Blood Draining Enchant - TTL Simulation

Lave wrote:i guess thats true for someone who doesnt get additional heal? when i get strucked down to 10k life by a 20k hit & the next one is coming i need >10k healing in those 2 seconds to survive. blood draining reduces that gap to like 7k.
sure, BD alone doesnt save me in this scenario.. but its still more likely for me to survive in the constant stream of hots/direct heals incoming

It's a little more subtle than that, but essentially you're correct. As I said in the original post, constant healing sources shouldn't change the overall result that blood draining's effectiveness drops off as the boss hit size increases relative to your AD range. While you can come up with where Blood Draining helps (like the one you proposed, assuming you're getting 8k healing between every boss attack), you can come up with others where it boosts you out of AD range (i.e. a boss hit takes you to AD threshold - 7k - blood draining proc, and the next one kills you as a result). The question is always whether it's creating more of the former situations than the latter, and by how many.

Healing extends the range in which blood draining does help. By adding healing, it shifts the whole distribution up. Just to make sure we're not overlooking anything, let's just add healing to the simulation and see what happens. For the next two plots, I've added a constant 8k healing after each "combat round" (i.e. boss hit event + possible blood drain proc). I've also corrected one oversight in the code - before blood draining could proc multiple times, but since we're really only interested in sequences of hits, you wouldn't have time to build up the 5-stack again; as such I've now limited it to a single enchant proc per series.


Plot with AD threshold on the x-axis instead of max health
The first thing I want to point out is that by adding 8k healing, you've become invulnerable (effectively) to hits less than around 11k, because once you're in AD range you'll be hit for less than 8k. This also means I had to modify the simulation to end at some point, because otherwise it would go on forever - it now caps out at 15 combat rounds. So the dark blue section below 11k is effectively "invincible" territory, where you'll live pretty much regardless of the enchant.

The region just above this is where the enchant shines - you'd have died by ~2k health in this region without the enchant, so it saves you an awful lot of the time.

More importantly, look at what healing has done to the upper half of the plot. Now, the enchant is giving you some benefit for hits up to 22-26k, depending on max health. This is roughly 8k more than the last plot, consistent with the conclusion that constant healing doesn't change the distribution, it just shifts it up some. So your intuition is correct, in that for a particular hit size, blood draining gets stronger as your incoming HPS from outside sources increases.

Also note that again, Blood Draining will never decrease your overall TTL, at worst it just doesn't increase it any. At the upper end of the boss hit size spectrum, it trails off to zero effectiveness just like it did before.

For the theorycrafters out there, I would propose the following metric to incorporate heals into the situation:
(AD_Threshold + Incoming_Heals - Boss_Hit_size) > 0
If that value is greater than zero, Blood Draining has an effect. The larger the value, the better the effect, at least until you reach the limit where you're receiving more heals than damage. If the value is near zero or negative, the enchant won't help you very much.

For me, I would estimate the average boss hit at around 14-15k, which would make blood draining a relatively effective enchant (again, assuming 8k heals between boss hits). I distinctly remember Hodir hitting me for around 13k with Frozen Blows last night in my frost resist set, and I think the melees were roughly the same. As my healers spellpower and my armor/mitigation improves, it will only make the enchant better.

Obviously this is all hand-waving math at this point - we'd need to look at some WWS parses to see whether 8k is a reasonable estimate, and to be more accurate with our weightings we should work on defining a probability density function the way cordelia suggested. But we're now talking about finer details that shift the value around, not the overall conclusion. I think it's pretty clear that Blood Draining isn't a terrible enchant, and that the assumption that it's bad because of AD leapfrogging is incorrect.

After looking at the results here, I'm tempted to try and get my hands on another Last Laugh just so I can have a second one with Blood Draining on it for bosses. I'd still like to see how it compares to an avoidance enchant like 26 agility; I suspect that for really large boss hits avoidance would win out, but for hits in the 15k range Blood Draining might be more effective.

TLDR Summary:

• Adding healing to the equation extends the range in which Blood Draining is useful.
• A good rule of thumb is (AD_Threshold + Incoming_Heals - Boss_Hit_size) > 0. The bigger the value, the better the enchant is.
• Blood Draining never reduces your TTL. The extra AD Leapfrogs it causes will always be at least offset by the extra AD saves it causes.

Kelaan wrote:I think we're less likely to be IN the AD (or enchant) activation range, but that damage reductions on incoming hits (whether from shields or from defensive cooldowns/HoSac/DSac) could make the boss hits effectively fall to a more favorable size as far as AD and the enchant are concerned.

How long do these calculations end up taking, anyhow?

Any damage reduction effect is analogous to reducing the base boss hit size, which always makes Blood Draining better, just as you surmised.

The calculations don't take very long at all, a few seconds even for the mesh plots. They took far longer this morning when I was letting them go on for thousands of combat rounds (or more), until I realized that was silly.

What really takes time is the analysis. The computer can crank out numbers very fast, but it can't tell you whether you've made a mistake in your algorithm, or tell you how to interpret them. Once I have the data, I take a look at some test cases to make sure they make sense, and make sure I understand what the numbers actually mean.

Squishums wrote:So any small heal on a pally tank below 35% is a double edged sword.

But does this mean BD is better than we instinctivly feel because the possibility of lowering our TTL that make us so nervous is just the same as what healing causes......or that small steady heals are in fact bad and nobody should stack hots on a pally tank, instead just spam big enough heals that if it took you out of AD range it would make up for the extra dmg youll take?

It means BD is better than most people assumed, for the reason you gave. It's very easy to imagine one bad scenario and draw a conclusion from it that's incorrect because you failed to notice a compensating good scenario.

