New Darkmoon Cards - Page 9

Charlatan · 07/09/07, 7:52 AM

I just got my Crusade trinket this morning, and in testing, I noticed a strange thing. I'll be attacking something, and getting the buff, when all of a sudden, I'll get a second icon for the buff, with the count on that buff starting at zero. So, for instance, I might have 5 stacks counting down, and 2 stacks counting down. Sometimes they have different timers (i.e., the first stack will be at 4 seconds while the second stack is at 6 seconds) but oftentimes they have the same countdown time. I was thinking it might happen when I attack different mobs, but no, I've seen this happen when attacking one mob (got a stack of 5 and then a stack of 1, same countdown timer). I've never seen the combined number of stacks exceed 10, so I don't think it's a bug, but the one thing that strikes me a bit strange is that since these appear to be distinct stacks of the buff, and the refreshing works on the newer stack, the older stack will expire (meaning I won't have a full stack of 10 for part of my fights).

I've not heard people describe this, so it struck me as a bit weird. Anyone know what's going on here?

songster · 07/09/07, 8:05 AM

Originally Posted by Carlaena

But does the 'only stacks till 20' still apply? It's been removed from the description and replaced by a timer. In theory, could I keep stacking it till i hit 100% crit?

How long have you got? To get it stacked to 100% will on average take approximately 10000000000000000000000000000 times the age of the Universe.

Calculation: you require 104 stacks to get from 20% crit to 100% crit. Odds of doing so can be approximated as (0.4^104) = 4 x 10^-42. At one hit per second, the average streak will last about 5 seconds (= 5 hits at ~20%).

4 x 10^42 x 5 seconds = 2 x 10^43 seconds.

That's ~3.65 x 10^38 years. Age of the Universe is about 1.3 X 10^10 years.

So, a ballpark estimate is that to stack your crit to 100% would take about 10^28 times the age of the Universe.

It's a very ghetto estimate - anyone with the leet spreadsheet skills to do better? I'm probably only accurate to within a few orders of magnitude.

Carlaena · 07/09/07, 8:42 AM

Originally Posted by songster

How long have you got? To get it stacked to 100% will on average take approximately 10000000000000000000000000000 times the age of the Universe.

Calculation: you require 104 stacks to get from 20% crit to 100% crit. Odds of doing so can be approximated as (0.4^104) = 4 x 10^-42. At one hit per second, the average streak will last about 5 seconds (= 5 hits at ~20%).

4 x 10^42 x 5 seconds = 2 x 10^43 seconds.

That's ~3.65 x 10^38 years. Age of the Universe is about 1.3 X 10^10 years.

So, a ballpark estimate is that to stack your crit to 100% would take about 10^28 times the age of the Universe.

It's a very ghetto estimate - anyone with the leet spreadsheet skills to do better? I'm probably only accurate to within a few orders of magnitude.

You miss my point.

Peoples calculations of the wrath trinket (that I've seen so far) are worked out on the basis that the buff stacks up to this glass ceiling of 20. Hitting 20 times without a crit is not as unlikely as it may seem, but does it stop there?

To calculate the effect of this trinket correctly, do we now need to take into account its cumulative effect on your base crit rate, all the way up to 100% crit rate? (the only measurable points where we know this trinket will have no effect).

Or does it in fact still only stack the previously stated 20 times, which renders my posts both pointless and troublesome

I'm no good at math this complex, just trying to stimulate those that are

Antarius · 07/09/07, 8:53 AM

Originally Posted by songster

It's a very ghetto estimate - anyone with the leet spreadsheet skills to do better? I'm probably only accurate to within a few orders of magnitude.

Using "leet spreadsheet skills" to multiply .8 * .79 * .78 * .77 ...... *.02 *.01

I received an answer of 7.16 * 10 ^ -42

Assuming the base 20% crit rate and a maximum of 20 procs, the chance of actually making it to the 20 stack is only 0.0516 %

At a 25% crit rate, that moves to 0.0107 %

Or said another way: approximatly 1 out of every 10,000 crits will require the full 20 stack.

