Query (Resists)

[gaeelornie]

If everyone on server or zone uses the same random number gen, such as in BB, my Druid gets same .341234 as my necro gets .341234, whats the time between updates on the generator? is it in secounds, or is it used as in when X amount of spells w/ chance of resists are casted then the random number generater generates again, ?

 

[Garluk]

A new number is pulled from the random number generator (RNG) every time something happens. Every action is independent from all the others as far as the RNG is concerned. The actual number crunching that goes on is simply mind boggling. It is not easy to grasp just how powerful a modern computer truly is.

A quick example: you swing, RNG spits out a number, multiple calculations are made based on your attack/level/mobs's defensive skill (level vs you, AC, etc). If you pass all those checks, it's a hit. RNG again to determine the amount of the hit. RNG again to see if you pass your double attack check. If you pass that, then RNG again to determine hit/miss. If hit, then another RNG to determine amount. If your class is capable of triple attack then another cycle. If you are attacking from the front arc and the mob is capable of riposting there is another RNG check to see if any of the hits that would have landed get riposted. This is highly simplified, but that's just for one weapon for one person for one attack cycle.

Now imagine that you have 15 melees, many of them dual wielding, all hasted out the wazoo attacking at once. It would literally take days to calculate by hand what happens during a single 3 minute fight. And this is just the dmg calculations. The cpu is capable of handling millions of instructions per second. Spitting out a new value from the RNG is just a tiny drop in the bucket.

The proof that a new RNG value is used in each action is that you don't hit for the same amount every attack cycle and that if you do happen to land both swings on a successful double attack that you hit for different amounts (usually) on each.

Sorry for babbling on but most people have no idea just how incredibly powerful their computers are.

 

[gaeelornie]

So!

So how is it that sometimes I dont land on a mob 6-7 times in a row, if it was calculated randomly the odds of that happening are quite small especially if my normal resist % are quite low around 30%

 

[calaran]

In brief, the RNG (random number generator) is flawed, as are all computer-based RNG's. Nothing is random.

If you want a more detailed analysis of the probabilities and whatnot, see the 4th and 5th pages of this thread:

http://www.shardsofdalaya.com/forums/viewtopic.php?t=7350&postdays=0&postorder=asc&start=60


But the simple answer is, all RNG's are flawed and as such there are glitches and sometimes patterns. If your CHA is not just insanely high, expect lots of resists. If you're fighting mobs with their native damage type (casting fire against a fire elemetnal, or poison against a snake) expect to be resisted a lot. If you're fighting in some high end areas, they just have high innate resists. It's going to be tough, going to be challenging, but for the most part, it is working as it should. However, to contrast that, a problem was brought up recently and when they looked at it, they realized that one zone actually DID have higher resists than it should have, and they've tweaked it back down a good bit.

So, post specific zone, mob, damage type you're trying to do, what spell, how high your charisma is, etc... The GMs can then look at it and see if they feel it's out of line. Most of the time, it is not out of line with what they want to see, and as such, things will stay the same; however, you may be the lucky one to discover an area that is out of whack, and it'll be re-scaled after they look at it.

 

[gaeelornie]

If everyone on server or zone uses the same random number gen, such as in BB, my Druid gets same .341234 as my necro gets .341234, whats the time between updates on the generator? is it in secounds, or is it used as in when X amount of spells w/ chance of resists are casted then the random number generater generates again, ?

... niether of you answered my questions to quote a master /snicker "Dont shit up my post"

 

[calaran]

If everyone on server or zone uses the same random number gen, such as in BB, my Druid gets same .341234 as my necro gets .341234, whats the time between updates on the generator? is it in secounds, or is it used as in when X amount of spells w/ chance of resists are casted then the random number generater generates again, ?

... niether of you answered my questions to quote a master /snicker "Dont shit up my post"

I'll tell you this much; being a cock to those who are trying to help you is one way to get quickly ignored by the whole community.

You already had that question answered so I was addressing the OTHER question you asked.

The RNG is hit each time something is done. Period. It's not like a number is grabbed and then held onto for the next 10 seconds or some such crap. You were already told that. If you can't read, that's not my fault.

To repeat in case you missed it. There IS no time lapse between the updates. None. It's not based upon the # of spells used. Not at all. So read this slowly....

THE RANDOM NUMBER GENERATOR IS ACCESSED EACH TIME AN ACTION IS TAKEN THAT NEEDS A RANDOM NUMBER.

