This is why you didn't get a powerscroll with top damage...

[AlyM]

Hello there! I'm going to be explaining why YOU didn't get a powerscroll when you made top damage, or second, third, etc!
I'm sick of hearing "It's broken because the top x didn't get scrolls!" and we've heard the GM's say a hundred million times.. "It's a lottery"... yet I still hear people "That's bullshit! You should get a scroll with that much damage" blah blah...

I dedicate this post to my lovingly jerk boyfriend, Kareem. You want to call me a liar one more time?
K. Moving on.

The powerscroll system works like this:
1. You go to the champion spawn and attack and kill eligible mobs that spawn.
2. Each point of damage causes score. The skill in which you cause damage gives you a % of damage caused. While we do not know the exact percentage given, we know that some templates are better "champ builds" than others as you tend to wrack up more points in that. Usually tamers get the lowest score per damage because... well, we have massive tanks that eat face.
3. These points are added up at the end. A random number is rolled between the points.... whoever got that point gets a powerscroll...

So why doesn't top damage get a powerscroll?
I will show you an example...
This is a scoreboard from Baracoon, probably a week ago.

NOTE: While the champs were still "Broken", the system is the exact same. Basically the old system would "reset" damage after world saves. Nonetheless, the system of handing out powerscrolls worked exactly the same, people just did different amounts of damage.

As you can tell, 5th place was the first person to get a powerscroll. Little miss 1,2,3, and 4 were up there all alone... Why did WeEzzY at 15611 points get a powerscroll when Success had 1.95x the damage?

[[SIDE NOTE: Faction scroll and third scroll were combined in that picture, because there is no "forth" scroll given out and there are factioners present]]

Let's do a little math. First of all, let's add up all the points.
There was a total of 235,425.65 damage points scored at that champion spawn.

[[Side note; there was a third page not screenshoted as the players had <20 points.... the math therefore will have minuscule (.00x%) differences... the point is the same]]

Player 1 (Success) scored 29,750 of that 235,425.65 damage points. That is 12.537%....
That is the approximate % chance of her getting a powerscroll... Doesn't that number give you a little more insight?
I'm not done yet.

You would usually assume that if this champion was repeated, scores the same, Success would be guaranteed a powerscroll after 8 times, right? (Getting this number because 100 divided by 12.537% is 7.976, rounding up to 8)
Wrong. Using statistics, doing it 8 times would only give her a (rounded up) 40% chance of getting a scroll....
Wait WHAT?

Let's redo this champion spawn the long ass way, the way the system does it....

There are 235,425.65 damage...
Each player gets the numbers they scored in that bracket... Here's scoreboard 1-10 in an example.

1) Success ~ 0 - 29,750
2) Artemis Entreri- ~ 29,751 - 58,300
3) Kum Dumpster ~ 58,301 - 83,844
4) Tard the Paladin ~ 83,844 - 107,244
5) Incar Fyreblood ~ 107,245 - 130,169
6) Count Vaxagore ~ 130,170 - 151,012
7) Nabucondosor ~ 151,013 - 168,969
8) Xantara ~ 168,970 - 185,791
9) WeEzzY ~ 185,792 - 201,402
10) Demitra ~ 201,402 - 213,090

[[Note: I rounded numbers up to make things more readable... Tiny tiny percent differences... .00x]]

So, for the sake of understanding, we assume those players OWN those numbers. Success owns the numbers 0 through 29,750... Artemis owns the next section of numbers up to 58,300... So on and so such..

The system now rolls a huge ass set of dice... [[Easier method: random number generator...]]
*shakes huge cup full of dice...*
Each number is weighted equally. Every single number from 0 to 235,425(...point six five but let's just round)...
*cup still shaking*
*number is drawn: 199,732*
*shakes that huge cup another time*
*number is drawn: 118,983 *
*shakes that cup one more time!*
*number is drawn: 147,327*

Wow! Lucky #9 owns that first number! Congratulations, you get a powerscroll!
Whoo! Lucky #5, you get one too!
Oh man #6 you're lucky too!

(Anyone getting this? Anyone at all?!)

