Southperry.net
Chaos Scrolling Probabilities (C++) - Printable Version

+- Southperry.net (https://www.southperry.net)
+-- Forum: Social (https://www.southperry.net/forumdisplay.php?fid=14)
+--- Forum: Rubik's Cube (https://www.southperry.net/forumdisplay.php?fid=58)
+--- Thread: Chaos Scrolling Probabilities (C++) (/showthread.php?tid=8055)

Chaos Scrolling Probabilities (C++) - »-Chris-> - 2009-02-07

Doing something with my brother, wondering if this works. Basically finding the chance of getting a certain attack, or range of attacks using chaos scrolls. I'm just learning how now, so I have no clue myself. =x

Quote:#include <cstdio>

void series(int total, int count, float temp);

float per = 0;

int main()
{
int total = 2;
series(total, 1, 1.0);
printf("%f\n", per);
getchar();
return 0;
}

void series(int total, int count, float temp)
{
for(int i = -6; i < 6; i++)
{
int tempTot = total;
float t = temp;
if(i == -6)
t *= 0.4;
else
{
t *= 0.6 * (1.0/11);
tempTot += i;
if(tempTot < 0)
tempTot = 0;
}
if(count < 5 && tempTot != 0)
series(tempTot, count + 1, t);
else if(count == 5 && tempTot >= 10)
{
per += t;
}
}
}
Thanks,
Chris

EDIT: Adding on stuff now so can change the amount of scrolls and the base easier.

Question: Is it -5 to +5 on all stats on the item? How does this work?

Say for a pink cape.

2atk
15 def
15 mag def

Outcome:

7 atk
20 def
20 mag def

Or-

7 atk
18 def
17 mag def

I've heard different possibilities, which one is the best maximum output?

Chaos Scrolling Probabilities (C++) - kleptophobia - 2009-02-07

Chaos scrolls are the same odds for each outcome? I thought it was +-1,2,3,5 to all stats present on an item. I do not know if we have sufficient data to determine what the probability of each outcome is, although one would assume the + are equal to the -.

Chaos Scrolling Probabilities (C++) - »-Chris-> - 2009-02-07

I believe it is an equal chance for all 11 possible outcomes.

Chaos Scrolling Probabilities (C++) - kleptophobia - 2009-02-07

So does that mean +-4 is a possibility?


Edit: You're response should be yes, as you list +4 and -4.

Chaos Scrolling Probabilities (C++) - »-Chris-> - 2009-02-07

No,
+5, +4, +3, +2, +1, 0, -1, -2, -3, -4, -5.

Thus 11 different possibilities.

Chaos Scrolling Probabilities (C++) - »-Chris-> - 2009-02-07

Here, in python.
Quote:def ScrollProb(start, maxcount, minnum):
per = scrollProbLoop(start, 1, 1.0, maxcount, minnum)
print (per)


def scrollProbLoop(total, count, prob, maxCount, minnum):
per = 0
for i in range(-6, 6):
tempTot = total
tempProb = prob
if(i == -6):
tempProb *= 0.4
else:
tempProb *= (0.6 * (1.0/11.0))
tempTot += i
if(tempTot < 0):
tempTot = 0
if(count < maxCount and tempTot != 0):
per += scrollProbLoop(tempTot, count + 1, tempProb, maxCount, minnum)
elif(count == maxCount and tempTot >= minnum):
per += tempProb
return per

Appears to be it.

Chaos Scrolling Probabilities (C++) - Shardsniper0 - 2009-02-07

The chances of +5 is even lower then I thought? How does it work that way?
+5+4+3+2+1+0-1-2-3-4-5
If the scroll has a 60% chance of working:
there is 11 outcomes.
The Chances of getting +5 stats is like:5.45454545%? So the chances of getting 5 chaos scrolls to spit out a perfect item is .00000125125125%? : (.

Chaos Scrolling Probabilities (C++) - Stereo - 2009-02-07

[Image: maple2494ma4.jpg]
-3
-2
-1
+1
+1
+2
+2 (20 hp)
+4 (40 mp)


[Image: maple0001sn8.jpg]
-4
0
0
+1
+1
+2
+3

[Image: maple0039an5.jpg]
-4
-2
-2
-1
+2
+2
+2

So in 20 chaos changes:
-4: 2
-3: 1
-2: 3
-1: 2
0: 2
1: 4
2: 6
3: 1
4: 2


I'm not altogether certain +5/-5 can happen with equal likelihood... anyone got pics?

Chaos Scrolling Probabilities (C++) - Harrisonized - 2009-02-08

Stereo Wrote:I'm not altogether certain +5/-5 can happen with equal likelihood... anyone got pics?[/color]
It is possible, because Chaos scrolls work EXACTLY the same way as the Maple Syrups back during the event with them. I myself, tested this, by first scrolling HP onto a Maple Leaf, and then using the Syrups to change the number of stats I had.

