Thundering Skyfire Diamond - Page 5

• Aldriana · 05/30/07, 2:39 PM

So, I went and tested this for a while last night, and I noticed something interesting:

At no time during testing did a proc ever refresh itself, nor did consecutive procs ever occur less than 40 seconds apart (shortest gap between two procs was 40.5 seconds). Now, admittedly I only had about 35 total procs during my testing, so it's possible I just got unlucky - but it seems to me that there might be a hidden cooldown on the proc of perhaps 30 seconds. Has anyone ever seen a proc refresh or two procs closer together than this?

Anyway, assuming the proc does have a 30 second cooldown on it, I went through and analyzed the proc rate, and found that with two weapons of rather different speeds, the proc rate was about 3% for both. So I'm wondering if this is perhaps a fixed % proc with a hidden cooldown rather than a PPM. Can anyone confirm/deny?

Shaker · 05/30/07, 3:10 PM

Which is similiar to findings found before, so PPM is same for melee and range it seems. Now about hidden cooldown, it never what I saw refreshed while still active but it has procced 9secs from each other (shortets one I have noticed) so if there is one, it is not very long.

Post 57 above, 5000 autoshots, .98 ppm, 161 procs.

• Aldriana · 05/30/07, 3:27 PM

Ah, missed that somehow. Guess I just got unlucky.

sp00n · 05/30/07, 3:31 PM

Originally Posted by Kabuto

Sorry about that was a bit sleepy.

"Did your "poor weapon" crit on average for 2833" - No.
"Did you have the same helm equipped, just without the diamond while examing non diamond damage?" - First batch of data was with a 24ap diamond. Second set with the Relentless.

Yeah, I definitely missed out a few factors in that basic run through. However, the bonuses to crit damage are accurate. As stated the huge damage bonus must be attributed to the relative weakness of said mobs in blasted lands.

So if your average crit wasn't 2833, but the bonus damage of 85 is accurate, then something has to be wrong with the 3% damage bonus in the tool tip.
Either that or I missed something in your posting.

You don't have a combatlog for your testings anymore by chance?

Zeblök · 05/30/07, 4:03 PM

Gave me (not a combat rogue) on blastedlands mobs:

7,5% optimal upptime and a 1.09% procrate
on 2009 autoswings inc (miss 169, 8,41%.)
With Spiteblade and Searing sunblade (dual mongoose)

Now i have no ide how a dps cyckle would affect this.
Would love some help on that

.

In addition, unraveler seams to have the allmoust the same upptime (7,1%) with my (critt 446, 22,20%)

Sarj · 05/30/07, 4:26 PM

Alright, posting my data from a little over an hours worth of Blasted Land killing, with Procwatch I'm showing a 0.8 PPM. Not once through the entire amount of time did I have procs come closer together than 39 seconds. Meaning to me that there is a 45 second internal cooldown. For this to be correct that would mean that the cooldown starts as soon as the proc has happened (45-6=39).

The first test I did was 38 minutes and 17 seconds with a Dragonmaw (Mongoose) and a Gladiator's Shanker (Potency). 39 seconds in between the closest procs and 1 minute 56 seconds in between the longest. The average proc was 48 seconds apart.

Second test (have a screenshot for this one) was 18 and a half minutes long using a Malchazeen (no enchant) and a Gladiator's Shanker (Potency).

http://www.imagedump.com/index.cgi?pick=get&tp=502818

Third test (SS of this one as well) 16 minutes 25 seconds with Dragonmaw (Mongoose) and Shanker (Potency). Didn't actually write procs down this time around as they came up but I figured I'd provide another SS for consistency.

http://www.imagedump.com/index.cgi?pick=get&tp=502820

Anyways, what I and Cili (Rogue that also was testing it with me) have come up with is that it is a 45 second internal timer and the procs per minute are 0.8.

(Sorry for the pictures, I couldn't get them to pull up with "Insert Image.")

Edit : I also wanted to note that one of the reasons the person who posted above might have a 1.6 PPM is because he or she was not only showing Gains for the haste. As such it was counting both the gain and fade of the combat log, equaling 0.8 PPM when cut in half.

Zeblök · 05/30/07, 5:26 PM

Is that me who gotten 1.6ppm ?
In that case, eh how is ppm calculated, if not #procs/min , i get 0.75procs / min
but it might not be the same

River · 05/30/07, 8:47 PM

Originally Posted by Sarj

Edit : I also wanted to note that one of the reasons the person who posted above might have a 1.6 PPM is because he or she was not only showing Gains for the haste. As such it was counting both the gain and fade of the combat log, equaling 0.8 PPM when cut in half.

