You both offer contrasting views.

Providing strong data would be the best way to guide a final decision on the effective of the 4 piece bonus and it's relation to PVE vs PVP gear.

[quote=the KRIS;1844375]Just curious how the most recent change to Mastery (2.5% per point to 2.1%) will affect modeling. Per Redcape's post on page 15:/QUOTE]

It should be completely linear from my understanding, however, I would hope this change was done so that when the mastery damage is affected by Inq it will only be slightly better than Haste/Crit as opposed to significantly.

Re-reading that part of the OP it sounds like its suggesting the 6 second mark if you only use 3HP for Inq. I've personally been running clcInfo setup to refresh with a 3HP Inq at under 3 seconds or when it falls off with whatever HP I have available. (The 3 second mark was chosen so that in the worst case I can CS and then Inq right after it falls off.)

I'm curious to know if Exemplar models 1 and 2 HP Inq to determine the best refresh point or just assumes a 3HP Inq. If you're switching from only using 3HP Inq to sometimes using 1 or 2 with the 4pT11 then I agree that it changes playstyle a bit, but if you're already playing that way then the bonus just means fewer Inq and more TV.

In either case, figuring out how it may or may not effect rotation priorities is largely an out-of-game exercise for most people I would imagine. Once the theorycrafting is done, just plug the results into clcInfo (or similar) and mash buttons during actual encounters while focusing on the fight mechanics. I doubt I would notice too much of a qualitative difference with or without during a fight (though, obviously I'd see bigger dps numbers on recount after the fact.)

Without 4pT11, it is best to wait for 3HP to use Inq, because otherwise you are using more of your GCDs to keep inquisition up. this is why Exemplar didnt model 1 and 2 HP inq. In example, lets keep inquisition up for 1 minute. If you use 3HP inquisitions it would take 2 GCDs, 2HP inq would take 3 GCDs, and 1HP inq would take 6 GCDs. So, with 2HP inq you are losing out on 1 DPS GCD during that minute, and with 1HP inq you are missing out on 4 DPS GCDs during that minute.

with 4pT11 it is more complicated because you now have to factor in GCD usage and HP efficiency. lets use 2 minutes for our example (so everything is neat and even). using 1HP inquisitions would cost us 6 GCDs and 6HP, 2HP inq would cost us 4GCD and 8HP, and 3HP inq would cost us 3GCD and 9HP. I would think that using 3HP inq would be superior, but I cant say for certain. Also, I didnt account for inq overlap in these examples.

Factoring overlap would skew the data, but it would depend on the duration of Inquisition that you were refreshing (i.e. 1 GCD of 2 HP to replace 10 or 15 seconds).

You also would need to consider when you are using Inquisition. Using it during GAnK/AW would reduce your damage output, unless you cast it at 3 HP (making it exceed the combined duration of 29 seconds of GAnK + AW + Zealotry, assuming you use AW at the 21 second remaining duration on GAnK), despite the reduction in Hammer of Wrath damage, should it go Live. Its not to say that not having it up either would increase DPS, but you just avoid losing a GCD for a damaging filler.

Not only that, but also an increased duration. Compare it to Savage Roar-- Given how difficult it will be post-patch to have consistent HP, you would think that the duration of Inquisition would be around 40 seconds baseline.

While correct, these numbers drastically simplify the situation. First, we aren't completely GCD locked, so just because it takes more GCDs doesn't necessarily mean it will be a dps loss. Second, the "3 second rule" I use ends up with over half of all Inq casts being 3HP anyway with the bulk of the remainder 2HP so you really need to compare a sort of "blended" value. Finally, you aren't accounting for the increased HP efficiency that the "3 second rule" can bring, resulting in more TVs.

I tried to run some napkin math to illustrate the point, but it got complicated so I broke down an wrote a rotation simulator. This is a quick first pass, but the results seem reasonable, so I don't think there are any major problems with it.

Settings:
no latency (didn't implement this yet)
no haste or mastery from gear (for no particular reason than that's how I ran it at first)
3.6 speed weapon (with full buffs: +10% from raid and +9% from talents give a swing speed of just over 3 seconds)
(GCD spell haste will full buffs: +5% from totem and +9% from talents)
8% mastery (base chance)
no 4pt11
8s Judgement (no PvP bonus)
30s Consecration (no glyph)

10 hour fight duration (to even out any randomness)
no cooldowns used
no time spend below 20% (no HoW at all)

Results:

6s	3s	delta	Ability
11989	11989		Melee
2378	2400	+0.9%	Exo
3697	3764	+1.8%	TV
7057	7000	-0.8%	CS
3424	3405	-0.6%	J
1863	1848	-0.8%	HW
986	972	-1.4%	Cons
1265	865		Inq @3
0	433		Inq @2
0	124		Inq @1
1265	1422	+12.4%	Inq Total
3795	3585	-5.5%	HP for Inq
90.90%	97.08%	+6.8%	Inq Efficiency
98.39%	98.38%		Inq Uptime

Working from the bottom up, we can see that Inqisition is up about the same amount under either strategy. However, the 3s strategy makes noticeably better use of HP for Inquisition (100% efficiency means you get the full 10 seconds per HP used, so the 6s strategy gets an average of 9s per HP and the 3s strategy gets about 9.7s per HP). Due to the higher efficiency and similar uptime, this means that the 3s strategy is using less HP overall. This frees up HP that can be used for TVs instead.