Just because Success owned the most numbers, she had less than a 13% chance at getting a powerscroll.
Its not COMPLETELY random, the more numbers you have, the HIGHER CHANCE you get of having a powerscroll... but at no point ever are you GUARANTEED* a powerscroll.

[[*disclaimer: if you're the only factioner there and you have done any amount of damage you shall get a powerscroll. OTHER THAN THAT YOU ARE NOT GUARANTEED.]]


[[Any higher-ups wanna tell me I'm right so that people don't think I'm just some crazy ass kid? @Bryce @JoeB ]]

 

[besttherewas]

BLASPHEMY!!! BURN HIM!!! WITCH!!!! said every cry baby trammie

 

[Bunnky]

Impressive breakdown.

Sent from my LGMS323 using Tapatalk

 

[AlyM]

BLASPHEMY!!! BURN HIM!!! WITCH!!!! said every cry baby trammie

Yeah, that's probably what i'm gonna get. Some one person (or two or three..) will try to shut down the entire breakdown with a single sentence...
But I'm right... and I know it... I just don't understand how nobody else gets it...

Just because you buy 82,693 lottery tickets doesn't mean you're gonna win.
No, because a million other lottery tickets were sold...

Impressive breakdown.

Thanks.

 

[TriXa]

a^2 * b^2 = c^2 ?????
Flubininator's easy explain:
20 People on Scoreboard.
you roll 4 times between 1 - 20.
BÄMMM and you got your winner's

 

[VFingerDiscount]

Its weighted rolls though as she described in earlier post based on % damage dealt.

Sent from Baja 1997 - ICQ 12013388

 

[AlyM]

People have thought many times it was "random" like that, but it is not simply that random.
It is weighted, and the more points earned gives a higher chance, not a guaranteed chance.
Most of the time, powerscroll distribution is weighted towards the top. Had it been 100% random as the above, why did no one on the second page get one? Why are they not rather equally spaced out? Completely random causes things to go in a more orderly fashion, actually.
Roll a number 1 to 20 a total of 100 times. You will not end up with 17 "ones" and 22 "sevens". You will end up with a nearly equal distribution.
The more times you roll that lovely set of number, the more equal it will be. Consider it if you rolled millions of times. You'd likely get
the same or close (within 10) amount with ones as you did sevens.

This system does not work like this. This system is weighted. If you rolled hundreds of times, most of the numbers will be at the top, while a few scattered towards the bottom.

 

[halygon]

You know I hate to ask, especially after your great post. Where did you base these calculations from?

 

[omnom]

They're mostly accurate, but the system was definitely broken for a time. And not just what Alyssa mentioned in regards to resets after world saves (not sure where she got that from), so skepticism is warranted.

 

[Bobaganoosh]

They're mostly accurate, but the system was definitely broken for a time. And not just what Alyssa mentioned in regards to resets after world saves (not sure where she got that from), so skepticism is warranted.

There is something in the patch notes Bryce posted about scores reseting in champ spawns


http://www.uoforum.com/index.php?posts/253967

 

[AlyM]

You know I hate to ask, especially after your great post. Where did you base these calculations from?

What calculations? I tried to give straight reasons for why I did this or that in the calculations. What specific part?
I based the entire thing off the two screenshots I posted, but I could easily do it to another 50 champ spawns.
Its simple math, nothing insane. The most detailed thing I did was binomial distribution to say after repeating a champ 8 times I'd still only have a 40% chance...

They're mostly accurate, but the system was definitely broken for a time. And not just what Alyssa mentioned in regards to resets after world saves (not sure where she got that from), so skepticism is warranted.

But really, the system worked the same way, damage would just be reset. Therefore some people deserved larger numbers than what they got. The system still worked the same way, and can be applied to future champ spawns. When I get a new scoreboard for a champ spawn, I will repeat this post with a new one to help sort things out. PS, You might not have been online but Adam worked really hard to figure out that damage reset problem. He said some defrag code in world saves from a previous developer caused damage to be reset each time. That's why you could come in last wave (...after a world save, or such) and get top damage easily.
Nonetheless, the system worked the same way, it was just flawed because people did not get the numbers they deserved.

