How to analyze loot

[kalanen]

Falk,

Is a field trip completely out of the question? I could freshen up your approach a little. There's bunch of things that cannot be seen in any recordable data, like:

- faster loot variance + shorter cycle length in crowded areas vs. long, static phases in silent ones. (hp cycle based system, scalable in any level; not random, not timer)

- reflections of overall loot level in any area (universal cycle; not random)

- "pre echoes" of peaks with same mob in another area (mob cycle; not random, against predetermined loot classes)

- center point of dmg*n pecs based loot (loot level with very little spread)

...to mention few.

I think you would spot these easily.

[falkao]

Originally Posted by Immortal

You refer to wait times between globals etc. You're assuming that it is one time-dependent function (we know it's not time dependent, but hp dependent, but we assume that the average hp_reduced is directly proportionate to time. I don't have access to the logged data yet so I haven't looked at this myself. However it would play along the lines of deconvolution of the global signal received {f(x),f(x+a)...f(x+a.n)}

What I had in mind was the following.

As we know loot has a distribution. Therefore, a value above 50 PED has a certain probability to observe it. If I assume that this prop. is 2.5%, I would expect a global in every 50th-60th kill. So the relation to time is given indirectly by the number of kills between to globals. We have observed avatars globalling quite close in time. There was for example one getting 6 globals in 1 hour on Ambus (or a mob with HP > 1000), if I remember right.
Unfortunately, the recorded global data doesn't have any info about the number of kills between globals. However, to get 6 globals in one hour with a rate of 2.5 globals out of 100 kills, he would have to kill about 240 creatures. So he needs to kill at least one creature in 15 secs. To kill an Ambu, more then 15 secs are needed and therfore we concluded that there must be global grouping.

All is still a random process, but instead of having an overall global rate of 2.5% there seem to be two states. One in which the global rate is nearly 0 and one in which it is much higher then 2.5% (say 25%). So for example in 20 kills you have a rate of 25% with 4 globals and in 180 you'll get nothing. So overall it is still 2.5%.

Instead of using the number of globals out of a number of kills, knowing the global rate, we can use the observed waittime between globals instead. This has the advantage that we can use global data.

So instead of using a distribution for the number of globals, we have now a distribution of waiting times. This is a time related function but stated as a distribution.

If the underlying process is a poisson process (i.e. the number of globals per time unit has a poisson distribution with rate r), then the distribution of waiting times has an exp distribution with mean 1/r. This is what I described some posts back.

Furthermore, I did check the states and analyzed the loot ranks, but have still to complete it.

Atm we don’t have an explanation why this happens. So it might be related to the area, or the ava itself or both. With your model you would try to model something like that in relation to time. That would come next, but first we have to identify the factors.

[falkao]

Originally Posted by kalanen

Falk,

Is a field trip completely out of the question? I could freshen up your approach a little. There's bunch of things that cannot be seen in any recordable data, like:

....

Absolutely not, and I would appreciate it to do one with you or a group of persons. Should we set up something for this weekend?

[kalanen]

Originally Posted by falkao

Absolutely not, and I would appreciate it to do one with you or a group of persons. Should we set up something for this weekend?

There's no silent enough areas for some demos during weekends, but spreaded group of hunters with variable killspeed, eco and phase of personal performance in looting could bring this discussion forwards. Maybe share some hints too.

Atleast I'm up to it. Sat night?

P.S. There was a typo in my original post. "I" can fresh up shit but "it" (meaning fied trip) could.

[Doer]

Originally Posted by kalanen

There's no silent enough areas for some demos during weekends, but spreaded group of hunters with variable killspeed, eco and phase of personal performance in looting could bring this discussion forwards. Maybe share some hints too.

Atleast I'm up to it. Sat night?

P.S. There was a typo in my original post. "I" can fresh up shit but "it" (meaning fied trip) could.

I want to come.

[falkao]

Originally Posted by kalanen

Atleast I'm up to it. Sat night?

sounds good to me. What time? And should we fix here what to hunt or should it stay secret, to avoid a bias

Originally Posted by Doer

I want to come.

great. Maybe you can post your time as well.

[Jimmy B]

Hmm Sat night. Are you sure you guys don't want to join England WoF strike support instead?

[kalanen]

Originally Posted by falkao

sounds good to me. What time? And should we fix here what to hunt or should it stay secret, to avoid a bias

Me thinks it's better just to gather up and try to find silent spot with somewhat hunted mobs.

I'm starting at ~7pm MA time.

[falkao]

Originally Posted by kalanen

Me thinks it's better just to gather up and try to find silent spot with somewhat hunted mobs.

I'm starting at ~7pm MA time.

sry Kal, was not available for the proposed time slot. So I did some testing on Hogglos, but we should still try to do a field test together.

Killed a total of 127 Hoggs in 3 runs. I've got only 2 types of loot

1) Normal loot, i.e. loot < 3*HP/1000 PED, or
2) a global

Out of the 127 kills I've got 9 globals. Most one global in every 10th to 15th kill. I've got them while other where also globaling or not. So its hard to say what pattern it did follow. The last 32 kills where bad. No global in them but a lot of globals for others hunting nearby. So let's see if I can recover from this tonight.

Now a question: I do not hunt Hogglos very often, but if I remember right they dropped also pedders. Atm it seems that there is only low loot and globals. Anybody else with a similar observation?

[Doer]

Originally Posted by falkao

sry Kal, was not available for the proposed time slot. So I did some testing on Hogglos, but we should still try to do a field test together.

I thought that was you i saw globaling.

Now a question: I do not hunt Hogglos very often, but if I remember right they dropped also pedders. Atm it seems that there is only low loot and globals. Anybody else with a similar observation?

I personally have hunted Hogglos once since they dropped PEDs (like, shiny gold things), but i've that time (about two months ago) we had no globals and one ~40 PED loot that probably would have been a global on any sane loot day for us (another team was pulling in big ESIs and globals every few minutes during that time). So, yeah, they drop non-globals if you're in a bad loot period (whatever that may be).