FANDOM


So... my last treatise on this subject (http://marvel-war-of-heroes.wikia.com/wiki/User_blog:DrPsyPhi/PROC_rates_--_Skipping_to_the_good_parts) had a few supositions that hadn't been tested (the skilled 10 high and the rel. high skill 1, among others). I took advantage of some recently acquired cards and ran a test of 100 battles against the same opponent (spars, actually) and have some revised numbers. Summary of results then some notes on testing below. I may do some additional testing; it has been suggested that rarity has an effect on proc rates, but I have not yet seen that evidence (opportunities for further research also below). For the sake of the discussion, see also: http://thetalkinggeek.com/procrates/

Revised Proc Rates (italics indicate untested):

Proc Rates for the Modern Player

Rate Center Wing 3 proc
High10 63-70% 100% % %
High 54% 100% 43% 75%
R.high10 60% 100% 45% 82%
R. high 50% 100% 41% 69%
Ave10 50% 100% 41% 69%
Ave 40% 80% 35% 46%
R. low10 40% 80% 35% 46%
R. low 33% 65% 31% 29%
Low10 30% 60% 29% 24%
Low 20% 40% % %

So the big changes from http://marvel-war-of-heroes.wikia.com/wiki/User_blog:DrPsyPhi/PROC_rates_--_Skipping_to_the_good_parts is the High10 and R. high but I've also revised the Ave and others because it's really starting to look like it's about a +10 proc rate bonus for skilling up. Where I have a bunch data (special thanks again Shinji) I have {for this table at least} selected a nice "round" number which is why there are so many multiples of 10. It's possible that actual proc rates are more like 41.5% instead of 40% say BUT >It feels really unparsimonious that the set proc rates (which would be set by the programmers -- remember, PROC=Programmed Random OCcurrence), >it would be difficult even with a large number of tests to reliably detect the difference of a percent or two (however, I think we have decent confidence interval for the numbers otherwise), >I should also confess that my recent test of R. High skill 1 activated 53 times in 100 spars but I but 50 anyway (up from 47% in previous guestimate)

Where we have only one series of tests or where all the tests give a consistent non-round number that fits the other patterns (see the high/high10) I've kept those in. So I would caution that all this is still a work in progress and subject to further revision. At the same time, I would reassure the majority of readers to not expect major revisions (just us wonks who might want to revise up or down a percentage or two).

New tests that informed the above were done with a New Mr. Fantastic (high) at skill 10 (63 activations in 100 tests), a Smartypants Spider-Man (high) skill 1 in center (100/100), and a base UR Arc Reactor Iron Man (Rel. High) at skill 1 (53 activations in 100 spars). {I only test 3 at a time to ensure I'm getting at PROC rates free of interference from the 3-proc limit}


Future research, for those interested in looking to contribute:

  • Put a low (any skill) in center. Put an average uncommon or rare in one wing and a L+ ave in the other wing (AT THE SAME SKILL LEVEL!!!) and run a hundred tests. See if there is a notable (+3%) difference in usage/activation by rarity and get us some data on Low at the same time! :)
  • Do another round of 100 spars/battles with a skill 10 high and a skill 1 high in the wings (same rarity) for confirmation.
  • Put a rel. high skill 1 in center and run it through a hundred tests or more to see if it ever doesn't proc. It may be anywhere from 47-53% and the double center bonus will really help be sure if it's over or under the 50% mark.

Ad blocker interference detected!


Wikia is a free-to-use site that makes money from advertising. We have a modified experience for viewers using ad blockers

Wikia is not accessible if you’ve made further modifications. Remove the custom ad blocker rule(s) and the page will load as expected.