Bot Drop Curious

[Elithas]

This.

However, we are not talking about a single random occurrence and probability of that occurrence (i.e. raid to bot drop %) - we have to factor two other criteria into the formula (there maybe more that FC is using; but we know at least these two):

1. That all general types have the same probability of spawn -and-
2. That each general has the identical drop rate.

Only if these two factors are identical can we apply a model to predict ALL raids with ALL bot drops and place that predictability within the standard deviation.

Since 1 and 2 don't appear (in my raids) to be identical, the ability to predict all raids seems flawed from the beginning. And this "assumes" these are the only other factors to be taken into account - which again may be a giant leap of faith with the FC coding.

I would have to see substantial (if not overwhelming) proof that they aren't identical in both spawn chance and drop chance. Until then I would rely on Occam's Razor.

It wouldn't be so horrible for a dev to just come outright and say what they are though.

Is it possible to hook into AO with a program that could track loot by corpse and upload that data to a SQL database somewhere? You could have fairly reliable data from (potentially) hundreds of people and thousands of raids if it were.

[Hides]

problem I'm having is that there is no more ICE to buy, so I do not know what my drop rate would be now!!

[Pennypacker]

problem I'm having is that there is no more ICE to buy, so I do not know what my drop rate would be now!!

There are other ways to acquire CRU.

[Hides]

yes there is, but I do not want to get 40% or so after 3 hours of running though a mission

[cybertron]

problem I'm having is that there is no more ICE to buy, so I do not know what my drop rate would be now!!

I still have a stockpile of ICE. Send me a tell ingame and we can work something out

[Wakizaka]

yes there is, but I do not want to get 40% or so after 3 hours of running though a mission

IP Run Speed

[Utopian]

I would have to see substantial (if not overwhelming) proof that they aren't identical in both spawn chance and drop chance. Until then I would rely on Occam's Razor.

It wouldn't be so horrible for a dev to just come outright and say what they are though.

Is it possible to hook into AO with a program that could track loot by corpse and upload that data to a SQL database somewhere? You could have fairly reliable data from (potentially) hundreds of people and thousands of raids if it were.

Why would you use a perception of equal-ness to a verified true state of being equal?

We are talking about probability and statistics and how those are implemented into the AO code to simulate true randomness. PC games use RNGs to attempt to provide random chance events - however, those are just code - and some do not produce the desired results.

And - again - based on probability and statistics mathematics, we don't need 1000s of events to demonstrate failure to fall within predicted ranges within an acceptable deviation.

Additionally, if we are trying to determine if faction, city size, profession, toon levels, etc affect the drop rate, we would need to use data only from identical raid types - or you just skew the data and create more confusion.

I am not saying there is a difference, merely by my own data, it doesn't appear that a) the general types share the same probability of spawn and b) the general types share same drop rate.

[Elithas]

Do you understand what Occam's Razor is? It answers most of your questions, or arguments, or whatever you want to call them.

[Chair]

We are talking about probability and statistics and how those are implemented into the AO code to simulate true randomness. PC games use RNGs to attempt to provide random chance events - however, those are just code - and some do not produce the desired results.

And - again - based on probability and statistics mathematics, we don't need 1000s of events to demonstrate failure to fall within predicted ranges within an acceptable deviation.

Additionally, if we are trying to determine if faction, city size, profession, toon levels, etc affect the drop rate, we would need to use data only from identical raid types - or you just skew the data and create more confusion.

I am not saying there is a difference, merely by my own data, it doesn't appear that a) the general types share the same probability of spawn and b) the general types share same drop rate.

Unfortunately, I did not keep a count of our raids, but with respect to the "randomness" or lack there of regarding which general spawns and what percentage of each general drops lead-bots we have had the following experience with raids since early to mid February. If I was to hazard a guess, over approx. an 8 week period we probably did in the neighborhood of 200 ground raids.

Lead-bot drops - 3 arithmetic, 4 enduring, 2 observant, 2 spiritual, 1 supple, and 3 strong

These lead-bots did not drop at anything close to a consistent rate. They dropped more like the game was turning drops on and off for periods of time.

However, I would agree with Utopian and others that leadbot drops and general-spawns do not appear to be statistically random. They may approximate randomness if there are enough people doing raids and/or if all the factors Funcom uses in the code are occurring with a high enough frequency to make the code work as intended. Not that I know what the FC factors are in the code or have access to data to test the statistical randomness. Nor do I plan to do 1000's of raids on my own to get more data.

Without the code or the data - all that any of us can do is to give our own respective opinions anyway. We are all due our own opinions afterall.

Sage

[Utopian]

Do you understand what Occam's Razor is? It answers most of your questions, or arguments, or whatever you want to call them.

So you want to seem smart?

Occam's razor (also written as Ockham's razor from William of Ockham, and in Latin lex parsimoniae) is a principle of parsimony, economy, or succinctness used in logic and problem-solving. It states that among competing hypotheses, the hypothesis with the fewest assumptions should be selected.

Please explain why you believe this has the slightest bit to do with a discussion of random number generating algorthms in an MMO and the basic topic at hand?

We are talking about pure math and the ability to use probability and statistics to gauge whether true randomness exists in an event.

[Gatester]

In order to disprove that it is a random drop you would have to do several raids in a short period of time, but not only have occurances of no bots dropping but also occurances of several bots dropping in a row. Just having "no bots" often is not a sign of anything, given that 10% (the assumed value) is actually incredibly low. Players are posting with no luck, an abudance of luck, and average luck which is all averaging out in the end.