I tested only the gain, the test string was "gain Skyfire Swiftness" off memory.

Edit: And there was no one around besides my tank who didn't have any haste items.

Zurgat · 05/31/07, 2:59 AM

Got procwatch added to favourites. i'll do some testing in the weekend,
autoattacks - offhand only,
autoattacks - main + offhand
autoattacks - main + offhand + SS Spam
autoattacks - main + offhand + SS Spam + Slice and dice.

used the gem at gruul last night, seemed to proc decently enough.

• Aldriana · 05/31/07, 4:02 AM

So, after seeing my initial test and Sarj's, it seemed to me that the trinket really was acting like it had an internal cooldown. So I decided to do a bit of testing on it; specifically, I parsed an entire night's raiding for TSD procs (specifically, TK trash up to void reaver, 3 pulls of Void Reaver including a kill, and then 6 bosses in Karazhan). In that time, 72 procs occurred, with the following characteristics:

1) No refresh procs at all. The longest continuous period of TSD haste was 6.203 seconds, which is within the range for a single proc.
2) No procs less than 40 sec apart. That is, the shortest period of time from the start of one TSD proc and the start of the next was 40.672 seconds; on 7 different occasions there were procs less than 45 sec apart, but never less than 40.

Statistically speaking, there's just no way that that's anything other than a hidden cooldown of 40 seconds - I can't come up with any other reasonable explanation. I can't explain Noobiisa's results - all I can say is that either he didn't see what he though he saw, or they hotfixed in a hidden cooldown at some point - but I just don't see any other explanation for the results I've seen, other than that there's a 40 sec hidden cooldown.

So, assuming for the moment that it is, in fact, a 40 sec hidden cooldown, what's the proc rate? Well, grabbing the data I took last night and throwing out attacks that occurred within 40 seconds after a proc, I find that with a single 2.6 speed weapon, TSD procced 20 times out of 605 eligible hits, and with a single 1.3 speed weapon it procced 15 times out of 575 eligible hits. Grinding some statistics on this (which I won't bore you with) shows us that the answer is one of two things:

1) It's a 1 PPM trinket with a 40 sec cooldown, or
2) It has a fixed proc rate of ~3% and a 40 sec cooldown.

Unfortunately, both are within statistical bounds, so, alas, no conclusion can be reached. It would require a larger data set to know for sure. But I'm reasonably convinced at this point that it does have a hidden cooldown.

LiteSabre · 05/31/07, 5:25 AM

Originally Posted by Aldriana

So, after seeing my initial test and Sarj's, it seemed to me that the trinket really was acting like it had an internal cooldown. So I decided to do a bit of testing on it; specifically, I parsed an entire night's raiding for TSD procs (specifically, TK trash up to void reaver, 3 pulls of Void Reaver including a kill, and then 6 bosses in Karazhan). In that time, 72 procs occurred, with the following characteristics:

1) No refresh procs at all. The longest continuous period of TSD haste was 6.203 seconds, which is within the range for a single proc.
2) No procs less than 40 sec apart. That is, the shortest period of time from the start of one TSD proc and the start of the next was 40.672 seconds; on 7 different occasions there were procs less than 45 sec apart, but never less than 40.

Statistically speaking, there's just no way that that's anything other than a hidden cooldown of 40 seconds - I can't come up with any other reasonable explanation. I can't explain Noobiisa's results - all I can say is that either he didn't see what he though he saw, or they hotfixed in a hidden cooldown at some point - but I just don't see any other explanation for the results I've seen, other than that there's a 40 sec hidden cooldown.

So, assuming for the moment that it is, in fact, a 40 sec hidden cooldown, what's the proc rate? Well, grabbing the data I took last night and throwing out attacks that occurred within 40 seconds after a proc, I find that with a single 2.6 speed weapon, TSD procced 20 times out of 605 eligible hits, and with a single 1.3 speed weapon it procced 15 times out of 575 eligible hits. Grinding some statistics on this (which I won't bore you with) shows us that the answer is one of two things:

1) It's a 1 PPM trinket with a 40 sec cooldown, or
2) It has a fixed proc rate of ~3% and a 40 sec cooldown.

Unfortunately, both are within statistical bounds, so, alas, no conclusion can be reached. It would require a larger data set to know for sure. But I'm reasonably convinced at this point that it does have a hidden cooldown.