Docjowles · 07/09/07, 8:57 AM

For what it's worth, the Thottbot entry appears to include data-mined info on how many times a spell can stack. It specifies 20 times for the Wrath card as of this morning. Whether or not that's up to date, I don't know; Wowhead doesn't seem to include stacking information at all.

I also have to agree with songster. He just showed that "hitting 20 times without a crit" when you have a stacking +crit mod and non-zero base chance really is as unlikely as it seems.

Cadfael · 07/09/07, 9:13 AM

Originally Posted by Charlatan

I just got my Crusade trinket this morning, and in testing, I noticed a strange thing. I'll be attacking something, and getting the buff, when all of a sudden, I'll get a second icon for the buff, with the count on that buff starting at zero. So, for instance, I might have 5 stacks counting down, and 2 stacks counting down. Sometimes they have different timers (i.e., the first stack will be at 4 seconds while the second stack is at 6 seconds) but oftentimes they have the same countdown time. I was thinking it might happen when I attack different mobs, but no, I've seen this happen when attacking one mob (got a stack of 5 and then a stack of 1, same countdown timer). I've never seen the combined number of stacks exceed 10, so I don't think it's a bug, but the one thing that strikes me a bit strange is that since these appear to be distinct stacks of the buff, and the refreshing works on the newer stack, the older stack will expire (meaning I won't have a full stack of 10 for part of my fights).

I've not heard people describe this, so it struck me as a bit weird. Anyone know what's going on here?

Are you wanding between spell cast ? If so, your wanding proc's the melee/ranged part which stacks up to 20 and gives AP (not much worth to you I guess) while your spells proc' the spell part which stacks up to 10. (Just mouse over the buffs to confirm)

The Crusader card effectively has 2 entirely different proc-buffs, and both can stack independently from each other. I read somewhere that this is really cool for Paladins since they do spellcast and melee a lot while farming but I can't place the value of the card for them.

For casters this is relatively useless, as we're interested in the spell damage being up and nothing else.

snape · 07/09/07, 9:15 AM

Originally Posted by Charlatan

I just got my Crusade trinket this morning, and in testing, I noticed a strange thing. I'll be attacking something, and getting the buff, when all of a sudden, I'll get a second icon for the buff, with the count on that buff starting at zero. So, for instance, I might have 5 stacks counting down, and 2 stacks counting down. Sometimes they have different timers (i.e., the first stack will be at 4 seconds while the second stack is at 6 seconds) but oftentimes they have the same countdown time. I was thinking it might happen when I attack different mobs, but no, I've seen this happen when attacking one mob (got a stack of 5 and then a stack of 1, same countdown timer). I've never seen the combined number of stacks exceed 10, so I don't think it's a bug, but the one thing that strikes me a bit strange is that since these appear to be distinct stacks of the buff, and the refreshing works on the newer stack, the older stack will expire (meaning I won't have a full stack of 10 for part of my fights).

I've not heard people describe this, so it struck me as a bit weird. Anyone know what's going on here?

Did you check your character screen and see if the buff did actually stack in your bonus damage number?

songster · 07/09/07, 10:27 AM

Originally Posted by Carlaena

You miss my point.

No, you miss mine.

Originally Posted by Carlaena

To calculate the effect of this trinket correctly, do we now need to take into account its cumulative effect on your base crit rate, all the way up to 100% crit rate?

No. Taking into account stacks up to 20 gets you something like 5 decimal places of accuracy. How much more accurate do you want?

In fact, we can turn this one round and ask "How many stacks are relevant?" Let's say that to be relevant it has to add at least 0.1% to your DPS - i.e. at least 1 DPS to someone who's doing 1000 DPS.

Assuming that a 1% increase in crit rate is very roughly equal to a 1% increase in DPS, the increased DPS contributed by allowing a stack size of n in your model is:

p(n) * 0.77% * n

That is, the probability of stacking up to n, times the 0.77% added crit you get from increasing the stack from n-1 to n.

