Shield Barrier theory crafting, need some input

[Airowird]

So with the new Rage changes GC posted yesterday, it's pretty much impossible to NOT have maximum Shield Block uptime, which means Shield Barrier usage will become the difference between a good and a great prot warrior. (on top of not sucking at your rotation, keeping SB cooldowns rolling, managing cooldowns for big boss breaths and raging on the VoIP if your choice)

This also means that SBar theory crafting becomes more important, which is where I need some ideas from you guys.
Right now, I have always considered average damage reduction per hit, but with SBar there are more possibilities...
a) Increase base health pool by a portion of the shield, depending on Time To Live
-> extra HP = TTL / (time till 60 extra rage) * SBar shield value
b) Treat the shield as additional "damage reduction", based on how hard the boss hits
-> extra DR = shield absorb per second / effective boss DPS
c) Treat the shield as an extra chance to 'block' for a certain amount, based on boss swing & shield values
-> SBar 'block' = (boss swing timer / time 'till X rage) * (absorb value / full or blocked swing)

Obviously, simming would be the best option, but for simplicity, I'ld prefer one of the above for my spreadsheet, which is kinda annoying because (a) assumes some sort of EHP scenario and disregards rage generation, (b) assumes you average it out, which is counter-intuitive to it's stacking nature and (c) well ... is just a pain to get right.

Any thoughts on how I should view SBar when it comes to EHP, total damage reduction & burst reduction??

Also, would you consider bleeding off rage pre-emptively with chunks of 20 Rage or just pop one when you're at 90/110 rage to prevent capping?



PS: Shout-Charge-Shield Slam (ZR if you miss) = start off with SB instantly, you can even get it just from Charge+Berserker Rage (or Shout) if you got 2pT14 and the Bullrush glyph. How bursty is that for rage income? I really think they should've just upped the def stance rage generation to ~5 every 3 seconds

[Tengenstein]

I'd treat it as option a, extra HP, and then work out the RPS

[Airowird]

So more like "extra HP = absorb per sec * boss swing timer" ?

[Tengenstein]

something like that, It's gona be a real pain to model.

[Quietsch]

I honestly believe, that there aren't any... well, really working solutions for this in a spreadsheet. The effect of SBarr basically boils down to overall damage reduction (so something like RPS over 6.66 * avg_vengeance/estimated vengeance cap)(Which is equivalent to extra HP in a generell regard). Of course, it depends on what you want to show (TTL, overall RPS/damage income).

so, EHP: Not included, but maybe a... well, additional cell, which shows the effect/EHP with SBarr.
TDR: as mentioned before
Burst reduction: ... Basically the same as for EHP, but i'd differentiate between the SB up/down cases.

Also, would you consider bleeding off rage pre-emptively with chunks of 20 Rage or just pop one when you're at 90/110 rage to prevent capping?

Depends on the situation. If i know, i'll have enough rage to get SB up, when it's available, i'd do the more appropriate (huge hit incoming -> pool rage for high SBarr - low inc, use 20 rage). If I'm not sure, if i'll have enough rage for SB, I'll use it only at 90/110 rage. (Or even at the cap, if I just had a critblock/it's a chaincaster)

[Tengenstein]

One thing to keep in mind is that you can't overwrite an SBar. If Sbar is already up, even with a value of 1, and you have a 60+ rage Sbar with a bajillion vengeance behind it, you can't put a new Sbar up until th eold one is consumed.

[Quietsch]

Which is wrong - you can overwrite it, but only if the new SBar absorbs more, than the old one.
/e or at least, if the new uses more rage, than the old - but i'd doubt it's based on rage consume...

[Tengenstein]

ah fair enough, must have been some momentary vengence dips when i tried it

[kopcap]

Whats is order of consumption between pws and sbar?

[truculent]

actually, after last nights discussion in shoutbox, I already started working on something. Ill throw it up here tomorrow when i have more information.