So saying that it is either 1PPM/3% with a 40 second cooldown, does that still leave the gem at better than swift skyfire/relentless earthstorm? The haste-freak in me says yes, but I'm getting kind of skeptical with all the evidence coming in that it isn't quite as good as everyone thought it was.

Noobiisa · 05/31/07, 10:02 AM

Did some new tests and with a way faster weapon (Used 2.9 for previous data and today 1.9 green I grabbed from AH), so my effective auto speed was 1.2 for this.

Shots fired: 2025
Procs: 34
Time: 44m50s
Hits/proc: 59.5
PPM: 0.7.

Seeing what Aldriana stated about hidden cooldown I parsed my combatlog aswell and the quickest upon succession procs I got was:

5/31 15:34:16.906 You gain Skyfire Swiftness.
5/31 15:34:52.453 You gain Skyfire Swiftness.

So it concludes with his evidence of hidden cooldown and I will tend to say it is atleast 30secs, we have proven it is no more than 40secs, but can even be as high as 35secs.
With this results and a way faster weapon my hits/proc increased alot and is in relation with the wpn speed of my bow what I used before and what I used now so I can tell you 100% that the proc chance is based on PPM.

So as conclusion it has 1PPM per minute with a hidden cooldown of 35-40secs, so with the hidden cooldown now fairly well confirmed from more than 1 source. This makes it up for a new discussion if it is really better than Relentless cause of the hidden cooldown. But for now anyone thinking of this gem, it has 1PPM with a hidden cooldown of no less than 35secs, closer to 40secs.

So with this new hidden cooldown and confirmed that it is PPM the procrate, new theorycrafting: With my weapon (2.9 speed) and 1PPM it has 4.83% chance of proccing, which equate roughly of proccing every 21 shots. With specials between each auto and having no other haste effects on me whatsover and shooting 57 shots per minute, means after 19.5secs it will proc. Now with the hidden cooldown not pinpointed exactly, lets assume highest cooldown of 40secs means it will be able to proc after 40secs from proc, so every 60secs it will proc. 6secs every 60seconds it will be up 10% of time of which I will have 22% haste on me. Which is 2.2% overall dps inc for me. Assume I do 900 by myself, that is dps inc of 19.8 (alot less than before hidden cooldown

).

To compare with Relentless Earthstorm, 3% crit dmg inc, I crit 33% of the time and with 33% crit it means exactly 50% of total dmg done will be crit dmg (dont get me wrong in the sense that I say 50% is due to crit chance, I say 50% of dmg done with 33% crit will be suspect to the 3% dmg inc). 900dps as total, 450DPS of that is then due from crits, hence 450*0.03 = 13.5 DPS inc and then with the additional 12agi which adds roughly 5dps aswell I might tend to say Relentless is better once you start to breach the 33% crit mark.

P.S. Sorry about initial post where I said I had a proc in 10secs from each other, that data was from a while back and seems they have hotfix something cause it was such a good gem and think Blizz realised that it overshadowed any other meta there is for physical dmg dps. Just wish they will tell us this...

Abaxial · 05/31/07, 11:28 AM

This was posted by someone on the WoW rogue forums, yes he didn't post his log but I don't see why he'd lie about his findings.

Tonight I managed to get two Skyfire Swiftness procs in a span of 27 seconds. That's quite a bit lower than my previous most favorable fight which showed an average cooldown of 30 seconds between proc effects.

My running totals for tonight show 6338 hits and 176 procs in 97 minutes of combat (~1.8 PPM).

-Sapp Tichondrius

Also, with all of these items with haste buffing your chance to proc Combat Potency, is anyone finding it a bit awkward to keep a consistent cycle going?

kennyxedge · 05/31/07, 11:42 AM

I've had one for a few days now, and it seems to proc pretty often. However, I have had no luck whatsoever getting ProcWatch to work with SimpleCombatLog. If you've got any tips/tricks for making ProcWatch work, I'd be more than willing to post some screen shots as soon as I get them.

Shaker · 05/31/07, 12:01 PM

6secs every 60seconds it will be up 10% of time of which I will have 22% haste on me. Which is 2.2% overall dps inc for me. Assume I do 900 by myself, that is dps inc of 19.8 (alot less than before hidden cooldown ).

You're probably factoring in instant attacks here (though I don't know much about post-TBC hunter shot rotations, so I could be totally off), but haste won't affect instant attacks, so you'll only get 2.2% of haste-affected DPS, which is an even lower number.

That being said, haste rocks for rogues, especially with combat potency, and unless there's some VERY surprising numbers, I don't think TSD is going to be dethroned.