Modelling the first stack (80% chance) increases your DPS by (80% * 0.77% = 0.616%)
Modelling a second stack (79.25% chance) increases your DPS by a further (79.25% * 0.77% = 0.488%)

The table plays out thusly:

Stack 	Cumulative 	%age DPS 		Cumulative 
size	chance to stack	contributed by 	added DPS (%)
			marginal increase 	from stack
			of stack size	of size n
0			
1	0.8		0.616		0.616
2	0.63384		0.4880568		1.1040568
3	0.497310864	0.382929365	1.486986165
4	0.38636081	0.297497824	1.784483989
5	0.297188735	0.228835326	2.013319315
6	0.226309222	0.174258101	2.187577416
7	0.170591891	0.131355756	2.318933173
8	0.12727861	0.09800453	2.416937702
9	0.093982526	0.072366545	2.489304247
10	0.068673032	0.052878234	2.542182482
11	0.049650602	0.038230963	2.580413445
12	0.035515075	0.027346608	2.607760053
13	0.025130467	0.01935046	2.627110513
14	0.017588814	0.013543387	2.6406539
15	0.012174977	0.009374732	2.650028632
16	0.008333772	0.006417004	2.656445637
17	0.005640297	0.004343029	2.660788665
18	0.003773923	0.00290592	2.663694586
19	0.002496072	0.001921976	2.665616561
20	0.001631683	0.001256396	2.666872957
21	0.001054067	0.000811632	2.667684588
22	0.000672811	0.000518064	2.668202653
23	0.000424275	0.000326691	2.668529344
24	0.000264281	0.000203496	2.66873284
25	0.000162585	0.000125191	2.668858031
26	9.87707E-05	7.60534E-05	2.668934085
27	5.92426E-05	4.56168E-05	2.668979701
28	3.50776E-05	2.70097E-05	2.669006711
29	2.04993E-05	1.57845E-05	2.669022496
30	1.1822E-05	9.10291E-06	2.669031598
31	6.7267E-06	5.17956E-06	2.669036778
32	3.7757E-06	2.90729E-06	2.669039685
33	2.09022E-06	1.60947E-06	2.669041295
34	1.14105E-06	8.78611E-07	2.669042173
35	6.14115E-07	4.72869E-07	2.669042646
36	3.25788E-07	2.50857E-07	2.669042897
37	1.70322E-07	1.31148E-07	2.669043028
38	8.77329E-08	6.75543E-08	2.669043096
39	4.45157E-08	3.42771E-08	2.66904313
40	2.22445E-08	1.71282E-08	2.669043147

From this we observe:

1) The total DPS increase from the trinket is about 2.7% if your base crit rate is 20%

2) Modelling anything more than 10 stacks is a waste of time. The probability of getting that many stacks is so low that it won't even add 1 DPS to your total

Carlaena · 07/09/07, 11:25 AM

{deleted}

Carlaena · 07/09/07, 11:34 AM

Originally Posted by songster

Stack 	Cumulative 	%age DPS 		Cumulative 
size	chance to stack	contributed by 	added DPS (%)
			marginal increase 	from stack
			of stack size	of size n		
11	0.049650602	0.038230963	2.580413445
12	0.035515075	0.027346608	2.607760053
13	0.025130467	0.01935046	2.627110513
14	0.017588814	0.013543387	2.6406539
15	0.012174977	0.009374732	2.650028632
16	0.008333772	0.006417004	2.656445637
17	0.005640297	0.004343029	2.660788665
18	0.003773923	0.00290592	2.663694586
19	0.002496072	0.001921976	2.665616561
20	0.001631683	0.001256396	2.666872957
21	0.001054067	0.000811632	2.667684588
22	0.000672811	0.000518064	2.668202653
23	0.000424275	0.000326691	2.668529344
24	0.000264281	0.000203496	2.66873284
25	0.000162585	0.000125191	2.668858031
26	9.87707E-05	7.60534E-05	2.668934085
27	5.92426E-05	4.56168E-05	2.668979701
28	3.50776E-05	2.70097E-05	2.669006711
29	2.04993E-05	1.57845E-05	2.669022496
30	1.1822E-05	9.10291E-06	2.669031598
31	6.7267E-06	5.17956E-06	2.669036778
32	3.7757E-06	2.90729E-06	2.669039685
33	2.09022E-06	1.60947E-06	2.669041295
34	1.14105E-06	8.78611E-07	2.669042173
35	6.14115E-07	4.72869E-07	2.669042646
36	3.25788E-07	2.50857E-07	2.669042897
37	1.70322E-07	1.31148E-07	2.669043028
38	8.77329E-08	6.75543E-08	2.669043096
39	4.45157E-08	3.42771E-08	2.66904313
40	2.22445E-08	1.71282E-08	2.669043147

Very interesting reading. Thanks for your time and effort.

