Loot autosplit

[Carnalismo]

Wondering about how the loot is divided within the group when you pick it up (talking about the sell-loot). I've heard that it's supposed to be random, so it will be divided equally, but to me it doesnt seem that way. When hunting with full groups, I almost never get any of the sell loot, just once in a while. And I can clearly see a pattern of who's getting the most. My question is how it's really decided?
I know it can't be classes or dmg output, since I very often group with lowerlevels that are the same class as me, and they're getting a "normal" amount of loot, and I'm getting none. And it hasn't been like this all the time either, just over the last 15ish levels, maybe I got some kind of cursed item on me

Reason I'm asking is not because of the money you get for the sell-loot (although thats nice). But since DF is coming up, it would be kind of sad hunting for five hours, and always ending up with all the others having ten times the seals that I have.

Anyone have any ideas about how it's actually divided? Or am I the only one who has noticed this?

 

[Alrindel]

It's random. It may not always seem random, but that's just the nature of randomness.

DAoC crafters are the same way, always trying to find the secret pattern that will help them figure out when is the best time to craft in order to gain skill points. But it's just plain random.

Anyways, remember that you can switch autosplit off if you want.

 

[Roo Stercogburn]

Isn't it only the group leader can switch autosplit off?

 

[Sortbane]

you can set your own autosplit settings, but only choice is to set it so you wont get any items with /autosplit S

 

[Seyrcim]

I'd say get used to it. I also have a skald and realy, after about lvl32-35 I hardly get sell loot, and when I do get it, it's always the same kind of sell loot (like, I get all onyxes). I also remember a friend who always got the poisonous funguses in spindel (he still gets to hear about his love of funguses (but then... he IS a zerker )

We had a little trip in DF last week on gorre and we got quite a lot of those diamond seals (don't remember the exact name tho). All in all the group (8 ppl) got ~30 seals and I ended up with two of those. So all in all, a bit below average, but still not 'unfair'


[edit] Uhm, strange sentence... :sleeping:

 

[Tigerius]

The problem is that DAoC doesn't randomize very well. Not anything anyone can exploit I'm sure, but most people who look at numbers much enough will realise that the random engine is pretty poor at being random.

In DF it would be adviceable to give all seals to leader, but that probably won't work with the amounts recieved and stacks unable to be splitted afterwards.

 

[old.chesnor]

Random numbers are notoriously difficult to produce with a computer. This game does its best, but the problem isn't with the numbers, its with the humans who play the game. I.e. I play the game for a 5 hour period, I fight 100 even cons mobs without death or incident, suddenly I miss 4 hits in a row and die. A poor random number generator, or just poor luck ?

You see if you flip a coin 10 times, you are statistically as likely to get heads 10 times, than you are 5 heads and 5 tails. Thats probability and thats what humans are just not very good at understanding.

And on the DF seals issue, if they drop in your bag, they are yours and should be yours. Handing them over to the group leader so he can distribute them fairly will end in tears, believe me. Daoc may appear to favour some people, and shaft others, but it really doesn't

 

[- Pathfinder -]

Which is why I prefer guild groups. No loot problems

 

[old.yaruar]

Originally posted by old.chesnor

You see if you flip a coin 10 times, you are statistically as likely to get heads 10 times, than you are 5 heads and 5 tails. Thats probability and thats what humans are just not very good at understanding.

Actually no you aren't you are far more statistically likely to get 5 and 5 due to the fact that there is only 1 combination which gives all heads and many which give 5 heads and 5 tails.

Pascals triangle and all that (tries desperately hard to remember any maths...)

 

[old.chesnor]

Yaruar is the human I was referring to obviously. When I flip a coin, it can land in 1 of 2 states. Heads or tails. There is a 50% chance of each state. The coin has no memory, and does not know what state is has previously landed in. If you throw 5million heads in a row, you chance to throw a heads next is still 50%. The probability to throw a given state is always the same, 50%

What you are talking about is predicting that I will throw 5 heads on a row. You would have to throw the coin many times to see this pattern. But this pattern has exactly the same statistical likelihood of appearing than any other pattern that you predict.

Try it (if you are really bored), the likelihood of throwing HHHHH is exactly the same of throwing HTHTH (or any other pattern for that matter). Its confusing, but its right. Humans just struggle to get their heads round the concept.

 

[Vell]

Ah, now this is two seperate problems you are posing Chesnor. The first one, where you ataed you are just as likely to throw 10 heads as you are to throw 5 heads and 5 tails - Yaruar is right, it's NOT just as likely.

