Math and Spawn Rates

[Snagglepuss]

I've noticed that we have some mathematically oriented minds on the forums (Tecmos, Loraen, Koros, Ele, etc). It seems like every few weeks someone is asking about what is normal for a spawn rate as far as what is a good run or bad run of variance and to what extent.

I've played a enough poker to understand things like risk of ruin and how variance can be your best friend or your worse enemy. Many older players at the table will always say to a bad beat, "I'd rather be a lucky player than a good player..." From the point of view of a professional poker player, you should look at each table siting as the piece of one big game that never ends. This way, your actions that have positive expected value will pay off over time...

Now, what I want to know concerns spawn rates. I was camping a mob that "according the wiki" had a 5% chance to pop on a standard outdoor timer. Being static, I had no reason to worry about any other variables or PHs. The bugger was killed immediately as he spawned, then I would reset my timer. Between two people, we camped the mob 29 hours straight this weekend to no avail. I guess we saw around 265 to 270 spawns.

Obviously, this is a run of bad luck, but what I'm interested in knowing is how bad. If you normally distributed its chance to pop, what would be considered outside one standard deviation? Also, should I look at this knowing the old gambler's fallacy and interpret my chances of camping a mob as "one big camp." It's hard to tell yourself after a 12 hour session that he's got to pop within the hour, but we all know fortunes have been lost from the words, "I can't keep getting these cards. One more hand; I know I'll get some good cards soon!" All actions are independent of the outcome, correct? Every chance is 5%, even if you have gotten no pops 267 times in a row? Wouldn't that be as unlikely as getting 12 pops in a row though? Still 5%...

Could someone explain some math or statistics about how long a camp like that could take and still be within the realm of "reasonable bad luck"?

[Vayder]

If the spawn chance is really 5% and you were killing the correct PHer you're results are literally right around 1 in a million (1.13 in a million rounded).

This is far more likely than getting the named to pop 12 times in a row however (which is around 24 in a quadrillion chances).

[Itap]

Reminds me of raster camp in lguk. I remember helping a friend do it back in 2001 and we sat down there for around 36 hours straight before he popped.

[Ele]

Sounds like a pained soul thread. That camp leaves plenty of time to ponder the deeper things in life.

You've already addressed the biggest mistake I've seen people make about these kinds of camps, which is the gambler's fallacy. "Just one more spawn, I've been here forever, it has to happen soon!" Each and every spawn is an independent event (in most cases) that has no bearing on the what the last place holder was. It was 5% last time, it is 5% this time.

With a 5% chance to occur, there is a 95% chance that it does not occur.

[Itap]

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

"Just one more spawn, I've been here forever, it has to happen soon!"

Lol, I think this is just what people to say to keep themselves from slipping into senility.


What hurts more is when soandso goes in there and gets spawn within an hour.

[zanderklocke]

I sort of just posted this on another thread, but I'm wondering where people pull most of their data from spawn rate percentages to drop rate percentages. I'm not a computer wizard, so I'm not sure if people have readily available access to source code somewhere? Or, is this proprietary information, and I'm just being naive?

[Snagglepuss]

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

I sort of just posted this on another thread, but I'm wondering where people pull most of their data from spawn rate percentages to drop rate percentages. I'm not a computer wizard, so I'm not sure if people have readily available access to source code somewhere? Or, is this proprietary information, and I'm just being naive?

Zander, I got mine from the wiki, but I'm not sure how accurate it is. I've heard that the data is pulled from a source other than P1999. Anyone confirm?

Ele, I was refering to the Rotting Skelly, but yes, both camps are very... well.. let's say you can get some good reading done haha.

If you had coin that was skewed to land on heads 95% of the time and tails 5% and you flipped it every 7 minutes, what is the normal distribution of times you'd get a tails? It would be cool to see some projections where you could compare to eq spawns to get an idea of where you are falling on the luck curve. Isn't the AC like 10% in OoT?

[eqravenprince]

My guess is it is not a 5% chance of spawning. 1% sounds far more likely.

At 1% chance of spawning
After 100 kills: 36% chance of not getting spawn
After 200 kills: 13% chance of not getting spawn
After 300 kills: 5% chance of not getting spawn

At 2% chance of spawning
After 100 kills: 13% chance of not getting spawn
After 200 kills: 1.7% chance of not getting spawn
After 300 kills: .2% chance of not getting spawn

At 5% chance of spawning
After 10 kills: 59% chance of not getting spawn
After 25 kills: 28% chance of not getting spawn
After 50 kills: 8% chance of not getting spawn
After 75 kills: 2% chance of not getting spawn
After 100 kills: .5% chance of not getting spawn
After 200 kills: .0035% chance of not getting spawn
After 300 kills: .000020753% chance of not getting spawn

[webrunner5]

We are talking a Random Number generator here on some of this stuff. Good luck guessing that at times. And as the rumor mill goes most people including Devs have never figured out how Sony did it yet.

[Frieza_Prexus]

The RNG on live was considered to be notoriously streaky. My inner conspiracy theorist suspects that something has been done here in an attempt to recreate the frustratingly "off" nature of EQ's RNG.