[Airowird]

So here's what I'ld come up with so far:

SBar absorb per rage = 2*(AP-2*Str) / 60 Rage
SBar absorb per second (aps) = absorb per rage * rage per second consumed by SBar

For EHP & Burst models:
min boss swings to kill you * boss swing timer * aps = extra HP, add average 'extra HP' buffer to health pool.
Technicly this isn't a guaranteed health buffer, but I would assume that if you're below 25% HP, you're going to mash the hell out of your Shield Barrier button anyway, so in reality, SBar is being undervalued (but again, you'ld need to sim it to be more accurate).

For damage reduction/TP:
boss swing timer * SBar aps = average absorb per boss swing (= "absorb" when talked about below)
divide this by average damage per boss swing = damage reduction from SBar, add this into total.
The problems that arise here are
a) Damage reduction is going to be more dependant on environment.
b) It makes the TP formula more complex, we're going from HP / (1-TDR) to HP / [(1-TDR)(1-SBar_DR)]
with SBar_DR = absorb / (raw boss swing * (1-TDR)), or better written: TP = HP / (1-TDR - absorb/raw boss swing)
c) I have to assume that you consume SBar completely within it's 6s buff time (again a reason to use smaller fragments)
d) I have to redo ALL my freakin math on stat values AGAIN!! FML

[Takethecake]

I don't think it's really worth while / feasible to model shield barrier into a spreadsheet. It could be done in a rough manner but the data it would give would really have to be taken with a grain of salt. The numbers given by the spreadsheet show ideal DR and damage intake but SB has SO many conditions and variables it would really only be useful in a sim type environment.

Best case scenario I would create a new HPS field and throw shield barrier data in there.

[Airowird]

Actually, the numbers given are already an average. Sure, how much that average changes per stat is pretty accurate, but I would hardly call any number from any kind of spreadsheet or even a sim "ideal". Neither SB nor SA has a lot of conditions, if you have the rage for it, you should SB, if you have too much rage, SA to dump it. Not hard or complicated imho.

[Quietsch]

The problem with sunder armor... erm, shield absorb.. shield barrier conditions is, that you need to guess, at least to a certain extent, how much rage you'll get in the next timeframe until SB. 40 rage at 9s SB-CD can be considered too much, but if you get really unlucky, you get only something like... well, less than 60 rage in those 9s (Which should be nearly impossible with the recent rage changes...), so you'd have to delay SB. On the other hand, if you don't use SBar, you're only guaranteed to get enough rage for SB, at certain 'timestamps', depending on SS/Rev/Shout/Zerker rage CD.
The problem with SBarr (at least for a correct use) is not this much present in the actual game, but much more in a simulation/to a certain extent spreadsheet - in the game, it's much more intuitive when to use it.

[Tengenstein]

The other problem is that you don't get alot of vengeance until after you've taken the first boss spike. Madness for example, i can barely get a 50k Absorb up pre impale, but post impale I'm covering myselvf with 250k absorbs that the tentacles just can't get through.

[Airowird]

Well I did the math anyway and ... for TDR it's effect on stat values is barely noticable. Sure, it's not entirely accurate on damage reduction with stuff like Impales, but for an easy tank-and-spank, it should be somewhat within realistic numbers.

[kebess]

Whats is order of consumption between pws and sbar?

Good question. Intuitively I'd say the last buff you get is the first one that's consumed, but I have nothing to back this. You'll have to directly test it.

/e
While we are at it. How is CB determined ? Is there a ''3rd'' roll ?

[Airowird]

CB = that much chance on a block roll that you block for double amount, so yeah, guess 3rd roll after Avoidance and Block checks.

Intuitively i'ld say it's FIFO on absorbs: the one that's on you first, runs out first. It makes is slightly easier for the game to track them and it makes sure shields get used frequently. The expection here is that shield procs (not sure how Holy din Mastery is considered guaranteed or proc) come before 'direct cast' shields, because you want them gone first so they can proc again for the full amount.

[Eetabeetay]

Could someone give me the formula for the rage scaling on SBar? I don't know if I didn't look hard enough or what, but I couldn't find it. I assume that the optimal absorb/rage level is at 60 rage, but I hate to simply make assumptions.

[Tengenstein]

it scales linearly.