Is seal of insight really worth it?

[Esoterick]

I have seen a lot of tanks mention that they use seal of insight when tanking.

However I have to say I severely question it's efficacy as a survival seal for two reasons:

Firstly - the healing in a given window is extremely weak. It heals just under 1.5k every few seconds and I have never seen myself nor any other tank die by a margin of less than 15k which is considerably more than what the seal is capable of in a 5 second window.

Secondly - Not using Seal of truth(with it's respective glyph) means that you hinder holy power generation through taking an expertise hit. Delaying reach 3 holy power logically also means delaying the amount of time till you can use a 3 HoP WoG.

As an aside I only ever use it if I need the mana such as after casting holy radiance or after being battle ressed.

Has anyone looked into this at all or had similar thoughts? Hopefully I have not just missed some glaringly obvious theorycrafting into this topic.

[Jeremoot]

I've been meaning to take a look at this, and it should be quite easy to calculate the change in your W39 HPS when you change seals. If I make a mistake, please correct me.

In C code, your expected time in seconds to gain and spend 3 Holy Power is calculated as such:

Code: Select all

double x = 0, y = 0;
for (int i = 2; i <= 9; i++)
{
    x += p * binopdf(2, i, p) * 3 * (i + 1);
    y += p * binopdf(2, i, p);
}
return (float)(x / y);


You can then calculate your W39 HPS using the WoG formula. (Note that this is assuming 10% AP buff, 5% crit buff, Divinity, and Rule of Law)

Code: Select all

float m = 9006 + (0.6534f * AP) + (0.0594f * H);
return (1.26f + c) * m / r;


Where c is base spell crit, AP is attack power with only kings applied, H is buffed health, and r being your expected rotation time.

If you have 0 hit and expertise, your probability of hitting is 71.5%, using the 10 expertise glyph brings this probability up to 76.5%.

For our purposes I've chosen the following values:
c = 0.035 (base spell crit)
AP = 6800 (AP with only kings)
H = 180000 (buffed health)

With those stats, using the glyph increases your predicted HPS from 2488.408 to 2658.663 giving the glyph an HPS value of ~170. It should be clear already that the Seal of Insight provides greater healing in a W39 rotation, as the 5% from that glyph would already be about a 120 HPS increase.

Wowhead lists the Seal of Insight healing formula as (0.15 * AP.15 * holy power), but testing on a training dummy shows no increase in the healing effect with Holy Power. Without buffs, I saw a steady 1400 heal from Seal of Insight regardless of HoPo. Without an accurate formula (possibly something warranting a post in the Call to Arms thread), we can't get an accurate representation of Seal of Insight. This is disappointing, but for our purposes we will assume that the HPS gain from Seal of Insight is ~(1500/2.57).

Giving our total HPS gain from using Seal of Insight a value of ~832. This number would be even higher taking into account reckoning and other raid haste buffs.

In conclusion, we can say that Seal of Truth is only for situations requiring the threat component.

[inthedrops]

I tend to use Seal of Insight for it's mana regen more than anything else these days. Allowing me to use spells like Holy Radiance for example, or more consecration when it's useful for adds.

[theckhd]

I've been meaning to take a look at this, and it should be quite easy to calculate the change in your W39 HPS when you change seals. If I make a mistake, please correct me.

A couple of comments:
1) Wowhead's "holy power" is holy spell power, or just spell power as far as we're concerned.
2) We know the formula for SoI healing, it's already been tested. The version in the matlabadin repository is correct.
3) You don't need to guess at most of this, it can all be extracted from the single-target rotation sims. Looking at rotations #3 and #17:

WoG healing with glyphed SoT active is around 2700 HPS, (W39, #17 data set 1)
SoI healing with a standard 939 is around 2400 HPS, (939, #3 data set 2)
WoG+SoI healing with a W39 is around 5200 HPS, (W39, #17 data set 2)

Which would suggest that in the last data set, 5200-2400=2800 HPS is due to WoG. Note that this is using the SoI glyph, so WoG gets 10% stronger in that setup; without the glyph it would be around 2560. So we can estimate the SoT glyph as granting about 140 HPS, while the SoI glyph grants about 240.

Not that any of this really answers the question posed in the thread. It should have been fairly obvious that SoI would be a larger increase in total SHPS than the SoT glyph's effect on WoG. That still doesn't make it "better," just different. You have to gauge whether steady HPS with the risk of overheal and unlimited mana is more or less useful than having your burst heal available more reliably and more frequently, which is something that varies from encounter to encounter.