Is there any chance you've made a small error in the 'cumulitive chance to stack percentages'?

I calculated the % chance if getting stacks 1-20 (with my limited mathematical and spreadsheet ability). My numbers match yours exactly up to 11stacks = 0.049650602. After that yours spiral down a lot faster that I calculate they should.

I get:
12 = 0.038955862
13 =0.030564769
14 = 0.023981118
15 = 0.018815585
16 = 0.014762708
17 = 0.011582821
18 = 0.009087881
19 = 0.007130352
20 = 0.005594474

I've no idea if these would make much difference of the effectiveness of the trinket though. I'm no whizz at maths, so may have made an error myself.

songster · 07/09/07, 11:38 AM

Originally Posted by Carlaena

Going by the numbers that Schizzle just posted up, every time you crit(and the stack resets) you have a 1.631% of going on to reach a 20 stack of trinket buffs.

Yep, looks like Antarius was using the quick and dirty estimate of 1% crit chance per stack - it's actually 17/22.1 ~= 0.77% per stack. That gives you a very slightly higher chance of getting to a 20 stack. Doesn't change the fact that it still won't even add an average of 1DPS to your overall output.

Look at the overall lessons: the trinket does not add much DPS, and mathematical modelling of anything deeper than a 10-stack doesn't give you any real improvement in accuracy. It is an extremely poor item, and only useful to people with a fast attack speed facing a high-resilience opponent. The change from a max stack of 20 to unlimited stacking, if it's even real, is irrelevant in every way.

Antarius: The "leet spreadsheet skills" I was after were the ones to transform the probability of stacking up to 100% into the expected time. I used a very dodgy estimate of about 5 seconds per non-crit streak. To accurately sum up the expected time exceeds my abilities. But I'm quite confident that it's many times the age of the Universe, so it's not really relevant...

songster · 07/09/07, 11:44 AM

Originally Posted by Carlaena

I get:
12 = 0.038955862
13 =0.030564769
14 = 0.023981118
15 = 0.018815585
16 = 0.014762708
17 = 0.011582821
18 = 0.009087881
19 = 0.007130352
20 = 0.005594474

I've no idea if these would make much difference of the effectiveness of the trinket though. I'm no whizz at maths, so may have made an error myself.

There's a mistake somewhere - the difference between each of these is a factor of about 0.78 (i.e. a crit rate of ~22%). That corresponds to only about 3 stacks up from the base rate of 20, and it's not increasing from stack to stack. Check your formulas :-)

Carlaena · 07/09/07, 11:46 AM

I'm going to go back to my usual lurking status. Made a mess of my numbers, big time.

Looks like you're right. I wasted 400g on a crap trinket

I'll start saving badges for bloodlust broach.

Charlatan · 07/09/07, 2:00 PM

Originally Posted by Cadfael

Are you wanding between spell cast ? If so, your wanding proc's the melee/ranged part which stacks up to 20 and gives AP (not much worth to you I guess) while your spells proc' the spell part which stacks up to 10. (Just mouse over the buffs to confirm)

The Crusader card effectively has 2 entirely different proc-buffs, and both can stack independently from each other. I read somewhere that this is really cool for Paladins since they do spellcast and melee a lot while farming but I can't place the value of the card for them.

For casters this is relatively useless, as we're interested in the spell damage being up and nothing else.

Aha! I bet I was auto-attacking with my sword just out of habit. Yeah, I suspect I was seeing the melee *and* the spell buffs (they go away so quickly it's hard to determine).

