Bot Drop Curious

[Gatester]

I meant when it becomes ~10% to not have a bot drop based on probability. Given 10% chance of a bot dropping each raid I was looking for the point at which not getting a bot over a series of raids was equivalent to the chance of getting a bot to drop in a single raid.

.9 ^ 21 > 10% > .9 ^ 22

Of course, not having done statistics in a while that probably was incorrect and what you posted looks familiar Noobas, but now I am a bit confused so I'm not gonna stick with not trying to compare one raid to another when it comes to probability


People win the lottery, and the chances of that are incredibly slim. People have won the lottery several times, and the chances of that are inconceiviable. Not getting bots after a long time is nothing more than hitting the lottery of bad luck

[McKnuckleSamwich]

I meant when it becomes ~10% to not have a bot drop based on probability. Given 10% chance of a bot dropping each raid I was looking for the point at which not getting a bot over a series of raids was equivalent to the chance of getting a bot to drop in a single raid.

.9 ^ 21 > 10% > .9 ^ 22

My mistake, I misunderstood your post.

The math in my post is correct though, regarding the other aspects of drops etc.

What nobody has discussed yet tho and what is so much mroe relevant is the drop rate on ground+ships

which is significantly higher.

Chance of no drop on ship raid is:

0.9*(0.9*0.9)

0.9 on ground, 0.9 chance of no coons, 0.9 on admiral

= 0.729 chance of no drops

which is a huge difference compared to 0.9 on ground.

Chance of going more than 5 raids in a row without a drop on full ship raids is

0.729^5 = 20%

Meaning, in any 5 ship raids, you're nearly always going to get a drop... even if it is a crappy bot.

[Vanillashake]

no its not.

...

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!

Lol, learn to math.

For one, 1/10 means just that, 1 in 10. Not 1 in 8. There is no break point, your odds on any single raid, regardless of previous streaks/lack of streaks is 10%.

Odds of a series of events are different. You have an incredibly small chance of 100 raids in a row without bots, but that does NOT mean that if you've done 99 without a bot that your odds are sky high the next raid will drop anything. You have a 3%* chance of going 0 for 30, but that does NOT mean that on raid #30, you have a 97% chance of getting bots. This isn't poker, there is no ace in the deck that has to show up eventually. Your math would be right if it was, though. Not getting bots does not eliminate not getting bots from the droptable.

It is far simpler than some of you are trying to make it.

As an aside, I've got 7 of 9 before and I've also gone 0-24. Neither were luck, because my overall rate that I've tracked is 12.8% over 200some raids, which is a small sample, really. That's just how it works. At least Lazy gets it.

[Gimpynoob]

Can I just throw a little wrench into all this theorizing? Thank you.

I see that everybody is assuming something that MAY not be true. That the drop rate remains constant.

Game developers often use RNGs that take previous drops into account and "adjust" the drop chance in accordance. This can have a huge impact not only on the "luck" of people depending on the time they play but also on the server population. Joining the population of the 2 servers into one may result not only in less bots per day for the same amount of "farming" of the 2 previous servers but also in people running out of luck earlier and farmers who had previously had a drop rate closer to average become more "unlucky".

[Vanillashake]

You're giving Funcom a little too much credit.

[Gatester]

@Vanillashake, Noobas was simply discussing the mathematical probability in response to my own post, just statistics and not some predictable occurance. Anyone who understands statistics at least on a basic level knows that statistics and probability themselves never determine outcomes.


I wonder how the occurance of double bot drops affects the rate as well. Is it 10% chance and then if a bot is to drop is it another 10% chance? Is a triple bot drop possible too?

Either way, the only time players should be worrying about whether the droprates are broken or not is if no one on the server is getting drops.

[Leetlover]

Droprates looks as b4 but 42 and ai-raids aren't played very much afaik. No bots for j00.

[nums214]

^This. Random chance is as random as ever.

I learned long ago there is a pattern to the randomness, and it has to do with your AI level, your Gear, and your rank. Certain toons will get 1/1200 and others will get 4/5 all day long.

[Gatester]

I learned long ago there is a pattern to the randomness, and it has to do with your AI level, your Gear, and your rank. Certain toons will get 1/1200 and others will get 4/5 all day long.