That was your original question, and was answered by the first person to respond to you. You then asked another question which I answered and then you decided to be an ass. ::EDITED:: I took that out because it would prolly be seen as a threat ::EDITED::

 

[Minerios]

Re: So!

So how is it that sometimes I dont land on a mob 6-7 times in a row, if it was calculated randomly the odds of that happening are quite small especially if my normal resist % are quite low around 30%

I think you don't understand statistics very well...

Stupid example: pick a coin. Flip it. It has a 50-50 chance of head or tail. That doesn't mean that after 10 throws, you'll have 5 heads and 5 tails. You may have that; you might have a 7-3 pattern; you might even go 10-0. The 50-50 chance of head or tail means that the more throws you make, the more the total of heads or tails will tend to even out. That doesn't mean you have more chances of getting a "tail" after getting a "head". It's a tricky nuance, but it's fundamental to understand statistics. Every time you toss a coin, you have just as much chance of having it land on one side as you have of having it land on the other, regardless of what the past throws were.

The human mind seeks patterns because patterns make sense and we like to understand what goes on. Another example: you shuffle a deck of cards. You would certainly be terribly impressed if the cards ended up being sorted by color. And you'd be absolutely bewildered to find it sorted both by color and numbers! But what our human mind often fails to recognize is that this pattern has just as many chances to come up as any other pattern. The difference is that these other patterns are all lumped together as "random" in our minds, even though they are just as random as the one with the cards sorted by color and number. It is our perception that makes it appear exceptionnal.

...

Did I make sense?

 

[Birlic]

Re: So!

I think you don't understand statistics very well...

Actually, Oualawouzou, it is you that doesn't understand statistics very well
First of all, this is all probablity theory knowledge and not statistics.
Gaeelornie is right, the probabily of that chain-resist pattern happening is very low because the distribution of "resists along several attempts" is binomial and not Bernoulli. As for your deck of cards example, don't give the human intuition less credit than it deserves! Try to compute the probability of getting all cards of the same color with the very basic definition of probabilities:
p = num_favorable_outcomes / num_possible_outcomes
and you'll see that the probability of that happening is .25^13 ~= 1.5e-8
i.e. one in 67108864 trials.
No wonder our human mind is impressed when dealt with all cards of the same color

 

[Minerios]

Re: So!

Thanks for the precisions. I shouldn't talk about stuff I could barely grasp in college (and has since forgotten it large parts).

Consider my previous post null and void. ;(

 

[gaeelornie]

When you flip a coin yes, you have a 50/50 chance of it landing on one side (almost it depends on the weight of one side of being equal to other ie one side weighs .01 gram more its more likely to be on downside etc) so yes if you have a 6 side die (Dice in singular dont feel like checkin if that is spelled right) and all sides being equal you have a 1in6 chance of it being a 1 and you have a 1 in 6 chance of it being a one the next time... but when you start going hey I'll roll it 6 times and it should be a 1, 2, 3, 4, 5, 6, in any diff order, but it wont be it might be a 1, 1, 1, 1, 2,3. or anything else but the odds of six 1's in a row are quite small much less then 30%, therefore either its not reset every cast or the RNG is kinda screwy or its much more complicated then you are even trying to think it is, does the resist chance go from a 1in6 to a 3in6 on the secound cast? because statistically it should not be a 1,1,1,1,1,1. many times at all, like almost... never as far as I know will happen once in a blue moon, so what i was really askin is someone who KNOWS the EQ engine and how these things work to answer because, I read the other post, the whole thing and my question was not answered, its like on live a JC with the skill maxed has like a 1% chance to fail (or less) on something that triv's at 21, but sometimes you still fail on a full stack in a row which has a very very very low probability of happening though... it happens and you hear about it happening alot on live, and it happens here with resists etc that I was just wondering about how the engine itself worked, not how you thought it works and a quick review of elemetery statistics

 

[Garluk]

If I recall correctly from previous posts, people with CHA that is normal for their level are getting in the neighborhood of 20ish% resists. That's 1 in 5. It's not that unlikely that you could be resisted 4-5x in a row at that rate. Aggravating, yes. Get you killed, possibly. But not a statistical stretch for it to happen.

 

[gaeelornie]

The statistcal odds are much smaller then what it works out to be in reality I'm gonna try to find a good spot and just spam a level spell on a low level greenie with no cha gear on and see what happens, cause 5-6 resists in a row just really shouldnt be all that common