how do you cast under the influence of heroism, and more importantly, why?

galzohar · 01/06/08, 8:22 PM

If you're not lacking mana, MB/SWD on cooldown... If mana is an issue, go crazy on MBs outside of BL and use the rest of your mana on BL... Easier to plan if the BL is used at the last 20% combined with molten fury otherwise it'll take some more "planning ahead". But with good gear and consumeables not to mention resto shaman support you shouldn't have mana problems and spam your abilities whenever they're up, BL or not.

ObservingLife · 01/07/08, 11:33 AM

Originally Posted by Caligula

The issue I have with that model is, if SWP only does 157 DPS, then why cast it at all? MF spam should theoretically ALWAYS lead to higher DPS. Except that it doesn't. While you've calculated the theoretical DPS increase at the beginning of the cycle. You haven't calculated the DPS at the end of the cycle. I'm not saying your wrong, it just seems incomplete. You don't seem to be taking into account that SWP actually ticks for the full 24 seconds from the beginning of the cast. If you wait an extra 1.5-4.5 seconds to cast it, you may be increasing your current DPS, but what about 24 seconds down the road? 48 seconds down the road? If you wait an average of an extra 3 seconds to reapply pain due to placing MB and SWD higher in your priority, you may be gaining 327 or 607 damage initially, but are you gaining or losing that damage in the end by not having SWP ticking for that last amount of time? I mean, what you're really doing is adding an artificial cooldown to SWP in order to lower the cooldown (over the course of 4-5 rotations) of MB.

Good posts though. I'm sure there is someone who has access to an automated script or something that can properly model both priority systems to see which comes out higher. If I feel up to it I might go Dr. Boom for awhile and see which comes out ahead for me. Anecdotal evidence doesn't seem like it will be enough to convince me or anyone else. :D

edit: After some more thought, I am thinking that you are correct, but there is going to be a point in + spell damage that putting MB+SWD ahead of SWP will be a decrease in DPS due to spell scaling alone. It would be great to know what that point is and if it falls anywhere along a spectrum where we should be concerned (somewhere between 0-2k shadow damage) about which priority system to use .

I realize I did not explain the situation correctly. You keep Pain up because for 24-27 seconds, you have tacked approximately 200 dps onto everything you do for a single cooldown. From a pure cooldown perspective, Pain, upwards of 4500-5500 damage for 1.5 seconds of casting, is the bets spell available. However, you only get that dps on that cast every 24/27 seconds. You are losing "seconds" of uptime by postponing refreshing it, however, you will put it back up, thus the DPS calculations to determine losing the ~200 dps for a second or two to use a much higher DPS spell. This is why the situations I brought up are so narrow, and as stated, will only come up occationally and lead to a relatively slight DPS increase, but do change the priority system.

Can the two people dividing the damage by the cooldown time explain why you are doing this? As far as I can tell, this only decreases the gain, not the total DPS. If in the next three seconds I can SW:D/MB then refresh, or refresh then SW:D/MB, I am costing myself 1.5 seconds of MB time, which will be filled with MF + SW:P uptime. Thus, the difference I am costing myself is the MB/SW:D value subtracting the MF + SW:D value in that time frame. *That* difference would then be divided by the cooldown, for theoretical purposes (its discontinuous, as I discussed in an earlier post: you have to do a certain number of "optimizations" in your priority system to get a single additional MB/SW:D; anything before that may actually lower your DPS as you have cost yourself 1.5 seconds of pain uptime without gaining the additional dps of a mind blast, however when you get your additional cast you will get more.)

Also, as discussed in an earlier post, the DPS difference is almost completely flat across spell damage increases, and crit obviously increasing MB/SW:D. The look at spell damage coefficitients (primarily because MF has such a low coefficient, and Pain's per second coefficient isn't that great compared to a whopping 42.86% over 1.5 seconds) explains why this is relatively flat. In any reasonably discussable + damage levels (750 and up), MB/SW:D will always outstrip the pain uptime.

Any ideas on how to properly model the discontinuous gain function involving "gaining" additional casts?