Interesting Discovery about Critical Strike and %s

Many map makers say that the WC3 engine rounds the % to the nearest 5% (so the chance to perform a Critical Strike is actually 20% and not 15%), but this is not true.
I made several tests and map in which I had a unit dealing 10-10 chaos damage and having a 15% chance to perform a 2x Critical Strike; then I had another unit with 0 armor, many hit points a fast regeneration rate. Every time the seconds unit was damaged by a critical strike (the damage was exactly 20, easy to check) a counter increased by 1. I made the first unit attacking the second for exactly 1000 times (the first one had 0 attack cooldown and 0 attack animation).

The result was that my unit performed 148 critical strikes over 1000 attacks and this is very near to the 15%.
I made another test but this time the attacks were exactly 100, the results were 85 normal attacks and exactly 15 critical strikes.
So, in the end, the next time someone will tell you "WC3 engine rounds the %s", simply answer NO
Cheers

P.S: I used Blizzard's Critical Strike (the same of a normal blademaster), not a custom made one.

....

Your post is a comedy of errors. WC3 chance-to-procs do round to the nearest 5%. Thus 15% stays 15% (I assume you were meaning round to nearest 10%, or messed up on something else? god only knows). So you just went on to prove that 15% rounds to 15% and that the chance worked just fine. Congrats.

So the next time someone tells you that WC3 rounds the % to the nearest 5% (or is it down? never bothered to try), simply answer YES.

(This should probably be deleted to prevent it confusing the hell out of people)
(edit: and by this I mean the thread, because the thread starter should really get a clue about this stuff before starting a whole thread about it working exactly as it should when it was never doubted)

I started this thread just because I have heard a lot of people saying that the Wc3 engine rounds the %s to the multiples of 10 (not 5, sorry) and so a Critical Strike with 15% chance would fire 20% of the times.

Nobody says that.

Never heard that before either.

First time I've heard that.

Same here.

-Av3n

Oh, well, lol, I'm sorry
I don't remember where I have heard that, anyway I apologize for the useless post...

I've never even heard about rounding to the nearest 5%, proof please?

I'm pretty sure Wc3 rounds to the nearest 5% for abilities like crit strike.

The easiest example is a 1% instant kill ability: in my experience (haven't tested for ages, but i doubt it's been patched) it will never fire.

That would be crappy if it was true. Really crappy. Someone please test this.

Okes, i'll do it now...

EDIT:

Aim: to see whether it is true that Critical Strike (and presumably similar abilities) round to the nearest 5%
Environment: identical units, one with 2% chance, the other with 3%.
Hypothesis: if the rounding is true, 2% will never fire, but 3% will.

What happened: 2% never fired, 3% did.

Conclusion. 2% was rounded down to zero. Test was not set up to accurately find the firing rate for 3%, it did however seem to be ~5%.


Interesting though: Drunken Brawler: the melee stats are not multiples of five...

All this time Dota's Radiance (8% Evasion) has been a 10% then.

This is Confirmed.

After running various tests, i can confirm that a 2% Critical Strike will Never fire and that a 3% Critical Strike will fire more than 3% of the time, it will fire 5% of the them to be precise.

So i ran some more tests.

Critical Strike, Bash and Drunken Brawler all round to the nearest 5% for the Critical Strike componant. Evasion does Not round to the nearest 5%.

Interesting and throughly crappy. I'll run my own test tomorrow to get the exact proc rat of crit strike at 2%, 3%, 11% and 17%. Purely for my own peace of mind.