PDA

View Full Version : Rare enemies; why people would think it's 1/400



Tycho
Jul 25, 2004, 10:32 PM
This is just a theory; I'll assume the actual rate is 1/512.

I'm looking at the number of enemies it takes to get a rare monster, since most of the time, if it doesn't have the drop, you'll just restart the quest and go to the same place again to pipe, until you get a new rare enemy.

The chance you'll get a rare monster instead of the usual one in 512 or fewer monsters, is:

1 - (511/512)^512 = 0.63248

The chance you'll get one in 256 or fewer enemies, is:

1 - (511/512)^256 = 0.39376

The chance for it to take at most 256 more, is:

1 - (511/512)^768 = 0.77719

See the differences in the cumulative chances per 256? It's:


1 - 256: 0.39376
256 - 512: 0.23872
512 - 768: 0.14471

So, even if the 1/512 rate is true, it's still more likely to take fewer enemies, than more. So that's why I think other estimates are always lower than the commonly accepted 512.

I may be wrong though. Even if I believe these calculations are correct, the assumptions might just be because the odds actually -are- better.



<font size=-1>[ This Message was edited by: Tycho on 2004-07-27 09:55 ]</font>

Jason
Jul 25, 2004, 11:08 PM
I had no idea you are good at math.

Tycho
Jul 25, 2004, 11:53 PM
Well, a bit. Just finished high school, so the least I could do is use the stuff I learned.

Stan64
Jul 26, 2004, 06:29 AM
Also heard of the beat theories?

rena-ko
Jul 26, 2004, 06:37 AM
hell no, dont start about that silly beat theory again...

Azbats
Jul 26, 2004, 06:44 AM
I always thought the best way to get a rare monster to appear was to set up your BBQ in the middle of the room and start cooking some stakes.

Worked for me [although there was this time I had to get the sausages out... hmmm]

In all seriousness though, I have found that up to an hours worth of piping usually brings in a rare monster. So 1/400 etc... it's all a matter of patients. And that is a virtue after all!

Stan64
Jul 26, 2004, 08:55 AM
I always get a rare atleast in an hour if I do the beat piping. o.o

Superguppie
Jul 26, 2004, 09:06 AM
Hm, sounds like we have another believer of the beat theory there. Sorry Stan64, I've seen so many people go against any relation to the time of day (which beat time is!) I am pretty much convinced it won't work. And before you can convince anyone it does work, I think you will have to do the following:
- Pipe for 24 hours straight and write down the beat for each rare enemy you get. Do take care to keep the number of pipes (and enemy spawns) per beat the same.
- Repeat this for at least 100 days.
- Make a graphic representation of the data you have gathered and look for clusters of occurences at certain beats.

You're probably not mad enough to do this, but until you do, don't expect anyone to believe you.
And this is a mild reaction, compared to what some others could do. Some are so fed up with the whole beat theory thing that little line of yours could lead to some serious flaming...

And don't start "I always pipe at this or that beat and get my rare" because that's a crap argument. If you always pipe at the beat and never off it you it's no surprise you get some at the beat and can't tell you didn't get it off beat. The self-fullfilling prophecy is a soap-bubble most of us can pop by just looking at it.


<font size=-1>[ This Message was edited by: Superguppie on 2004-07-26 07:37 ]</font>

Jess_01
Jul 26, 2004, 10:04 AM
theoretical it is right that you can expect that the chance that you get a rare enemy by the 256th try is higher than the chance that you get a rare enemy by the 512th try
but I don't think that's very relevant for the game

Stan64
Jul 26, 2004, 01:27 PM
I know the arguments and fights that have been. I strongly think that certain beat periods yield more percentage to find a rare enemy than others. Maybe dropping rares too. I have some happenings that makes me believe so.

Hrith
Jul 27, 2004, 04:21 AM
On 2004-07-26 11:27, Stan64 wrote:
I know the arguments and fights that have been. I strongly think that certain beat periods yield more percentage to find a rare enemy than others. Maybe dropping rares too. I have some happenings that makes me believe so.
that's where you're wrong, your experience means nothing.

KaFKa
Jul 27, 2004, 04:31 AM
sigh...

the beat time is not calculated into the equation. proven as fact, have a nice day...

FattyMcGee
Jul 27, 2004, 05:25 AM
Fewer monsters, not less monsters http://www.pso-world.com/psoworld/images/phpbb/icons/smiles/icon_wink.gif

I don't really care whether it's 1/400, 1/512 or 1/1000. Either way I'm going to hunt rare monsters til they drop me the item I want, and it's all a matter of chance anyway. If I do kill 511 of a monster, that 512th one will still have the same chance of being rare as the very first one.

It's all chance.

--fatty

heyf00L
Jul 27, 2004, 12:51 PM
Actually it does rely on time. Random number generators use time of day as the seed for the hash function that makes the random number.
But, that's down to the millisecond so unless you can plan your pipes to the 1 /1000 second you won't be able to finding rare enemies based on time.

Tycho
Jul 27, 2004, 07:09 PM
But... well... if that's true (and I guess it's likely to be true), shouldn't people be more likely to encounter two rares at a time? Or would .000001 second have passed by the time the game starts calculating if the next monster is rare?

Superguppie
Jul 28, 2004, 04:48 AM
The time is only used as a seed. So even if you draw 2 random numbers from the same seed, you won't get the same number twice. Or actualy, you could, but it is highly unlikely. Also, last time I studied random number generators, it was customary that the seed was only used to start the generator. If that goes for this one too, that would mean only at the start of the session would the time be of influence. After that the order of actions drawing random numbers would determine what action gets which number. Of course it is possible to seed the generator with the time every time you need a number, and then you would really need split-split-split-second timing to get the same result. That is, provided there are no other influences.
Second remark is that even if you do take this as a time-dependency, it's still NOT the time in beats that counts. It is to be expected that all beats have an equal chance of having a time in them that will work.
And thirdly, even though the sequence a random number generator gives is fixed, it is unpredictable. Only when you get the seed (and other influences) exactly the same do you get the same sequence.

PyroticTigerX
Jul 28, 2004, 08:29 AM
I believe it's the same system they did on Shining Soul/Shining Soul II ... whichever item you can get from said monster/box changes on the fly, if you wait even a part of a second to kill that monster it'll drop something else, actually, I think the rate of change (on Shining Soul II at least) per second is like 4 or 5

PSO has loads more items, I bet they changed it to a much greater number than that, maybe even 20 or 30 a second, then again, it might still be 4 or 5, but when you hit one then it gives a table and listing down each table is what you could get (like one table might be all weapons, another all armor, a third nothing) and only one of those tables will have the rare item (maybe multiple times) and however many blank spots to fill in the gap

on rare monsters, I'm guessing more like 1/1024 or something, that's still 10 bits of random numbers, and a very low chance they'll all =1

think about it, Pal/Al Rappies are the most common 'cause you can get 7-8 in a room, how many times do you hafta pipe on average to get a Pal Rappy? my numbers seem to be a good 40+ a shot, that's 320+ rappies, but while you might be thinking that it would be 320/512 to throw in there, the actual chance that one would be a Pal Rappy is 320/163840 ... no better than 1/512, eh? each rappy has the same chance as the next, it's not dependant on the one you got before, it's an individual chance... go fig