 

[Garet Jax]

Is my head supose to hurt like this?

 

[AlyM]

Is my head supose to hurt like this?

Yes, its quite normal. When you get used to playing a game and then you go straight into mathematical functions, probability, and more, you just kind of...
hurt.

Think of it like this... when you haven't exercised in a while, then you just go full out and run a mile... you might be a little sore, your muscles might ache... Your brain is just another muscle... and it doesn't want to do math any more than you want to go running for no reason.

 

[halygon]

What calculations? I tried to give straight reasons for why I did this or that in the calculations. What specific part?
I based the entire thing off the two screenshots I posted, but I could easily do it to another 50 champ spawns.
Its simple math, nothing insane. The most detailed thing I did was binomial distribution to say after repeating a champ 8 times I'd still only have a 40% chance...

What I am saying is how do you know that it uses weighted percentages and that the percentages are in direct correlation with the damage vs overall damage dealt?

By your response is it safe to assume that your numbers are based off of your experience vs actually seeing the hard code that generates the results?

 

[AlyM]

What I am saying is how do you know that it uses weighted percentages and that the percentages are in direct correlation with the damage vs overall damage dealt?

By your response is it safe to assume that your numbers are based off of your experience vs actually seeing the hard code that generates the results?

From GM's responses in the past. One time in IRC a screenshot of the numbers was posted. This has since been changed, but the scoreboard basically multiplied the exact number of hit point damage caused by a certain number based on your skill. If you used a dexxer skill, it was multiplied by two. If you were an archer, by .75. If you were a tamer, .25. This has been CHANGED DRAMATICALLY since, but the same system is in place.. just not those numbers. The GM's have said before in IRC and on forums (on posts that seem to mysteriously disappear?) that the powerscroll handout is "like a lottery"... The more numbers you have, the higher percent you have of getting one... but you don't always get one with a high percent.
I do have the hard code that generates the results.
I do not know if the GM's wish for me to post that publicly though, therefore I will keep that private.
The basic gist (since not many could even read programming) is that it counts points from eligble MOBs (the waves), it doubles the points overall (I believe to give more fairness? make the rolls less or more random?), then does a double roll for the number.
But I've use information I have seen posted in game, on forums, and in IRC by GM's.

I hoped to get a GM to confirm what I'm saying as to provide a bit of foundation, none responded yet.

 

[TheFallen]

So what your saying is its all rigged.
J/K.
All very well laid out but to say it simply: a point = one ticket and there are as many tickets as the total points.
Two small variables only the points of players meeting the eligible criteria count (ie. Alive, within range). Once you've drawn a winner all your tickets are removed(which may or may not be a good thing).

 

[AlyM]

So what your saying is its all rigged.
J/K.
All very well laid out but to say it simply: a point = one ticket and there are as many tickets as the total points.
Two small variables only the points of players meeting the eligible criteria count (ie. Alive, within range). Once you've drawn a winner all your tickets are removed(which may or may not be a good thing).

Basically, although some people get more tickets because they have different builds.
But that's the basic gist...

 

[Bryce]

I know the exact numbers.
From those numbers there is an extra damage multiplier to equate points given.
I will not give these exact numbers.
People fixate on them.
There are modifiers for each template.
There's a lot of truth to what Aly said.

Bottom line here people need to drill into their heads is this:

The more damage you do, the more points you get.
Let's call these 'names on a piece of paper'.

The more of your names thrown into a fishbowl, the more likely one is to be drawn when you reach in and pull one out.

Is it a guarantee? Hell no. Just more LIKELY.
Multiple reasons for this, the foundation two are;
1) More likelihood of awarding those that put in the most time effort
2) Still provide a chance for those logging on and showing up at a spawn halfway in. Otherwise they'd just log back off assuming it was a waste of time.

The initial reason the scoreboard was put in was to show that the system works.

 

[halygon]

Cool -- Good enough explanations for me. Thanks @AlyM for explaining and @Bryce for confirming.

 

[AlyM]

There's a lot of truth to what Aly said.

Is there not truths? ;_;


MATHAMATICALS!