And in 4.1, WoG gets a 20-second cooldown, which changes usage patterns in SoI's favor. My 4.1 sims are suggesting that W39/SoT drops to around 1300 HPS, 939/SoI healing goes up to around 2900 HPS, and W39/SoI drops to around 400 HPS.

[Esoterick]

Thanks for the replies, was interesting to see the data behind it.

[Jeremoot]

1) Wowhead's "holy power" is holy spell power, or just spell power as far as we're concerned.

Oh, of course.

3) You don't need to guess at most of this, it can all be extracted from the single-target rotation sims. Looking at rotations #3 and #17:

I started guessing when I misunderstood Wowhead's usage of holy power. The HPS value of a W39 rotation and SoT Glyph seem rather conclusive with your numbers. I mistakingly thought the SoI glyph only affects the Word of Glory healing, which is why I only came up with 120 HPS.

And I'll check out that SoI formula, if I knew it was already on here it would have saved me some trouble.

[theckhd]

I mistakingly thought the SoI glyph only affects the Word of Glory healing, which is why I only came up with 120 HPS.

And I'll check out that SoI formula, if I knew it was already on here it would have saved me some trouble.

The SoI glyph does only affect WoG healing.

In general, the formula for anything paladin-related can be found in one of the modules of the matlabadin project, mostly in stat_model and ability_model.

[Jeremoot]

The SoI glyph does only affect WoG healing.

Ah, the tooltip gave the impression that it double-dips. If that's the case, 2560*0.05 comes out to be 128, which is how I got my 120 number. How are you getting 240 HPS as a result?

In general, the formula for anything paladin-related can be found in one of the modules of the matlabadin project, mostly in stat_model and ability_model.

aaaand, bookmarked.

[theckhd]

The SoI glyph does only affect WoG healing.

Ah, the tooltip gave the impression that it double-dips. If that's the case, 2560*0.05 comes out to be 128, which is how I got my 120 number. How are you getting 240 HPS as a result?

By mistakenly multiplying by 10% instead of 5%?

So it should be about 128.

[Jeremoot]

By mistakenly multiplying by 10% instead of 5%?

So it should be about 128.

That makes more sense!

Since you brought it up, your 4.1 simulations interested me, with 939SoI pushing so far ahead of SoT we will absolutely be taking EG over SotP. It's possible that you've answered this already and I missed it, but even with the GC change we would need an 80.17% melee success rate to expect 3 HoPo every rotation. That is, if you want to use 3 WoG every time it's available and you're not waiting to use it reactively. I don't think anybody is going to make hit and expertise their gearing strategy, so to ensure 3 HoPo WoGs on CD you'll have to settle with a 2 SotR in-between. Because waiting to use the EG-proced SotR would push our rotation out further, it seems the only beneficial time to use it would be in place of the 2 SotR. Am I correct in assuming this, or am I over thinking this and going completely bonkers? [/thread hi-jack]

Also, am I the only one that really hates the no HoPo on miss change? I mean, give WoG a cooldown, fine. Taking away my ability to feel confident in having a 3 SotR ready at the pull is crossing the line (Especially when you raid with an ego-tripping fire mage).

[theckhd]

Actually, I've come to completely the opposite conclusion regarding EG. I'll be speccing SotP when 4.1 hits.

[Jeremoot]

Okay, I'll bite. What's driving you away from the SoI939 approach with its superior healing?

[Durability]

I think it's less that and more that EG is absolutely awful in 4.1.

[Esoterick]

Is SoI really generating 2400 HPS?

I am seeing considerably lower figures even in a hit/expertise capped setup attacking a training dummy from behind?

I'm not very good an quantifying data. I am unsure if SoI works on a PPM type mechanic and also if it can only be triggered from autoattacks.

Following a standard rotation and avoiding WoG I was seeing just short of 540 HPS. Regardless of what I do I can see the heal procc just over every second at the most.

Again I suspect I am just missing something really obvious. Any ideas?

[Jeremoot]

I think it's less that and more that EG is absolutely awful in 4.1.

Right, but if you're prioritizing SoI, SotP is essentially 0. I would take awful over nothing.

Is SoI really generating 2400 HPS?

Using the matlabadin models with the numbers from your armory:

( 0.15 *(2251+7423) * 1.06 ) * (15*2.6/60) = 999.8079 unbuffed HPS from SoI.

What you were forgetting is that Vengeance is going to give you a whopping 18k AP. This brings your expected healing up to 2860.1079.

EDIT: Nevermind, got it.