Thanks for the input (and thanks to snape too for the brainstorming).

eludebear · 07/10/07, 11:53 AM

[Darkmoon Card: Wrath] (Ace - Eight of Storms)
Each time one of your direct damage attacks does not critically strike, you gain 17 critical strike rating and 17 spell critical strike rating for the next 10 sec. This effect is consumed when you deal a critical strike. (Stacks up to 20 times)

I saw a member of our guild with this earlier today, I had never really looked up these cards today, but from a pure arena stand point (from a rogue) this thing is amazing, looking back at when I did 3v3/5v5 on my rogue with around 38-40% crit with a mongoose or two up, I can remember countless times attacking a priest or warlock in full Resilience gear and not getting any crits which made me frustrated many a times. Having this trinket would have been nice to counter that.

songster · 07/10/07, 2:23 PM

Originally Posted by eludebear

[Darkmoon Card: Wrath] (Ace - Eight of Storms)
Each time one of your direct damage attacks does not critically strike, you gain 17 critical strike rating and 17 spell critical strike rating for the next 10 sec. This effect is consumed when you deal a critical strike. (Stacks up to 20 times)

I saw a member of our guild with this earlier today, I had never really looked up these cards today, but from a pure arena stand point (from a rogue) this thing is amazing, looking back at when I did 3v3/5v5 on my rogue with around 38-40% crit with a mongoose or two up, I can remember countless times attacking a priest or warlock in full Resilience gear and not getting any crits which made me frustrated many a times. Having this trinket would have been nice to counter that.

Did you not read any of the last several pages of this thread? You're one of the few cases where the trinket is any use at all (fast attacks, high resilience opponent). However even then I suspect it'll turn out better to get an AP trinket.

Consider: if they have enough resilience to significantly nerf your crit rate, they also have enough resilience that your crits do significantly less damage when they land. Also, the stacking of this trinket relative to your base crit rate is just too slow. You have to have a lot of misses in a row to get a measurable benefit out of it.

The only way to fix this trinket would be to make it stack much higher, much faster, and put some kind of cooldown on it. A melee version of Combustion, in other words.

eludebear · 07/10/07, 2:47 PM

Originally Posted by songster

Did you not read any of the last several pages of this thread? You're one of the few cases where the trinket is any use at all (fast attacks, high resilience opponent). However even then I suspect it'll turn out better to get an AP trinket.

Consider: if they have enough resilience to significantly nerf your crit rate, they also have enough resilience that your crits do significantly less damage when they land. Also, the stacking of this trinket relative to your base crit rate is just too slow. You have to have a lot of misses in a row to get a measurable benefit out of it.

The only way to fix this trinket would be to make it stack much higher, much faster, and put some kind of cooldown on it. A melee version of Combustion, in other words.

It wasnt so much just my crit for damage as it was my spec at the time. My spec was 30/31, Seal fate adrenaline rush with claw spec, which worked out amazing in arena when I actually got crits, when my SS's werent critting I wasnt able to put up Exposes and kidney shot and do as much stuff As I was wanting to, which was the annoying part.

Rane · 07/10/07, 7:48 PM

Originally Posted by Charlatan

I just got my Crusade trinket this morning, and in testing, I noticed a strange thing. I'll be attacking something, and getting the buff, when all of a sudden, I'll get a second icon for the buff, with the count on that buff starting at zero. So, for instance, I might have 5 stacks counting down, and 2 stacks counting down. Sometimes they have different timers (i.e., the first stack will be at 4 seconds while the second stack is at 6 seconds) but oftentimes they have the same countdown time. I was thinking it might happen when I attack different mobs, but no, I've seen this happen when attacking one mob (got a stack of 5 and then a stack of 1, same countdown timer). I've never seen the combined number of stacks exceed 10, so I don't think it's a bug, but the one thing that strikes me a bit strange is that since these appear to be distinct stacks of the buff, and the refreshing works on the newer stack, the older stack will expire (meaning I won't have a full stack of 10 for part of my fights).

I've not heard people describe this, so it struck me as a bit weird. Anyone know what's going on here?

I just got the trinket as well yesterday, and I noticed another peculiarity: Kill Command triggers the spelldamage proc of the trinket, which means with a bit of luck I can get the trinket to run with 5-6 stacks (50-60 spelldamage). Sadly I didn't have my Scorpid pet with me, but that may just be a a very nice boost to its damage output using Scorpid Poison.

