Tidise

Hopefully he didn't overlook something like Mongoose enchant on his weapons. However, I do think the Mongoose proc is called "Lightning Speed".

 

Svidrigailov

The procrate still ought be slightly higher than 1.3, because the combat log (and thus, presumably ProcWatch) doesn't pick up procs that refresh existing procs.

 

• Aldriana

Right, which takes us back to the 1.7 number I computed originally. Point being: first procwatch read 1.3, which means actually 1.7 (see my early computations). Second procwatch says 2.6, which, once divided by 2, puts us back at 1.3, and hence presumably an actual proc rate of 1.7 as well.

 

• Aldriana

So, I picked up the trinket tonight, and am ready to run some tests on it - can someone remind me how to print your combat log to a file? Thanks.

Ald

 

Northerner

/combatlog

 

Cyrithor

I'm not sure about this, but if it is the case then it does explain the higher numbers I saw. I'll run it again and be sure.

 

Trohck

Anyone have any updated information as to whether this is PPM-based or not?

 

Zuckas

Had my dragonspine for a couple weeks now with proc watch running. As can be seen from the "most recent combat" even though I am tracking only "haste" it's only registering the gain effect and not the lose. Also through my own testing it's not registering the 2 pc assassination bonus I have as a proc for "haste".



Armory profile link.

 

Vardisk

I've been stacking the haste lately too. Have had problems with accurate parsings though, due to the fact I have 3 effects called Haste (Assassination 2p, Dragonmaw, and Dragonspine Trophy). Is there a mod that differentiates between them?

Definetly fun getting everything stacked up at the same time though. Broken +260% total haste a few times now.

 

• Aldriana

I did some further parsing on this trinket today, and while the results I got weren't as conclusive as I might have liked, we can say a few things about it.

Test methodology involved blasted lands mobs and straight autoattacking with no proc effects on other than Dragonspine Trophy, and autoattacking until I got low on HP, whereupon I'd vanish, bandage up, and pull him again.

I kept only a single weapon equipped so as to not have to contend with MH vs OH issues.

I did this with two different weapons: Warp Splinter's Thorn (1.3) and Demonblood Eviscerator (2.6)

I assumed that any proc over 11 seconds in length was actually two overlapping procs, and, likewise, any proc over 21 seconds in length was actually 3 overlapping procs.

With Warp Splinter's Thorn, I attacked 2220 times and got 40 procs, for a proc rate of 1.8%. Applying elementary statistics indicates that the true proc rate is thus most likely between 1.25% and 2.35%

With Demonblood Eviscerator, I had 1215 attacks and 39 procs, for a proc rate of 3.2%. Again, statistics indicate that the true proc rate is between 2.2% and 4.2%

Converting this into PPM based on weapon speed, we find that Warp Splinter's Thorn procced once every 72 seconds indicating a true average time between procs of somewhere in the range of 55 to 104 seconds; meanwhile Demonblood Eviscerator procced once every 81 seconds, indicating a true average time most likely between 62 and 117 seconds.

Looking at this, a fixed % proc rate seems unlikely. The data doesn't rule out a fixed proc rate of 2.3% or so, although it does seem a bit unlikely.

Similarly, 1 PPM seems unlikely - it falls outside the range of the data for Demonblood Evisecerator, and is right on the edge of the range of Warp Splinter's Thorn. It is not impossible, but it does seem unlikely.

So far, the best model for the behavior seems to be a proc rate of around .75-.8 PPM (i.e. one proc every 75 to 80 seconds). However, more data would be needed to confirm this. I'll take additional data as I have time to see if I can pin this down further.

 

exog

Dont know if its relevant, but did you attack the mob for equal amounts of time with both weapons? Getting 39 and 40 proccs with those weapons seems awfully convenient assuming theres a weaponspeed contribution to the procc calculation.

 

• Aldriana

Um, more or less the same amount of time, yes. I did the same number of attack/vanish/eat cycles with each weapon to get a comperable amount of data for each.