Looking at specific number, the 3s rule uses 210 fewer HP for Inq which should translate into ~70 extra TVs and, in fact, we see that the 3s rule has 67 more TVs.

The downside, of course, is that the 3s rule is using more GCDs for Inq. It isn't a huge increase since 61% of the Inqs are still @3HP, 30% @2HP and 9% @1HP, but its still a 12.4% increase over the 6s case. This isn't automatically a DPS loss, but we do see some reductions in filler use (most notably with Cons).

I couldn't say without additional testing how much of the loss/gain on abilities is due to RNG even with a 10 hour fight time, but I think this data is fairly suggestive that the 3s rule provides better efficiency than the 6s rule, which leads to more TVs. It also results in slightly more GCDs used for Inq, but this mostly seems to effect our worst fillers, so it may very well be worth the extra TVs you get as a result. Not conclusive, but certainly interesting (which is why I asked about modeling 2 and 1 HP Inq in the first place).

Unrelated question: Does anyone happen to know if Inquisition's GCD benefits from spell haste?

Outside of 4piece I only model 30 second Inq. Unused GCD are so few in number (most of what look like GCD are only partial GCD, you do end up delaying something if you use them) that it's been a loss to uniformly refresh Inq at 1 or 2 HP.

With 4 piece it was just tacking on 10 seconds, the present model (not currently available) swaps to 1 HP mode for maximum benefit. The partial GCD lost are worth the extra TV gained (since it pushes some low end of the spectrum abilities away). Probably worth a 2 HP right before an AW phase.

Based on some comments (not chippy's), I suspect some people do not understand the recommended 6 seconds. If Inq is just at 6 seconds - ignore it, use TV. Below this value start thinking about refreshing. This means the majority of your refresh will occur in the 0 to 2 second range. Some few in the -2 to 0 (you lost Inq for up to 2 seconds) range. And some few in the 2 to 6 range. It's better to refresh at 4, or 5.9 seconds than let it expire for a prolonged period. Testing showed you tend to average 3 HP (or a HoL) within 8 seconds. This may be slightly too long now, it may be closer to 6 or 7. Personal play should find the optimum window for yourself.

The main point is, refreshing Inq at 10+ seconds is a horrible waste of a lot of HP. Letting it fall off entirely for 1 or 2 seconds every cycle is perfectly acceptable. 100% uptime is not a necessity. Goal would be 100% uptime with zero waste (always refreshing at 0 seconds remaining). Simply minimize waste either directly - long downtime is bad, as is early refresh.

Chippy - if you're using a 3 second rule and it's mostly good, try waiting that extra second or two for the third HP before you engage. See if this negatively or positively impacts your overall DPS. I expect either a wash or a gain.

Inq appears to take a full 1.5 sec GCD.

Not completely. There are plenty of scenarios where you would get the TV, and possibly the CS, with Inq still rolling. I know generally if I let Inq fall it's a situation where I have 3 HP with 5-6 seconds left on Inq and an AOE proc or all my fillers up. In a situation like that you are likely to get the mastery affected skills off, and only lose the Inq boost on a filler abilities tops.

While we're discussing Inq and when to hit it - it happens now and then that my first swing in a fight will proc exo, or I'll get a proc just as Inq falls off (if I have ~5s left I'll usually hit TV and let Inq fall off for a second or two before refreshing). Either way, I have 1 holy power and 2 button presses until I'm guaranteed another. Is it better to use the exo at reduced efficiency (no Inq), hit Inq at 1 HP and use the exo, then build up 3HP for normal duration, or hold onto the proc for a few seconds until I have 2 or 3 HP for a more reasonable Inq duration before hitting both?

I re-ran my sim with these settings and got a small increase in ability usage across the board. Inquisition uptime drops from 98.38% to 94.43% so its not automatically a dps gain, but certainly interesting in any case. I'll try to flesh out my model a little and add some actual numbers so I can do more useful comparisons and see what pops out.

Does anyone know at what point a mastery affected by Inquisition would be approximately equal to Haste/Crit? (approximate being within .05 dps per point of each other). I feel that Blizzard is trying to make our secondary stats equivalent to each other so that we don't end up mindlessly stacking one over the other two. I would hate to have a 2.1% mastery since that wouldn't be close enough to haste/crit without Inquisition affecting it.