 |
| Welcome to Elitist Jerks |
We're testing some new features on the site regarding OpenID registration and coordination with gamerDNA. If you experience any issues with registering an account, please take the time to fill out a report and send it to this e-mail address. We would appreciate any assistance you could provide in making sure everything is functioning as intended. Thanks!
If this is your first visit, please be sure to check out the FAQ and the forum rules. Users must register to post and new registrations are subject to a one day "mute" period to get acquainted with the community.
|
04/24/08, 7:09 PM
|
#1
|
|
Do Not Stand In the Wizards
|
Do dice rolls count decimal places, or round?
A simple thread that seeks to answer a simple question.
We know that the character sheet displays up to two decimal places for your percentage chances to hit, crit, block, parry etc. Has anyone verified that the combat system dice rolls are actually accurate to 2 decimal places, or are they rolling integers 1-100? Please don't respond with answers like "we can assume that" or "based on the character sheet." It's baseless to assume that Blizzard's coders have been 100% accurate on every detail of combat mechanics, especially on something as comparatively minor as this.
If anyone is up to the task of posting logs or evidence on this topic, be sure to track auto-attacks and special attacks separately, as they clearly use different dice-roll mechanics.
|
www.magegraf.com
Raiding is full of challenge. Sometimes there is fire. You have to not be in the fire.
"We agree with Communism." - Greg Street 2009
|
|
|
|
04/24/08, 7:25 PM
|
#2
|
|
Super Macho Man
Night Elf Rogue
Proudmoore
|
I think we can say pretty safely that they track multiple decimal places. How many decimal places that might be is of course, hard to answer; but the argument I'd make is this:
If it were rounding off to the nearest percent, it would be impossible to have a miss rate between 0 and 1%; that is, if you could miss at all, you'd miss, on average, one in every 100 attacks. However, rogues have taken data sets with miss rates demonstrably below this - the first one I found searching around is this post wherein a rogue attacks 10286 times and only misses 5, which is statistically significantly different from 1% miss rate and clearly not 0% miss rate. So it's certainly not rounding to the whole percent.
|
|
|
|
|
|
04/24/08, 7:28 PM
|
#3
|
|
Von Kaiser
|
A tank can still be crit with 489 defense and attacks can still miss if you are one point below the hit rating cap. I would say this is enough evidence that combat rolls are something other than integers from 1-100.
|
|
|
|
|
|
04/24/08, 9:58 PM
|
#4
|
|
King Hippo
|
You can also alter your UI display. If I remember correctly, the UI stores more than 2 digits, I forget exactly how many looked real, but you can display 14 or 16, but after a certain point you see noise. I believe the roll is a straight floating point.
The GetCritChance() UI function returns a floating point and can be formatted to more than 2 digits.
Last edited by Dontmindme : 04/24/08 at 10:01 PM.
Reason: More info
|
|
|
|
|
|
04/25/08, 3:25 AM
|
#5
|
|
Don Flamenco
Blood Elf Warlock
Turalyon
|
I'm curious as to why spell haste displays as xx%, as opposed to crit, which displayers as xx.xx%. Does spell haste also have decimal places, or does it round?
|
|
|
|
|
|
04/25/08, 10:37 AM
|
#6
|
|
King Hippo
|
Originally Posted by Krazen
I'm curious as to why spell haste displays as xx%, as opposed to crit, which displayers as xx.xx%. Does spell haste also have decimal places, or does it round?
|
Oddly enough, a guildie was just asking me that question. They had decided to experiment with spell haste and had equipped a single +10 spell haste gem and was unhappy that the display still said 0%. Given it's 15.76% rating to 1% and 10 rating should normally round up, it seems that spell haste might be floored like Defense, Expertise or the old Weapon Skill.
|
|
|
|
|
|
04/25/08, 10:53 AM
|
#7
|
|
Von Kaiser
|
I had 25 spell haste listed as 1% haste on the character sheet. I equipped a 5 spell haste gem, and was still listed at 1% haste. But, my shadwbolt cast time was reduced by 0.01s.
|
|
|
|
|
|
04/25/08, 10:58 AM
|
#8
|
|
Glass Joe
|
Programmatically speaking, all rolls are effectively a real between 0 and 1 scaled to the desired dimensions, are they not? For example, besides the combat rolls, the straight /roll (and /roll X-Y variant) actually returns a real, though it's displayed as an integer. Everyone remembers the "My 100 beats your 100" cases that were so prevalent back in 40 man days, right?
|
|
|
|
|
|
04/25/08, 11:10 AM
|
#9
|
|
Piston Honda
Night Elf Druid
Die Todeskrallen (EU)
|
Originally Posted by zliplus
Programmatically speaking, all rolls are effectively a real between 0 and 1 scaled to the desired dimensions, are they not?
|
It's usually implemented like that, yes.
|
|
|
|
|
|
04/25/08, 12:14 PM
|
#10
|
|
Glass Joe
|
Originally Posted by charriu
It's usually implemented like that, yes.
|
No, it's usually implemented returning 0-RAND_MAX (whose value can change based on the system and the algorithm used). A float isn't typical for a return type for a system level call. If you want a floating value it's up to the application to convert the returned value into what is appropriate for it's specific purpose.
|
|
|
|
|
|
04/25/08, 4:52 PM
|
#11
|
|
Divine Protector
Blood Elf Paladin
Mal'Ganis
|
Originally Posted by Dontmindme
it seems that spell haste might be floored like Defense, Expertise or the old Weapon Skill.
|
The haste tooltip is incorrect then, just look at your spells' tooltips and as you add 10 haste gems your spells cast time will get faster.
|
DK - Ashbane Failure is the condiment that gives success its flavor.
|
|
|
|
|