you should also count the number of non-MB casts, so instead of comparing values (which will be skewed because teleport patterns won't be the same), you can compare ratios, which will give you a much better indicator of the effects of pumping MB.

i thought about that too, but then you're introducing another variable. i tried to do pretty similar patterns (it didn't seem to make too much difference about where i was in terms of how often it uses mindblast, as long as i wasn't very close), and anyway then i'd have to keep multiple numbers in my head =/ which makes it a lot easier to lose track.

repitition...yeh, i mean that helps negate the chance thing a bit...but tbh, the tests were quite consistant, it wasn't as if it mindblasted 80 times in the first 5 minutes and 26 in the second, it was a little over 50 at the half-way mark. so i don't think it really matters too much. once i decide how many points i'll probably be using on my build, i might do more intensive tests, but for now 10 minutes seems to work pretty well.

i think the main thing i'm seeing so far is:
1) hard points make a big difference on frequency
2) it's not a linear relationship

oh, and thanks for the tip about hard points in master, i'll test that next.

This post was edited by darcanegel on Apr 21 2008 12:56pm

counting non-MB casts doesn't introduce another variable. turns your values into "real" numbers rather than absolutes, which make them much better for testing frequency. all using ratios does is hedge against two different teleport patterns that trigger a different total number of shadow "actions".

my teleport patterns were pretty random =/

the issue with testing against another skill is that there's a higher chance that i just "got lucky" and cast a certain thing more often...i.e., the number would be off if it swed more, or if it mbed more (or less). plus maybe it uses mindblast less often, but also uses sw less often...that makes it worse on both counts, but by that test it would seem to be just as good.

...plus i just don't want to keep those numbers in my head tbh.

anyway if you want to test it that way be my guest, i'm convinced my test is fine, the results have been pretty consistant.

oh look he's starting to get a clue

I agree that using a ratio would make things much more accurate. Due to the fact that it seems to be dependent on distance, position, etc (which would be very difficult to keep constant), using a ratio helps to cancel that out.

I'm also glad that someone else tested this (other than me). Saves me the trouble

hmm? i never argued sm mb isn't affected by hard points in mb...i just used warrior instead to avoid pumping it =/

any my test might explain why my 16 mb sin was underwhelming, it looks like having maxed mindblast is a huge improvement over "high" mindblast.

by counting that second element, you'd be able to see how many total "actions" there were, so you would be able to see whether it was a lower number of actions or just a lower frequency of mb.

you can also take a bit of the randomness out of it by teleporting into a corner with a stationary target some distance away, so your distance and relative position don't change. but then you'll have to do a few different tests with varying distances though.

This post was edited by Dumpling on Apr 21 2008 01:23pm

my understanding of shadow master is it uses the AI bonus as a improvement in picking a given skill, then the skill is used several times (she will usually keep creating the same trap for example instead of picking random trap skill each time)

Welcome to the world of science haha