You only have 1 combination of throws which will give you 10 heads -
HHHHHHHHHH

However, to get 5 heads and 5 tails, you can have:
HHHHHTTTTT or
HHHHTHTTTT or
HHHTHHTTTT or
HHTHHHTTTT or......

In fact, there are 26 possible combinations that will give you 5H and 5T.

So the likelihood of getting 5Hand 5T is 26 times that of getting 10H.
And there are a total of 176 possible permutations for a toin coss 10 times (not counting the very slim chance that it will land on its side)
In fact, the distribution of probabilities for this is:
10H - 1/176
9H 1T - 10/176
8H 2T - 17/176
7H 3T - 22/176
6H 4T - 25/176
5H 5T - 26/176
4H 6T - 25/176
3H 7T - 22/176
2H 8T - 17/176
1H 9T - 10/176
10T - 1/176


However, you statement

The probability to throw a given state is always the same, 50%

is also true. Ah the wonders of mathematics......


And now you say...

What you are talking about is predicting that I will throw 5 heads on a row. You would have to throw the coin many times to see this pattern. But this pattern has exactly the same statistical likelihood of appearing than any other pattern that you predict.

Isn't this exactly what we are talking about? Predicting the statistical likelihood of the people in the group getting an equal share of the loot, and then comparing that to what people actually got, to make a decision on whether autoloot is actually random or not.

Try it (if you are really bored), the likelihood of throwing HHHHH is exactly the same of throwing HTHTH (or any other pattern for that matter). Its confusing, but its right. Humans just struggle to get their heads round the concept.

This is also true - however, the likelihood of getting 5 heads is not equal to the likelihood of getting 3H and 2T.....

And when we're talking about getting equal shares of the loot, we don't really mind if person A gets loot 3 times, then person B gets it three times, as long as we both get the same ammount. SO in our Autoloot example - order does not matter. In your HTHTH example - order does matter.

Anyone else want to join in our lovely maths conversation?

 

[Trahg]

Chesnor you aren't beeing exact when you spoke about this, that's probably why yaruar reacted.
If we take your comparision with HHHHH and HTHTH... and replace them with HTHTH and HHHTT... sure, it's an equal chance to get one of those sequences, but they are only different in the sequence, they are statistically the same result... (statistics isn't about individuality, it's about general result (if i didn't make sense in that sentence it's because I'm not speaking english as primary language ))

Average: More likely to turn out more even divided among the possible outcomes
Step-By-Step: Exactly even chance to turn up in any possible outcome

What I'm saying here isn't that each fight I will hit exactly 80% (just making up a number) of my swings... I'm saying that it's more likely to get near that amount... and in all my fights i would have an average hitrate near 80%... but, since probability isn't rock solid, a fight will sooner or later result in one without a single hit...

Sumary (based on that it is an ideal coin ):
Probability that next flip will be heads: Exact 50%
Probability that all flips will be head: Logaritmically descending from 50%

(if i left something out it was prolly because i got tired of writing )

 

[Anu]

Whoooooooooo! GEEK ALERT Whoooooooooo!
Whoooooooooo! GEEK ALERT Whoooooooooo!
Whoooooooooo! GEEK ALERT Whoooooooooo!

 

[old.yaruar]

Originally posted by old.chesnor
Yaruar is the human I was referring to obviously. When I flip a coin, it can land in 1 of 2 states. Heads or tails. There is a 50% chance of each state. The coin has no memory, and does not know what state is has previously landed in. If you throw 5million heads in a row, you chance to throw a heads next is still 50%. The probability to throw a given state is always the same, 50%

I refer you to my earstwhile collegue vell...

Any individual combination in a specific order has the same probability, but any total number of throws in any order is a different probability.
HHHHHTTTTT
HTHTHTHTHT
are both as likely as each other, but are essentially 5h 5t.
Essentially you need to look at the different combinations before you can work out the total spread of results. There is always a skew to the results which is why usually you end up with results which show close to 50 50 with little deviation...

Unfortunately I am actually one of the humans who spent far too long being forced to learn this stuff to qualify me to be a beancounter.....

The one that I never got my head round personally was the gameshow with 3 boxes probability puzzle which someone had to sit me down and show me the proof before I got it....

 

[Myshra]

bad luck, you fought with the accountant and lost!

 

[Vell]

Originally posted by Anu
Whoooooooooo! GEEK ALERT Whoooooooooo!
Whoooooooooo! GEEK ALERT Whoooooooooo!
Whoooooooooo! GEEK ALERT Whoooooooooo!

