Moonkin Guide for Dummies! - Page 12

Lilija · 02/04/09, 3:12 PM

Originally Posted by Euthanized

I scanned through and couldn't find which talents are best to sacrifice to get mana regen talents?
My current spec only allows me about 2.5 minutes of sustained all-out dps, and at my guild's current dps output this is not enough time to down a boss.

Looking to drop Typhoon and GW which I don't think will be a problem, but also considering dropping IMF so I can 3/3 Moonglow and Dreamstate. Is this a big no-no? Looking at the "base spec" on the first page (and all consecutive pages) its on almost every spec, just not sure what else can be sacrificed for longevity with minimal dps reduction until my gear is up to par.

If you have replenishment (and you really should since even Blizzard stated that they consider replenishment sort of a must for raiding if you don't want to have mana problems) you should do good even with less mana talents. Personally I find putting 1 point in each quite optimal, however observing how my mana doesn't go down much in 25 mans I think I'll drop even more. As for classification of which mana talents are the weekest, I would said that without going much into details, the resto ones are better than balance ones.

Euthanized · 02/04/09, 4:21 PM

Originally Posted by Ashaera

Mana - Adjust as needed, 10man needs differ alot from 25man etc.
Omen of Clarity > Intensity > Dreamstate > Moonglow
In some specific gear cases Dreamstate (Atleast the first 4% point) might outshine Intensity.

Been following this in terms of which order to spec mana talents. The main question is which damage talents I can forfeit without too much loss of dps. AoE talents seemed to be in the first cut (can't find the post), but I am keeping Starfall for general utility. The main issue I am contemplating is if I will miss the extra 10% from 2/2 IMF, or if these 2 points can come from another area.

We are only doing 10-man at the moment, and we do have two raiders with replenishment, perhaps I should just follow your advice and take my 3 points from Typhoon and GW and put them in Dreamstate, leaving IMF alone.

*edited for accuracy

Eilt · 02/04/09, 4:36 PM

Originally Posted by Euthanized

Been following this in terms of which order to spec mana talents. The main question is which damage talents I can forfeit without too much loss of dps. AoE talents seemed to be in the first cut (can't find the post), but I am keeping Starfall for general utility. The main issue I am contemplating is if I will miss the extra 10% from 2/2 IMF, or if these 2 points can come from another area.

We are only doing 10-man at the moment, and we do have two raiders with replenishment, perhaps I should just follow your advice and take my 3 points from Typhoon and GW and put them in Dreamstate, leaving IMF alone.

*edited for accuracy

A quick armory shows you do not have all that much spirit as of yet (it comes later with gear). If you are having mana regen problems at this point I would recommend putting points into Dreamstate, not maxing out moonglow. Now as to what to give up it depends on your play style, if your guild runs into trash and AEs them down, and you do sarth 3d every week where killing adds at certain times is vital I would not give up GW. On the same note if you find situational use of typhoon (saving a healer or on skeles for gluth) you may rather keep that.

All that said, it pains me to see a raiding spec without Treants, if you use them right and they survive the whole time they do more damage than any other single GCD.

triarchelon · 02/10/09, 5:32 AM

Someone suggested I post this here:

Originally Posted by triarchelon

I'm not sure what made me go test this. I think I was wondering if I had to apply my Improved Faerie Fire to get the increased critical chance on a target. My testing seems to show that I do not get the crit if I apply it, only if someone else does.

Has anyone else tested this?

Target: Grand Master Training Dummy (level 80)

Starfires, no Faerie Fire
1553 hits
555 crits (26.32%)
2108 total

Starfires, my 2/3 Improved Faerie Fire
1654 hits
580 crits (25.96%)
2234 total

Starfires, another moonkin's 2/3 Improved Faerie Fire
1571 hits
614 crits (28.1%)
2185 total

I did some tests before this but do not have them documented and the testing was not a large enough sample to make sure I didn't have some bad luck streaks.