I've examined a dozen of pictures of Chaos scrolls before and after, and I've noticed that existing scroll patterns happen the most frequently. For example, a common one, when there are two stats, is ±1 and ±2, which follows the pattern of a 60% weapon scroll. Another common one is ±1, ±3, and ±5 when there are three stats that can be affected.

Back when Chaos scrolls first came out, I attempted to explain the effects of them. Thus, I examined the hundreds of screenshots of my Maple Leafs. After much investigation, I found out that these combinations were possible.

Single stat on armor:
±0
±1
±2
±3
±4
±5

Double stats on armor:
±1 and ±1
±1 and ±2
±2 and ±3
±2 and ±4
±3 and ±1
I have never seen:
±3 and ±4
±4 and ±4
±4 and ±5
±5 and ±5

Triple stats on armor:
±1, ±5, and ±3
I've seriously never seen another combination for this, though I know it should exist.
Usually, if it doesn't change in that order, it does one of the double-stat patterns to two randomly chosen stats.

The changes get MUCH more complex after three. However, I speculate you can group them off and call them "existing patterns"
Stereo Wrote:-3
-2
-1
+1
+1
+2
+2 (20 hp)
+4 (40 mp)
±1 and ±2
±1
±1 and ±2
±2 and ±4
[quote=Stereo]-4
0
0
+1
+1
+2
+3
±4
±1 and ±1
±2 and ±3
Stereo Wrote:
-4
-2
-2
-1
+2
+2
+2
±2 and ±4
±1 and ±2
±2
±2

Another thing I noticed with these is that there is a hierarchy of % in which a certain number of stats are changed. For example, in one stat scrolling patterns, ±1 and ±2 happen much more often than ±3 or ±4. Likewise, ±1 and ±2 happen more often than ±2 and ±3 for two stat patterns.

I hope this helps, and I hope the TS statistically breaks down the effect of a Chaos scroll correctly and posts a chart because that would be extremely helpful.
Stereo Wrote:
-4: 2
-3: 1
-2: 3
-1: 2
0: 2
1: 4
2: 6
3: 1
4: 2
Yea that seems about right. For some reason, 2 was always the most occurring in the leaves.

Chaos Scrolling Probabilities (C++) - »-Chris-> - 2009-02-08

6 atk shoe
24 def
7 mag def

Chaos Scrolling Probabilities (C++) - DeadHead - 2009-02-08

For the range -5 to +5, -2 and +2 are just two of the most common random numbers due to the nature of pseudo random number generators. The extremes (+5 and -5) most often appear the least.

BTW, I'm Chris' brother, the one who has been writing these programs.

I've got a question for you Harrisonized... around how many scrolls have you seen the effects of for the multiple stat patterns?

Chaos Scrolling Probabilities (C++) - Harrisonized - 2009-02-08

DeadHead Wrote:For the range -5 to +5, -2 and +2 are just the two most common random numbers due to the nature of pseudo random number generators. The extremes (+5 and -5) most often appear the least.

BTW, I'm Chris' brother, the one who has been writing these programs.

I've got a question for you Harrisonized... around how many scrolls have you seen the effects of for the multiple stat patterns?
I have not seen as many multiple stat patterns as I have seen single stat patterns, but I've seen quite a few. Back when they were first out, the most commonly Chaos'd items were FaceStompers and Pink Capes. I've seen a few on Zhelms, and that one Stereo posted with his Deputy star, and personally, I've used two chaos scrolls, both on my HTP.

A good place to search for scrolls and their effects are on Sleepywood, since after all, people posted there to show off when they used it "back in the day" that it was expensive:

+1 +2 on a 3 stat item
http://www.sleepywood.net/forum/showpost.php?p=8688264&postcount=79

-1 -3 -5 on a 3 stat item (It's -3 but you cannot have negative weapon attack)
http://www.sleepywood.net/forum/showpost.php?p=8692543&postcount=85

+1 +3 +5 on a 3 stat item (FS clean are 2/20/5)
http://www.sleepywood.net/forum/showpost.php?p=8710211&postcount=1

+4 -5 on a 3 stat item (which interestingly, I was unable to recall ><)
http://www.sleepywood.net/forum/showpost.php?p=8721244&postcount=129

-2 on a 3 stat item
http://www.sleepywood.net/forum/showpost.php?p=8722884&postcount=135

+3 +3 +3 +1 on a 4 stat item (which then could be broken down to +3 +3 and +1 +3 which is an existing scroll pattern)
http://www.sleepywood.net/forum/showpost.php?p=9071970&postcount=25

Good thread on single stat items (maple leaves)
http://sleepywood.net/forum/showthread.php?t=1490912

The thread on my HT Pendant
http://www.southperry.net/forums/showthread.php?t=5188

I didn't include the various ones in Basilmarket because those don't include the initial stats. Also, you might want to try searching again because there are a hella lot more that I probably missed.