So my enforcer ranked general in DB armor/ofab, some beta symbiants, and around AI 20 gets no bots in 63 raids and my 74 MP in full combined officer, AI 8 and enough axp for the next level, and ranked squad commander acquired 7 supple bots from 6 raids because of some pattern involving those variables?

Could you be more specific as to what factors contribute to this pattern, and perhaps how to increase drop rates based on those factors? Would you also say that two characters with entirelly duplicated armor, rank, and AI level would have the same drop patterns?

[Unkfix]

I learned long ago there is a pattern to the randomness, and it has to do with your AI level, your Gear, and your rank. Certain toons will get 1/1200 and others will get 4/5 all day long.

Shhh don't tell the secret! You forgot to mention the date factor.

[Manicmouse]

Could you be more specific as to what factors contribute to this pattern, and perhaps how to increase drop rates based on those factors? Would you also say that two characters with entirelly duplicated armor, rank, and AI level would have the same drop patterns?

You must test all permutations!

[Gatester]

You must test all permutations!

Or you could just test 3 duplicated characters. Simple to verify.

[Manicmouse]

Guess they don't have a sense of humour where you're from

[Unkfix]

In my experience, the characters in team, the characters looting, and the date all play a role in chance of loot. Not just in raids either.

[srompu]

Can I just throw a little wrench into all this theorizing? Thank you.

I see that everybody is assuming something that MAY not be true. That the drop rate remains constant.

Game developers often use RNGs that take previous drops into account and "adjust" the drop chance in accordance. This can have a huge impact not only on the "luck" of people depending on the time they play but also on the server population. Joining the population of the 2 servers into one may result not only in less bots per day for the same amount of "farming" of the 2 previous servers but also in people running out of luck earlier and farmers who had previously had a drop rate closer to average become more "unlucky".

If they were to keep a constant percentage droprate, I don't think population/merge would have an effect. As the number of people raiding goes up, so do both lucky and unlucky drops.

If for some dumb reason there was instead a time-dependent droprate, with say 10 supples added per server per day, then I think we'd see what you're describing.

On a slightly different topic, every time I set out to test the droprate, I always end up close back to 10% after 70+ raids. I have a theory that as soon as I start writing down general name, ql, and drops, the game starts behaving.

[Gatester]

Guess they don't have a sense of humour where you're from

Nah I got it but an orgmate was saying something similar about the drop rates last night though and I was wondering how exactly AO is generating their "random" values. Maybe it is taking into account some ridiculous variable like AI kills, your exact AXP value, or some other unusual factor.

The thought of testing every permutation though, it put a chill through my spine lol. Maybe we should start a thread theorizing what factors AO may use for RNG droprates and possibly even testing them to see if they actually apply? I have issues with things dependant on the characterse themselves, however, considering that I do not believe a mobs loot table is dependant on the player who killed it. The delay between a mob dieing and their body becoming lootable, however, leads me to believe AO is relying on something such as a time-dependant factor. This is if a mobs loot is not generated at the point it spawns rather than the point it is killed either.

I vaguely remember reading about a GM investigating a mobs loot while it was still alive, so whether this is true or not would further complicate things.

[Notcrattey]

Snip

I've heard numerous times that the loot is set on the spawning of the monster.

[Gatester]

I've heard numerous times that the loot is set on the spawning of the monster.

Either way, it probably does not rule out the possibility of city raids being dependant on the participants themselves, which is a terribly complicated variable to deal with unfortunately.

[Elithas]

Why would someone program in a variable that would be such a pain to program in? Or do you guys think it is some unintended consequence?

I understand RNG strings giving similar loot on similar days/time periods, but the rest of it, at a glance, just seems like total bull to me.

[Gimpynoob]

Loot being dependant on the toons characteristics is highly improbable but it being dependant on previous drops is not. RNG formulas that decrease chance of loot dropping for loot intended to be rare is not something new afaik. Such a mechanism would explain why after merge some players keep having similar drop rates and others started having much worst ones.

Its all theoretical not only because low drop rates variyng with time would require extended colaboration of people playing in very different locations but also because data would be unreliable.