I had the same gear on, always in moonkin form during the testing, and always used the same training dummy. No other debuffs were on the target, and no other buffs on myself other than moonkin aura. I started thinking that maybe 2% crit was not something I would notice, and decided to remove some crit gear (1.18% of crit) and test it out. I did straight Starfires on a non-debuffed training dummy, just as above. I definitely saw the difference in my crits, it dropped to 23-24% (two separate trials of less than 1k casts each).

So has anyone else tested this or am I overlooking something? Our feral druids are not always good about keeping up faerie fire. I didn't take the talent to increase my (or the raid's) hit chance on a target. At this point it seems better to spec out of it, since it clearly does not seem to be giving me any benefit.

If this isn't the appropriate forum to post this, my apologies.

--------

Originally Posted by klüger

Two things: This is more a question for the dummies thread, and this has been answered by blues and the verdict is that you get the crit bonus as long as there is a faery fire on the mob (no matter who casts it or how talented it is).

-------

Originally Posted by triarchelon

klüger -

Did you even look at the test results? They would disagree with what blues have posted.

The bottom part of my post got cut off where I discussed whether it might be worthwhile to keep Imp FF as long as another person was putting up a Faerie Fire, or whether it was a complete waste since I can't rely on them to put it up. (We always have a shadow priest or actually two in raids so don't need it for hit.) If I forego this, I have 2 points and I'm not sure which is more helpful in raids. I am trying to decide between putting one point in Owlkin, Brambles, or getting the third point in Eclipse. I probably don't get as great of use of my trees as I could, but I am working on improving it. As a sidenote, ignore my current spec if you look me up on armory. I am planning to change things around as soon as I have enough test data to show that Imp FF isn't working properly.

I'll go post this in the thread you suggested. I don't know how to just move it entirely.

Najtrok · 02/10/09, 5:37 AM

I think your dimesion of tested numbers is to low. Its like throwing a dice about ten times and complaining that its not an equl chance.

You have to get by far larger numbers (witch will include hours of testing and regeneration), until your numbers are big enough. Go check your crit chance for say wrath and then gather a huge number of spells until you see it goes near that number. Now gather the same amount with FF up and see if it varies.

2k hits are just too few.

Edit: For you last lines: 2/3 IFF is senseless, either come to conclusion it is worth or it is not.
Getting Eclipse full should also be your goal. If you dont like it and want just SF spam or so, skip it, but I think it's gonna be a great dmg loss.

Brambles sucks if you want to get it instead of Eclipse or IFF and so does owlkin (in PvE). And you should rely on FF casters, if they are feral. I am not entirely sure but afaik does FF produce a good aggro for a bear tank and a cat will get a good dmg boost (including its meele companions). So if there are a feral and an SP, dont bother keeping IFF up, they will for their own good.

Ashaera · 02/10/09, 5:41 AM

@Triarchelon -> Its been to long since I had math at any serious level, so cant give an estimate over how accurate your findings are, Id dare claim that they are far from conclussive though.

If you really think you are right then I suggest that you find a friend with resilience gear, have him adjust so you critting without a bonus from Imp FF would be impossible & spamnuke him to find out if you can force a crit.

----
& get that 3rd point in Eclipse before you start worrying about imp FF.

• Hamlet · 02/10/09, 6:11 AM

The standard error in measurement of a binomial process (such as crit rate) goes like the square root of the number of "hits" (roughly--this is an approximation for low p).

At 550-600 hits, you're looking at an error of +/- 24. Out of 2000 trials, that's over 1%--uncomfortably large compared to the effect you're trying to observe.

triarchelon · 02/10/09, 6:23 AM

I actually have the 3rd point in Eclipse right now, but was planning to drop one point of it when I respecced. Guess I won't. Thanks.

I had a bunch of testing done with 3/3 Imp FF but didn't have it saved. Since I started the testing with 2/3 I left it this way until I was done so I wouldn't have to re-do the tests. The previous testing was an even bigger difference in crit % than this was (I was gaining more if someone else put up the FF, and losing even more if I put Imp FF up.