The second statement you made (about not stacking HoTs on a pally) is demonstrably wrong from this data alone. If a small 2k heal at 35% never decreases your TTL, then there's no reason to expect that a HoT would be any different. In other words, heals are always a good thing, no matter the source.

cordelia wrote:Honestly, the Health distribution should be exceedingly easy to bootstrap from a combat log with simple unix shell skills. Too bad my shell skills are so rusty.

I was thinking the same thing - take a WWS parse of a reasonably long fight or series of fights, and figure out both the average boss hit size and the amount of time spent in each health range. I don't have access to a Unix machine at the moment though, so for me it might be faster to load the whole combat log into MATLAB and do a line-by-line filtering and summing algorithm. I don't know if/when I'll get the time to do this though, so if someone else feels inclined to work on that problem please feel free to do so.

theckhd wrote:

cordelia wrote:Honestly, the Health distribution should be exceedingly easy to bootstrap from a combat log with simple unix shell skills. Too bad my shell skills are so rusty.

....I don't have access to a Unix machine at the moment though, so for me it might be faster to load the whole combat log into MATLAB and do a line-by-line filtering and summing algorithm.

If you install Cygwin (and put its bin directory in your Windows path) you can have access to tools like grep, sed, etc from your normal windows command prompt. I feel crippled without it. They will install an xterm launcher as well, but I just use it in my normal windows shell, as I don't like the way they map the filesystem into "unix-land" in cygwin's bash shell.

Code: Select all

(15:34:00.92)> ls *.el -lFart
(...)
-r-x------+ 1 kelaan ???????? 920910 Apr  9  2008 color-theme.el*
(...)

(15:34:32.69)> grep "(defun" color-theme.el | grep "lethe"
(defun color-theme-lethe ()

(15:34:34.47)>

(After the fact, I realized that you can't tell this is on windows, considering I obfuscated my prompt and am looking at XEmacs lisp modules. )

Kelaan wrote:Wow. Awesome! I appreciate your analysis of what it means, and what we can derive from it.

You hit on an interesting point, though: healing makes the probability of being at various health totals different. I think the Next Step (haha) is an app that lets one apply healing strategies (chained HL's, HoTs, Shields), and show a the value (e.g., your 3d plots) of AD and the enchant. I think we're less likely to be IN the AD (or enchant) activation range, but that damage reductions on incoming hits (whether from shields or from defensive cooldowns/HoSac/DSac) could make the boss hits effectively fall to a more favorable size as far as AD and the enchant are concerned.

How long do these calculations end up taking, anyhow?

I think the most interesting way to work with healing styles is model those healing styles and then have them be added modules such that you could plug in two holy paladins or a holy paladin and a resto druid or a Disc priest, rolling lifebloom and AoE heals etc. Lot of work but would be very interesting and powerful.

Thank you for the analysis!

Essentially we want to know the relative benefits of blood draining and avoidance enchants and the difference is probably small. Empirical determination of probability distribution of health might introduce too big of a source of error in to the calculations. I could see it being affected by things like avoidance level, class of the healers and the spells they use, and time of tank death. Are you thinking about using succesful or unsuccesful attempts? Or maybe just the events between death and last time at 100% health? It might be better just to determine the frequency of situations where blood draining saves you or lets you take one more hit.

Rather than using probability distribution of tank health I was wondering if the calculations could be adjusted for probability distribution of incoming damage (even a made up one)? Or is this wrong way to go about it because the order of boss abilities has some restrictions? Is it possible to identify bosses where blood draining performs well/poorly? The example case where boss hits for a fixed amount that can potentially leapfrog AD makes a bad case for the enchant. if there are small hits (like AoE) in between the bigger hits the enchant would probably perform better.

Also, could you quantify the relative benefit of blood draining for a tank with and without AD in your model? The enchant is sometimes being dismissed because it's supposedly worse for us than for other tanks.

The tank health affects the choice of healing spells. If tank's health drops low healers will often use instant heals. At least from healer PoV this enchant looks promising

I think im way out of my league trying to understand some of the stuff you mentioned, so ill just asked....ask a paladin tank with 32k hp unbuffed, raid buffed im at 41k, is blood draining worth the enchant slot for my weapon or should i go with another enchant? id post my armory but atm im wearing my ret gear lol...

ill log on as soon as i can and try to log out with the prot set to update it
http://www.wowarmory.com/character-shee ... n=Holynite
wow-heroes.com has my prot gear on it

Seems like it to me. Theck's work seems to say (to me) that it's never a BAD enchant (as it acts like any OTHER small heal, and no one worries about HoTs taking them above AD thresholds, for example). Is it better than the +Hit enchants? Better than str, or agi? depends on what you need. Agility contributes to avoidance and block cap (and armor), +Hit enchants are good for threat (but useless once you're hitcapped, as I am), and Blood Draining looks like it's very good for general potential to save your ass. It probably will save your butt more often than Ardent Defender will.

In light of this blue post:

A bug with the Blood Draining weapon enchant has been fixed to allow the Blood Reserves to restore the proper amount of health.

One would assume that it is possible it was restoring to much health, maybe altering the data you gained for this enchant. Even if it was too little, I think another round of testing for this enchant is due.

Wondering if it still crits though.

mhm well hots can also get u out of ad range as well and as far as i am concerned this talent is not that effective when maintankin a boss ... same with the weap enchant ...maths are really nice and the analisis as well . Tbh since peop in this forum (winks to worldie) informed us that we could drop ad for high end maintankin last year , i don't mind about loosin a couple of triggers in it and well for my server that is the best enchant i can get to replace potency and moongoose.

Slight offtopic, i wonder why this topic seemb to attract spambots. Might have something to do with draining blood from innocent people?

HMM, i wonder what bug they fixed? it not healing enough? it critting when its not supposed to? or healing to much?

hmm

Thank you for this breakdown. I might just switch to blade ward after reading this.