For an example, try getting a 10-sided die to roll a 1 only once every 10 times. Getting 20 or 30 rolls without a single 1 is not impossible and while mathematically it may seem incredibly unlikely it is not really that far off. Each roll will always ONLY be a 10% chance to roll a 1, regardless of how many times you roll the die.


If anything, getting only 1 bot every 10 raids everytime is the most unlikely event possible.

[Elithas]

No, we aren't you were talking about developers, who are very human, programming code.

[Lazy]

ITT: people that don't understand how statistics work discuss statistics.

whether the drops are random or pseudorandom doesn't matter. what matters is that, over a period of a year, 1/10 of the city raids done in game will have dropped a bot. "true" random or pseudorandom or even preprogrammed drops are entirely irrelevant. also, as has been confirmed: the drop rates have not been altered in any of the recent patch and are thus the same as tehy were years ago.

[Gimpynoob]

ITT: people that don't understand how statistics work discuss statistics.

whether the drops are random or pseudorandom doesn't matter. what matters is that, over a period of a year, 1/10 of the city raids done in game will have dropped a bot. "true" random or pseudorandom or even preprogrammed drops are entirely irrelevant. also, as has been confirmed: the drop rates have not been altered in any of the recent patch and are thus the same as tehy were years ago.

This is true because, as everybody knows, secret nerf tables and buggy code causing unintended changes do not exist in AO.

[Lazy]

This is true because, as everybody knows, secret nerf tables and buggy code causing unintended changes do not exist in AO.

they do but threads like this one are franky ridiculous. "herp derp i didnt get bots in 15 raids clearly something is broken" is a tinfoil hat approach to playing that the game can do without. do mistakes happen? yes. has one happened here? no. do we need a dev to come tell us "no, really, drops are fine" every time someone goes on a less than ideal streak? (considering the "unlucky streak" is actually the most probable of all possible outcomes because of way that statistics work that people who usually post in these threads have no idea about) definitely not.

basically, working as intended and hypothesizing on how exactly the loot calculations are done for ai raids is largely irrelevant when the only data we have at the moment is, of all the city raids done on the server since ai launch, roughly 10% of them contained bots.

[Gatester]

It's been a while since I took statistics, but isn't it about the 21st or 22nd raid before the probability of getting a bot to drop exceedes the cumulative probability of not getting a bot to drop over that period?

[Lazy]

It's been a while since I took statistics, but isn't it about the 21st or 22nd raid before the probability of getting a bot to drop exceedes the cumulative probability of not getting a bot to drop over that period?

not quite. the probability of a bot to drop is always 10% (allegedly), regardless of what the previous results were. that never changes.

this is where people are making the mistake. over time, this will result in approximately the correct drop rates, but smaller time frames may vary wildly from 4 bots in 5 raids to no bots after 3 bps.

basically, no matter how many times you flip a coin, the probability of the next flip landing heads is always 50/50 regardless of what the previous combination of heads has been (indeed the previous heads may have been 1 million times, the 1million1th flip will still have a 50% chance of landing heads)

[Utopian]

It's been a while since I took statistics, but isn't it about the 21st or 22nd raid before the probability of getting a bot to drop exceedes the cumulative probability of not getting a bot to drop over that period?

Depending on the standard deviation acceptable - the lowest number of events that used is n=20.

[Vanillashake]

People are confusing the odds of a single event vs a string of events. The odds on any single raid are 1/10. The odds of you going 30 raids in a row is significantly smaller, but that does not mean if you've done 29 with no bots that your odds are magically higher. All it means is your odds of the set of the NEXT 30 being botless are 1/10*30.

[McKnuckleSamwich]

People are confusing the odds of a single event vs a string of events. The odds on any single raid are 1/10. The odds of you going 30 raids in a row is significantly smaller, but that does not mean if you've done 29 with no bots that your odds are magically higher. All it means is your odds of the set of the NEXT 30 being botless are 1/10*30.

no its not.

your math is bad.

you always need to base probability (odds) on the ratio between something happening, and the alternative (not happening).

The chance of any bot drop in any 1 raid is:

(0.1)^1

The chance of NO drop in any 1 raid (ground only) is:

(0.9)^1

The ODDS of getting no bot is

(0.1)^1 / (0.9)^1

The odds are always the ratio of probabilities.

Now, as for your 30 raid question,

The probability of getting NO drop in 30 raids is:

(0.9)^30 = 0.04 = 4%

Whereas, the probability of getting at LEAST one drop in 30 raids is:

1-(0.9)^30 = 96%


Gatester, your assessment is incorrect. The probability of going 7 raids in a row without a drop is roughly the same as that of going 6 raids in a row and getting one drop:

0.9^7 = 0.47 = 47%
1-0.9^6 = 046 = 46%

Meaning, you're just as likely to go 6 raids in a row and get 1 drop, as to go 7 raids in a row and not get any drops.

Over any 8 raids, however, the ODDs of EITHER of those two events happening, diverges, as shown below:

0.9^8 = 0.43 = 43% and as the number of raids goes up, this probability goes DOWN.
1-0.9^8 = 0.57 = 57% and as the number of raids goes up, this probability goes UP.

So, realistically, 8 raids is the break point. If you haven't seen a drop in 8 GROUND raids, your current run is below average on luck for those raids. If you have got a drop within those 8 raids, your luck is above average for those raids.


Hope that clears things up!