What is a reasonable sample size? I see the difference if I remove a 1.18% crit trinket, so it seems like I should see the 2% increased crit from the debuff. I suspect it will be more difficult now to find someone to put up their FF for me, but I can redo with wrath since it takes less time, and go for a larger number of casts. My guildies already think I am nuts. This has already been hours of testing and regen of mana and testing more. It's not just 2k casts. but 2k casts with no fairie fire up, 2k casts with my faerie fire up, and 2k with another ff up, plus about 4-6k more that I lost when I was 3/3 Imp FF.

Not sure if I have anyone with resilience gear that would remove x amount to make them uncrittable, and then stand there with me trying to get a crit for however long is necessary. This would also require 2 people other than myself to do testing as opposed to just me and one other person. Is there a reason people go this route with testing? How long do I cast trying to get a crit before I give up?

• Hamlet · 02/10/09, 6:44 AM

Originally Posted by triarchelon

I actually have the 3rd point in Eclipse right now, but was planning to drop one point of it when I respecced. Guess I won't. Thanks.

I had a bunch of testing done with 3/3 Imp FF but didn't have it saved. Since I started the testing with 2/3 I left it this way until I was done so I wouldn't have to re-do the tests. The previous testing was an even bigger difference in crit % than this was (I was gaining more if someone else put up the FF, and losing even more if I put Imp FF up.

What is a reasonable sample size? I see the difference if I remove a 1.18% crit trinket, so it seems like I should see the 2% increased crit from the debuff. I suspect it will be more difficult now to find someone to put up their FF for me, but I can redo with wrath since it takes less time, and go for a larger number of casts. My guildies already think I am nuts. This has already been hours of testing and regen of mana and testing more. It's not just 2k casts. but 2k casts with no fairie fire up, 2k casts with my faerie fire up, and 2k with another ff up, plus about 4-6k more that I lost when I was 3/3 Imp FF.

Not sure if I have anyone with resilience gear that would remove x amount to make them uncrittable, and then stand there with me trying to get a crit for however long is necessary. This would also require 2 people other than myself to do testing as opposed to just me and one other person. Is there a reason people go this route with testing? How long do I cast trying to get a crit before I give up?

Part of it is that it's such a weird result that people would need to see something pretty solid. I'd make sure you get someone with some statistics clue to help you set up all the details before you spend hours doing anything.

If you can go gearless against a high-resilience target who would be uncrittable, that would be a great way to test whether you were getting the 3% bonus.

Ashaera · 02/10/09, 9:54 AM

Originally Posted by triarchelon

This would also require 2 people other than myself to do testing as opposed to just me and one other person. Is there a reason people go this route with testing? How long do I cast trying to get a crit before I give up?

Since you want to prove that Imp FF doesnt give crit when you apply Faerie Fire it would only involve 1 friend. Anyone with high enough resiliences & preferably a healer. If you with gear+spec manage to land your crit chance at 2.5%->2.9% & dont produce a crit in 1000 casts then I suggest reporting it as a bug. Suggested test spec
You are trying to prove something that seems very odd, so I suggest that you take this approach instead of doing x thousand casts to show the difference between 30 & 32.

Erdluf · 02/10/09, 11:16 AM

Tria,

If your expected crit rate with 2/3 IFF was 28.2% (which is what your current armory implies) then you would have about a 0.96% of getting 580 (or fewer) crits out of 2234 trials.

That is really bad luck, but is not a "win the lottery" event. Since that is over an hour of casting (not counting mana breaks), I'm a little concerned (without seeing a log) that FF may have been down for some of those casts.

I can certainly imagine a coding bug that gives you more crit when FF is on the target, but not when IFF is on the target. Since most players aren't going to test thousands of casts, such a bug might go unreported.

I'd recommend you post your numbers in the bug report forums. If a bug like I listed is there, blizz can probably find and fix it quickly once they look for it.

