Greetz,

I built a few tools to calculate the value of proc effects. Based on data such as cooldown, proc chance and your casting behario, the calculators try to give it a statistically complete answer.
I hope this will spare you some spreadsheet work, when evaluating trinkets

The tool is available here:
http://zusterke.orderoftheathanor.eu/procs.php

The old version of this tool can still be accessed here:
http://zusterke.orderoftheathanor.eu/to ... t=procBeta


Suggestions and questions are welcome!
Zus

Hmm, maybe I better give a small demo. Imagine Bangle of Endless Blessings.

It has:
- a 10% proc chance
- with a 45 second internal cooldown.

We then want a value estimation. I could, of course, calculate the regen it would give me on a proc and throw that in the calculator. Rather... I set the proc value to 10 for the 10% regen while casting (lvl 80) it gives me.
The duration is 15 seconds and I estimate and average cast speed of 2.2

In the new calculator, I use 'method 2' and fill in these values in the respective fields. The results give me:
(In the old calculator, http://dieter.srs.be/zusterke/tool.php? ... =15&cs=2.2 )

It guesses the overall value to be:
3.18 - 2.52 - 1.82

Which is the % regen while casting. Now.. I'm gonna round it up to an average of 3 (since i want to avoid the floating point for now), and stuff that 3% in my regen calculator, combined with my usual stats:
http://dieter.srs.be/zusterke/tool.php? ... b=10&mp=mp

We can see how this trinket pushes back the value of MP5

OOOOO so cool!

You keep this up and soon nobody will bother checking EJ anymore.

Stickied

It seems the calculator undervalues proc values. What is probably the reason is the following. The proc pattern is like this:



If the fight takes 70 seconds, you would easily obtain 2 procs. However, the calculator takes the full cooldown into account and thus evaluates the proc as it the fight took 90 seconds. A 28% difference in this case which is quite a lot. This difference decreases gradually as the fight becomes longer or the fight ends near the end of a cooldown.

I could add an input field for time correction to the calculator, but I would overload an already complex interface with more input fields to obtain a number that isn't exactly more accurate.

Perhaps a better approach would be to add a 20% cooldown time correction to the "best" result, 50% to the "average" and 80% to the "worst" result. The interval of 'best'-'average'-'worst' would cover a larger variance then, but I believe that would be better model of the proc performance.

is this the procodile addon?

Nope

And it's still in development. However.. I have recently figured out a few things about how to represent some additional information so this will be updated within 2 weeks probably, and finally 'ready'.

is it going to be like procodile, or is it going to be niftier? just looking to see if there will be more features/info w/your add on. i like information!

Procodile is an ingame mod that tracks the performance and statistics of procs. What I'm building is a webtool that, given the proc chance and internal cooldown (and value of each proc) is capable of estimating the performance and the reliability or chance to score that performance. It is, in a way, an automated theorycrafting tool.

ahh! so it could be usful if someone were to be working on a super secret disc priest spreadsheet project?!?

thanks

GIEF!!!!

UPDATED!
The OP has been updated but here's the link if you don't want to scroll:
http://zusterke.orderoftheathanor.eu/procs.php

The biggest changes are:
- a separate calculator for each method or type of proc effect. I hope this will be more intuitive than 'guessing which parameters' are required.
- a new site design
- the math background on the calculator's mechanics is provided
- it has been written in Javascript with the following purposes in mind:
* It no longer requires you to submit the page to obtain the answer
* The javascript is embedded in the page, so you can just 'save the page' offline and use the calculators offline!
* The code is open source, allowing everyone to easily maintain their own copy if they want to
* The requirements for the webserver are lowered, making it easier for other people to host a copy if they want to.

Enjoy!

Just a little note: the "Proc Chance" labels should probably be "% Proc Chance". I was putting 0.25 (for 25%) in there and getting dumb numbers until I figured it out.

Done. Thanks for the feedback