Avatar of Fear spawn rate post merge

[Furitor]

Using statistics on Random Number Generators isn't 100% correct. You can statistic all you want, but you can't avoid pseudo-random number generator flaws.

Here are a few topics if you care:
https://www.kdnuggets.com/2017/06/su...andomness.html
https://www.random.org/randomness/

Plenty of topics showing that people try to avoid "streakiness" in their own games:
https://gamedev.stackexchange.com/qu...streaks-in-rng
https://gamedev.stackexchange.com/qu...ber-generation

Depending on how things are coded, you can even exploit randomness very easily depending on the game. Obviously, EQ would be fairly difficult, but for games that rely on save states and so on you can exploit it to oblivion (Pokemon is one such example).

So yes you *can* cause the software to have streaking RNG as by nature RNG isn't true random and therefore you can't just use statistics in a "perfect world" scenario to say "no you can't code streaky rng".

Now whether this is true on P99 remains to be seen as psychology also comes into play here, especially with tedious camps that mess with your brains like Manastone or Rubi BP camps.

[bubur]

they're not micromanaging spawn timers

[Phaezed-Reality]

Originally Posted by Furitor [You must be logged in to view images. Log in or Register.]

Now whether this is true on P99 remains to be seen as psychology also comes into play here, especially with tedious camps that mess with your brains like Manastone or Rubi BP camps.

seems like spell rng, droprate rng, among other things are all the same rng code, and when item drops, or fizzle, or gate collapse or charm break, it all happens at once for a few and then goes back to being random. every single experience of mine has been this repeatedly.

edit: tecmos does a better job at explaining it than me.

[Pootle]

Have to support Tecmos here, as a long time tradeskiller both on live (back in the day) and here on P99, I have experienced the RNG's streakyness many times.

If anyone wants to test this I suggest getting a tradeskill up to around 150+ then gathering supplies for 100 or 200 attempts and burn through them.

[Furitor]

I would assume all the "action" oriented RNG is the same (combat, casting spells, fizzles, charm breaks), but drop-rates and spawn rates may not be. Depending on how you seed any of the aforementioned you can run into some fun bugs.

That said -- I actually personally have experienced what Tecmos has with spell casts (fizzles especially).

I've tried a few things like stopping casting as soon as I see two fizzles to see if I would get a third one 30 secs or 1 minute later (and I did not fizzle). However, we'd need more testing here. If that's the case, then there might be some sort of clock seeding bug (if they're using the clock for a seed, which would be the easiest thing and also have it's own flaws).

Again, without seeing the source, it's very hard to tell (even seeing the source it might be hard to tell lol).

Then again... the streakiness might just be classic lol.

[Tecmos Deception]

Sense heading fails a ton of the time the first time you use it, even at 200 skill. Then it works 20 times in a row. Not random.

And about everything else feels similar.

Originally Posted by Furitor [You must be logged in to view images. Log in or Register.]

Then again... the streakiness might just be classic lol.

Yeah, I wouldn't doubt it.

I don't really mean to sound like I'm complaining, it's just hard to get out of the RPG mindset that stuff is randomly determined and expect goofy streaks, even after all these years of p99.

[Lordgordon]

There's actually a very little man on the other side of the screen deciding if you've earned your pixels or not

[tommydgun]

Do you even know how it works? You are saying 3 resists in a row... that doesnt compare to 32 PHs in a row observed back to back days... the cumulative probability of that many failed named spawns in a row is less than 0.01%. That happening twice in two days is not random lol. But trying to explain math to you is clearly impossible

[Tecmos Deception]

Originally Posted by tommydgun [You must be logged in to view images. Log in or Register.]

Do you even know how it works? You are saying 3 resists in a row... that doesnt compare to 32 PHs in a row observed back to back days... the cumulative probability of that many failed named spawns in a row is less than 0.01%. That happening twice in two days is not random lol. But trying to explain math to you is clearly impossible

Originally Posted by Tecmos Deception [You must be logged in to view images. Log in or Register.]

You can fight the same exact mob forever, have an average resist rate on a given resist type of 10%, but still see 3 resists in a row happen every couple hours for days on end.

You're talking about a "less than .01%" chance thing happening twice in two days. My example was a .1% chance thing happening every couple hours ad infinitum (aka a streak of 3 resists happening every ~100 spellcasts instead of the every 1000 you'd expect, on average). My example was inspired by actual logs and a thread I made a couple months ago, and IIRC my overall resist rate on a given mob was actually only 7-8% but I still saw 3 resists happen a few times and 4 resists in 5 casts happen once in the matter of only like 5-6 hours of play.

They're both examples of how what we call "RNG" on p99, whether it concerns spawns or drop chances or spell resists or melee swings or tradeskill skillups or whatever, is streakier than it should be if it were actually random. And it seems streaky enough to me, just on a gut feeling anyway, that it is unlikely to just be a pseudo-RNG that is accidentally a bit wonky.

[fadetree]

Tec and the others are right. Anyone who has done any tradeskilling can clearly see a repeated pattern-- long string of fails, then a couple successes almost one after the other, then another string of fails. This is so consistent that I am almost convinced there's some internal state information associated with the generation process...meaning that it's not just a bare RNG, there's some other processing going on in the code.
We don't even know whether the same RNG process is used for everything, or there are different routines for different activities. The last time this came up I was going to log in and do like one thousand /rand 1000, then process the log and feed it into a randomness analyser to get the chi-squared and other measures of how good the randomness is, but I got tired and wandered off.
I was also thinking that we could look at the code-- isn't the stock emu code available somewhere? I doubt that P99 would have custom re-engineered the RNG. Looking at code could explain a lot.