Question: How do I prove there is or is not a personal loot/expenses pool/tracker?

[-Archangel-]

I believe the loot is created at point of hitting the loot button. It is not sitting in the pockets of the mob.

being created at the moment it is looted and the ammount not being related on how it was killed are two different things, how you kill it seems to be a variable that determines loot.

 

[Fifth]

I believe the loot is created at point of hitting the loot button. It is not sitting in the pockets of the mob.

Think about it from a coding point of view - you'd need a huge database just to hold all the spawned loot for each and every mob in the entire universe if it was sitting in their pockets. If its generated at the point of pushing the loot button, then you dont need that whole database structure which would be pointless expense to build and maintain.

The fact that the particular server you were on was flush with peds doesnt mean the particular mobs in that spawn had the loot in their pockets already.

Yes yes, the "loot in mob" and the "push the button" theories. We don't know how it works but let's assume for a moment that there's no fundamental difference between fruits and globals.
Suppose HoFs/ATHs are coming from a "special fund" then most mobs can safely keep their meager "default loots", all i have to do is spread out a few hundred thousand peds over 50 mobs at the time. Maybe it's not few hundred K, depends on the global economy as MA has said many times, could be million peds randomly spread out at better times (have to take into account many will not be found before they time out). This can be maintained and it's not too hard to code.

Soooo... what we are looking at? It's kinda reversed minesweeper game - most squares are empty, some give hints "something big nearby" and when you carelessly step on the wrong one - BUM! well, in our case it's the right one... Hmm, seems i wandered back to a certain old loot theory that was popular before "personal loot pool" became everybody's favourite lol

Where the efficiency fits in this picture? With this theory it seems to be "outside", here it's simply how long can you manage before you step on the right square and blow up, and what's your losses when you make it there.

Hehe, i'm not sure i like or even believe this theory myself... but who knows, it might as well work this way...

being created at the moment it is looted and the ammount not being related on how it was killed are two different things, how you kill it seems to be a variable that determines loot.

Sounds good, as long as you're a hunter. Miners will have hard time figuring out how to "kill it right" thou Amped or unamped? Double bombing or not? Yes hunters have lot of variables they can change but miners have very few things they can do differently while dropping this one single bomb that brings them ATH.

 

[Jimmy B]

Think about it from a coding point of view - you'd need a huge database just to hold all the spawned loot for each and every mob in the entire universe if it was sitting in their pockets. If its generated at the point of pushing the loot button, then you dont need that whole database structure which would be pointless expense to build and maintain.

I have no strong opinion about whether loot is in the mob or not, but I don't agree with this logic. Presumably each mob has an Id, and fields that identify the mob type and maturity, whether stored in memory or db. Some sort of loot multiplier field or even a loot multiplier that is generated directly from the Id seems perfectly plausible to me. Then this multiplier is taken in conjunction with the kill information (how much HP was killed, etc) to determine the tt loot. Then the mob's droplist table is used to determine the individual items that make up the loot.

 

[Joat]

Mobs have a individual ID, but my guess is a mob is also a part of a cluster.
Just look at a LA, it has 6 clusters, from where a certain amount of
mobs are spawned.
So I doubt there are a setting per individual mob, rather per cluster.
My guess is that some type of multipliers influence these clusters, but not
all kind of multipliers.

 

[Uvas]

I have no strong opinion about whether loot is in the mob or not, but I don't agree with this logic. Presumably each mob has an Id, and fields that identify the mob type and maturity, whether stored in memory or db. Some sort of loot multiplier field or even a loot multiplier that is generated directly from the Id seems perfectly plausible to me. Then this multiplier is taken in conjunction with the kill information (how much HP was killed, etc) to determine the tt loot. Then the mob's droplist table is used to determine the individual items that make up the loot.

And if you think about it, this can all happen when you go to loot a mob, i.e., push the button. The type and maturity of the mob you just killed must for sure be part of the loot equation.

 

[Infinity01]

I think the loot droplist is decided at the time of looting....at least that has been my observation at cyrene..
There the scoutbot (iirc) drops low grade powersupply in a cyclical pattern... very short window of ~5 mins in every 30 to 60min. Initially, to kill more during that short window I just killed some and didnt loot them immediately, later (after the short window period) i found out that they didn't carry the low grade power supply. So, I think that was a good enough for me to conclude that the droplist is created at the point of looting.

 

[mrproper]

nope, i cant agree with that. thats sounds like a bug, assuming your using the correct team setting, loot should share stackables. (or maybe you are using one of the daft team settings). Bring up teaming is interesting point as i have always felt teaming is a major obstacle to person loot.

Assuming I am using correct team settings and take video footage next time, would you be inclined to accept that I am not a bloody freaking liar?

I can't believe you didn't notice this until now.

 

[mrproper]

I think I found two null hypothesis, goes like this:
- long term log, more than 100 loots per sample, more than 50 samples.
- A/B comparison with the same setup, only behavior changes
- Test A attempts to obtain the lowest loots only
- Test B attempts to obtain the highest loots only
- Test is valid if combined A+B overall return is 84-94% (90%+/-5%) (log curve adjusted)
- Confirmation of new hypothesis against personal loot pool if A return is below 70% (90%-15%)
- Confirmation of new hypothesis against personal loot pool if B return is above 100% (90%+15%)
- Confirmation of new hypothesis against personal loot pool if more than 75% out of 4+ independent players confirm the previous two points.

What do you guys think?

In case you are wondering how to obtain the lowest or highest loots (at least try to):
- loot 10 times before starting the log, decide to continue the log depending on result
- team up with someone and watch his loot in team chat, act accordingly
- watch global chat for global activity, act accordingly
- ask people about their current loot return, act accordingly
- monitor your in-test return and increase or decrease looting speed

 

[-Archangel-]

What do you guys think?

I don't see what you are trying to find out in the above.

Meaning what are the null hypothesis ?

Without it any further comment from me seems pointless

 

[Magix]

Sorry if this was already suggested:

Step 1: get two blueprints. It shouldn't matter which two, skill levels, etc.

Step 2: On each bp, do 1k clicks consecutively, tracking gains/losses on full quantity (for condition I would think at least 10k clicks would be necessary.) A breakdown per 100 clicks is fine, but not necessary. Calculate average profit/loss per click or per 1k, doesn't matter.

Step 3: Do another 1k clicks on each bp, but this time alternate them in sets of 100 (or 50, or 10, or 1, it shouldn't matter.)

If there is a personal loot pool:
1. The alternating sets will not be equal to the baseline sets. (Although I have seen series of crafts that do follow a 2 losses/1 win pattern, so three bp's can be used in this case.)
2. The gains in alternating sets should follow losses, and vice versa. The % chance that a gain run follows a loss run would be much higher than 50%.
3. The gains in alternating sets should make up for losses in the previous sets, and vice versa, within about 5-10%.

If you wish to control for COS, you can use qr bp's from the same level and book, but if there is a personal loot pool, QR, skill level, etc. shouldn't matter.

If there is no personal lootpool:
1. The alternating sets should be equal to the baseline sets.
2. There will be no pattern in the gains/losses. The % chance of a gain run following a loss run is very close to 50%, within +/- 5%.
3. The amounts of gains/losses will not equal each other, the variance will be +/- 10%+.

In this case, I would advise to control for cost per click, probably the cost per click between the different bp's should be less than 10 pec.

If you are getting a lot of negative runs (or only negative runs, actually), break even point can be recalibrated to 90-95% and profit/loss calculated from there.