I'm not a geek anymore. I used to be. At school, and at uni where I did a maths degree I was, but I'm a normal human now. So no geekiness from me anymore.

 

[old.yaruar]

Originally posted by Myshra
bad luck, you fought with the accountant and lost!

Actually I'm currently dole scum ;-)
Although I did recently look at statistician work, but my god, it's boring.....

 

[Vell]

/em growls

 

[old.tRoG]

it is more likely that you will get 4head and 6 tails, because the heads side weighs slightly more

 

[old.flit]

heads or tails?

try this for me, a friend showed it to me and proved himself right so.....

place the coin heads up each time on your hand before you flip it.

flip the coin 15 times and you should get more tails than heads (~9-10 tails, ~5-6 heads)

now, we worked it out as being a 2/3 chance of the coin landing on the other side to which it was flipped. and it works both ways too.

 

[old.dittytwo]

STATISTICS

90% of statistics are mad up on the spot

that’s with a 12% error ratio


and going back to the originally post loot I don't get any when in a group but get it all when I solo

seriously I must be really unlucky

I can not remember the amount of time I have been in a full group with 7 of the members saying MY good I am full I need to go and sell were I have the 4 items I brought with me when I joined the group (go figure that one out)


Charcy the unlucky minstrel

 

[old.Alliandre]

So what items have you got? I was thinking off making a crazy Mystic on the PvP server when it came out. Want to see how good I'd be

 

[old.xetius]

well, ive been playing long enough to get a good enough spread, and know about randomness and streaking, but I have yet to be the one to say that my bags are full and I need to sell up (Even though I have 2 bags of venom, leaving only 3). Usually when the first person wants to sell, I have 1-2 bags of stuff

 

[the_smurflord]

In my experience it is random, but heavily weighted towards the higher level members of any group. When I'm the lowest member of a group, I tend to get very little. When I'm the highest, I tend to get loads.
Overall this is fair, as the higher levels will be doing more to the mob than the lower levels, in the main.

 

[- Pathfinder -]

People with alot of stuff in their inventory are less likely to receive loot, but you end up with an uneven distribution as often as not anyway

 

[Blood]

daoc's random number generator is a pseudo random number generator, and it doesnt work very well, very often have streaks, ask a crafter, and he will tell you that you very often gain skills twice in a row, then noone, noone, noone etc.

and take a look at this screenshot (notice staff in question is 5 points higher than my skill, and wasnt a problem to make at all, untill i got this streak)

 

[Vell]

Re: heads or tails?

Originally posted by old.flit
try this for me, a friend showed it to me and proved himself right so.....

place the coin heads up each time on your hand before you flip it.

flip the coin 15 times and you should get more tails than heads (~9-10 tails, ~5-6 heads)

now, we worked it out as being a 2/3 chance of the coin landing on the other side to which it was flipped. and it works both ways too.

I just carried out your test and got more heads than tails, thereby disproving your friends theorem. Please note - your friend did not prove himself right - he merely showed that it worked on that occasion. In order to prove it, you have to show that it works every single time it is carried out. To disprove, you only need to show that it doesn't work once.

 

[Lambsbreed]

Originally posted by old.Blood|Prydwen
daoc's random number generator is a pseudo random number generator, and it doesnt work very well, very often have streaks, ask a crafter, and he will tell you that you very often gain skills twice in a row, then noone, noone, noone etc.

and take a look at this screenshot (notice staff in question is 5 points higher than my skill, and wasnt a problem to make at all, untill i got this streak)

some picture..

If i think right, their random numbergenerator are made out of different states in the game. Such as amount of gold divided on amount of people actually playing etc

prolly the most hard thing to do is making a perfect random numbergenerator, but when it comes to mmorpg's it's a tad easier when other players allways varies what they do, what they have and how much gold they have etc etc etc..

 

[old.yaruar]

Re: Re: heads or tails?

Originally posted by Vell

I just carried out your test and got more heads than tails, thereby disproving your friends theorem. Please note - your friend did not prove himself right - he merely showed that it worked on that occasion. In order to prove it, you have to show that it works every single time it is carried out. To disprove, you only need to show that it doesn't work once.

There is no such thing as proof, just conjecture and refutation.........

;-)

 

[old.job]

I love probability, statistically there is no difference in chance of last weeks lottery numbers coming up again that any other set of random numbers... but would you pick last weeks numbers???