Lazhar · 07/11/07, 6:22 AM

During several tries on kael'thas i noticed Darkmoon Card: Vengeance proccing on
unusual events: on thaladred it would proc with me way out of range. sometimes it was right at
a gaze switch, at other times it was during a gaze (in neither case was i the target).
on sanguinar (who i did not tank) it would proc on me attacking, not me being attacked.

i do not know if it's a bug or due to an internal spell, just wanted to add this info.

snape · 07/11/07, 12:32 PM

It's probably some pulse put out by the target - think Ouro in the old days (Magic Absorption specced Mages would get ticks of 5% regen even when no quakes were anywhere near them).

Irshish · 07/11/07, 1:58 PM

Originally Posted by songster

Did you not read any of the last several pages of this thread? You're one of the few cases where the trinket is any use at all (fast attacks, high resilience opponent). However even then I suspect it'll turn out better to get an AP trinket.

Consider: if they have enough resilience to significantly nerf your crit rate, they also have enough resilience that your crits do significantly less damage when they land. Also, the stacking of this trinket relative to your base crit rate is just too slow. You have to have a lot of misses in a row to get a measurable benefit out of it.

The only way to fix this trinket would be to make it stack much higher, much faster, and put some kind of cooldown on it. A melee version of Combustion, in other words.

Not to say Wrath is a good trinket, but has anyone done any calculations taking into account talents and abilities that trigger off of crits? For some classes a crit isn't just the damage of the crit.
* For feral druids and rogues it is an additional combo point.
* For a BM Hunter it would be a Kill Command AND 50 pet energy, which with Focused Fire has a good chance of leading to a Pet crit, which in-turn would trigger Ferocious Inspiration increasing the party's damage by 3% for 10 seconds.
* For a Survival hunter a crit would mean triggering Expose Weakness increasing physical DPS AND replenishing mana.
* Other classes with crit procs that I don't now about.

Initially I was just wondering how much a difference Primal Fury w/ Wrath would make, but looking at the things a hunter has going on with procs on crits, makes me wonder if the trinket might actually be beneficial to them to insure more consistent crits for procs.

Regan_ · 07/11/07, 2:08 PM

Originally Posted by Irshish

Not to say Wrath is a good trinket, but has anyone done any calculations taking into account talents and abilities that trigger off of crits? For some classes a crit isn't just the damage of the crit.
* For feral druids and rogues it is an additional combo point.
* For a BM Hunter it would be a Kill Command AND 50 pet energy, which with Focused Fire has a good chance of leading to a Pet crit, which in-turn would trigger Ferocious Inspiration increasing the party's damage by 3% for 10 seconds.
* For a Survival hunter a crit would mean triggering Expose Weakness increasing physical DPS AND replenishing mana.
* Other classes with crit procs that I don't now about.

Initially I was just wondering how much a difference Primal Fury w/ Wrath would make, but looking at the things a hunter has going on with procs on crits, makes me wonder if the trinket might actually be beneficial to them to insure more consistent crits for procs.

I've always wondered if Wrath is worth it regarding Improved Shadow Bolt uptime.

snape · 07/11/07, 2:15 PM

I don't think any amount of Crit synergy can push the card into the realm of viability. A Warlock would get far more use out of the Blessings/Crusade deck.

vyedma · 07/11/07, 2:35 PM

I typically just think of Wrath in terms of its average effect. For a caster with about 20% crit rating it will on average give you a boost of about 1.5% crit (~33 crit rating). Compared to other trinkets it's not all that great. For example, Xi'ri's Gift will give you 32 crit rating and an on use effect.

Obviously it gets a little worse for higher crit and better for lower crit. And you may get lucky/unlucky streaks, but over the course of a 10min boss fight, or over the course of a 2-4 hour raid I'd much rather have a more steadily reliable trinket.

snape · 07/11/07, 3:07 PM

Well, it IS steadily reliable, it's just reliably bad. Like driving a Civic, it's very consistent, but I'd rather have the Mercedes, which is consistent and much higher quality.