 

loopyj22

I did some testing today and saved the combat logs if you'd like to parse the info. I ran 2 5 min long tests. The first test was a 5s/5r cycle using full talents available to me except bloodfury (as it shouldnt have any effect on proc rates). The second test I ran was the same except for using 5s/5e cycles. The change in cycles was part of another test and shouldnt effect the outcome of procrate testing I would believe. Anyways....The last 5 minute session here is the procwatch info:

hits: 478
procs:17
time: 5:00
hits/proc: 28.1
procs/min: 3.3


I did not save the in from the first session but i remmeber the ppm was at 3.1

So the data was similar. Now I am new to all this and just wondering what exactly that means. I am full combat rogue with and was using AR to start with then spamming the cycles previous mentioned. I alos used blade flurry when it was up throughout the 5 mins. This data was teaken with full raid gear on and no other haste procs available through Assasition set, Blackout truncheon, etc.

Am I to believe that this is telling me thatbased on 2 weapons the proc rate is close to 50%? Reading back in the original post that would fit with the expected percentage of 25% with the buffed haste proc rating couple patches ago.

Thanks for all insight. Just trying to understand this trinket a little better and how the PPM rating works.

If you'd like a copy of the combat log PM me here with contact info and I"ll send it to you.


EDIT: MH spd = 2.6 OH spd = 1.6

 

alienangel

Sigh. Can anyone tell me if this trinket is even remotely useful to an MS warrior on a top ranked arena team? It never occurred to me that someone using a 3.8 speed weapon would have any interest in this at all until I saw who won the bid. Yes this is a post about class mechanics more than the trinkets mechanics, feel free to ignore me if it's too off-topic.

 

koaschten

I dare saying, "This is the ultimate DW trinket" shusssh MS Warrior

 

Adrianna

Quick math says around a 10.8% chance to proc per swing, assuming 1.7ppm. Speed increase is on par with top ranked flurry, 10 seconds gives you 3-4 hasted swings, depending on what procs it and when in the swing timer.

It's essentially like having flurry with an 11% crit rate. Seems like an "oooh, shiny!" bid, or a future scoop for a DW fury respec.

 

Infectedbull

This is my Procwatch from shadow labs. Granted it isn't comparable to the other one posted but some might find it interesting.
The last fight was from the last boss.

 

IceBox

I also finally got it yesterday and immediatly went to Blasted Lands to test it.

I came up with about 3.8 raw ppm but it counted every proc twice and I don't know why.

So with adding the refresh procs and dividing through 2 because of the double counted events I'd say 2.5 to 3.0 PPM.

I just like that thingy :>
It just rocks, especially when you got and 60 second peroid where it's up all the time. Hello 1.5spd Black Planar Edge!

 

Nevervate

Got one of these last night and spent a good amount of time gathering numbers.

Screenshot:


http://img508.imageshack.us/img508/5...nspine1hv1.png

Hits: 1206
Procs: 46
Time: 19:36
Hits/Proc: 26.2
Procs/Min 2.3

46procs/1206hits = 3.9% proc rate

 

Surion

I've had it for exactly a week now myself and in honesty, no. It might be nice for a MS warrior to get that haste effect every now and again however a DW class will get much more benefit out of it. If he was in my raid and bid I would have called bullshit.

 

Eldia

Anyone know if this trinket is changed in any ways in 2.1? e.g. hidden cd added so that it won't proc back to back, etc?

 

Vaxum

I've tested this on PTR and it definitely still procs back-to-back.

 

Grayson Carlyle

Why would this be the case? Uptimes for procs using a 2H are higher than DW because of the ratio of instant attacks to normal swings is higher.

 

Cel

It doesn't mean you hit more than dw though... how can it have better up time, unless it's a PPM rather than a %? (or is it a PPM and I missed that?)

"There is much pleasure to be gained from useless knowledge." - Bertrand Russell

Alpha is recruiting... go go.

 

The Iron Colonel

It's about 1.7 ppm, as I understand things.