For that 0.96% number I used Excel: =BINOMDIST(580,2234,28.2%,TRUE)

triarchelon · 02/10/09, 12:28 PM

Thanks. I will look for a friend and respec this. I never let FF fall off the mob when testing my own FF, though I think it did 3 or 4 times when someone else applied it. That would not go against what I'm trying to prove though, since their FF should be at best equal to me applying it, not better. My gear was slightly different than what you see in armory. I received a couple of upgrades after I started the testing and I did not test with Sundial of the Exiled equipped. I didn't log out in my testing gear last night.

I will look for someone patient enough to let me run this and I'll use Wrath instead since it is faster. I can also save the combat logs for anyone who doubts the results.

Thanks for the feedback everyone.

Erdluf · 02/10/09, 2:41 PM

Originally Posted by triarchelon

Thanks. I will look for a friend and respec this. I never let FF fall off the mob when testing my own FF, though I think it did 3 or 4 times when someone else applied it. That would not go against what I'm trying to prove though, since their FF should be at best equal to me applying it, not better. My gear was slightly different than what you see in armory. I received a couple of upgrades after I started the testing and I did not test with Sundial of the Exiled equipped. I didn't log out in my testing gear last night.

I will look for someone patient enough to let me run this and I'll use Wrath instead since it is faster. I can also save the combat logs for anyone who doubts the results.

Thanks for the feedback everyone.

To show that you aren't benefitting from your own IFF casts, you won't need a friend. Also, to prove that you are not getting a crit buff, you are better off if you are at a low crit%, and of course it is better to have lots of tests.

You may want to use hurricane for your tests, in caster (not moonkin) form. Hurricane has a lower crit rate (no Majesty benefit) and hits faster than Wrath. If you can fit three target dummies in a single hurricane, then

IFF * 3 + Hurricane*3

is 90 hits in under 40 seconds. With luck on Omen procs, mana may not be too terrible.

In this table, the top row is the number of tests in a trial. The left column is the "expected" crit rate. The rest is the likelyhood of seeing 2% less than the "expected" crit rate (or worse) for a particular trial. Just 500 tests at 6% "expected" crit is better "proof" than 2000 tests at 24% "expected". For instance, it says that if your expected crit rate is 12%, and you cast 100 Wraths, there is about a 1/3 chance that you will see 10 or fewer crits.

	100	500	1000	2000
3%	4.755%	0.075%	0.001%	0.000%
6%	14.302%	1.898%	0.199%	0.003%
9%	19.398%	4.625%	0.976%	0.053%
12%	33.369%	7.128%	2.092%	0.218%
15%	34.743%	11.572%	4.014%	0.598%
18%	35.711%	13.371%	5.263%	0.998%
21%	36.423%	14.819%	6.372%	1.429%
24%	36.959%	15.992%	7.336%	1.859%
27%	37.368%	16.944%	8.161%	2.267%

spiritryu · 02/10/09, 10:18 PM

I've been wondering for a while, and after a bit of work to finally figure out why I couldn't search Glyph of Focus, I've been wondering what was the value of Glyph of Focus for single target DPS vs maybe Glyph of Insect Swarm?
I remember you posted it being a possible good glyph, but that was before it was released and after checking it, it isn't on the list of Glyphs anymore.

Anyways, love your guide. Great info. =] Thx for all the great work.

EDIT: Oh and would you think the Glyph of Focus would become a better glyph should they make Starfall scale better with SP?

Ashaera · 02/11/09, 10:25 AM

I removed it when I found out it wasnt a minor glyph -- For a major glyph its absolutely horrible.

My last Patchwerk as example (2:33 - 1 Starfall cast):
Starfall = 11642 -- Assuming best case scenarion 11642*1.2 = 13970
Damage gained: 2328 pr 3minutes -- Now I was perhaps 5% crit below average so lets assume 2.5k damage pr 180 seconds & we land on 13.89dps.

IS = 45756 -- Im unsure to how the glyph adds its damage, so this is just pure guessing to illustrate that no matter what Glyph of Focus is not likely to be close. 45756/130*100 = 35197.
Damage gained: 10559 over 153 seconds. Adding 69dps, allthough I suspect my value of how it adds up is over the top this is still far ahead.

Naturally Starfal glyph will give more in the average fight but never even close to the IS one.