[Research] Map increased item quantity (IIQ% ) and its role in map sustainability

"
ScrotieMcB did an interesting thread a while back about streakiness in map drops (PSA: Understanding Map Drop Streaks) which somewhat agrees with the data you've provided here. According to their model (which, disclaimer, was only a model) quantity has little effect on whether a map "breaks even" and drops a map to replace itself. Instead, the gains go toward getting multiple drops per map, but even then, there's only a (calculated, not empirical) 20% difference or so between a 20% quant map and a 100% quant map. Given your small sample size compared to the likely drop rate of maps, it's not surprising that you didn't detect much of a difference.

Their model didn't take Pack Size into account, but assuming Pack Size acts as a separate More multiplier on map drops, it's not too hard to do the math. For example, a map with 50% IIQ and 30% Pack Size should drop about as many maps as one with 100% IIQ (1.5*1.3=2). A map with 100% IIQ and 15% Pack Size should drop as many as a map with 130% IIQ (2*1.15=2.3), or, alternately, it should drop about as much as a map with 75% IIQ and 30% Pack Size (1.75*1.3=2.3). In general, since mods that give Pack Size also give IIQ, it makes sense to prioritize Pack Size.


However, neither of these analyses take Bloodlines or Nemesis mods into account. I'd be very curious to see what effect those mods have, since they both increase drops (likely as a separate More multiplier) AND they increase the chance for a +1 drop or +2 drop. Since each +1 drop is worth 3 +0 drops for map progression purposes, it might make sense to roll for those mods.


Good post about possible factors.

Another one I would add is WHICH tier map is being run. I don't think every tier 9 map is the same in terms of number of monsters killed and such.

Hopefully aggregate would account for all of these outlier factors however, I do think a sample size of 400 or so should take a strong note of which maps, Even if we assume even distribution of each Tier, that is only roughly 100 or so per "tier", split that tier into which "map name" and now you have even smaller and smaller samples.



I could be 100% wrong about this, but I do think that their are too many factors to consider that an aggregate can account for (at least in terms of a relationship). The data would be too noisy.

"
Chris wrote:
This thread has been linked to us a lot of times today and we're worried that it might be spreading misinformation.

You said "Not all map drops were being included in the research data", and then conclude that higher IIQ doesn't help much. Can you post your raw sample data that includes all map drops? IIQ is linear on items and maps. Note that the map boss has a chance of an extra map, which can bias the data if you don't take it into account.

Our concern is that people will read the thread and think that IIQ doesn't help (when it significantly does, as your data should show when you look at all the drops) and then receive fewer maps, further increasing the complaints about drop rate.


I sincerely hope I do not sound like an asshole.

If you are worried about misinformation spreading, why don't you release the correct information?

<3 Free Tibet <3
"
Opinionated wrote:
"
Chris wrote:
This thread has been linked to us a lot of times today and we're worried that it might be spreading misinformation.

You said "Not all map drops were being included in the research data", and then conclude that higher IIQ doesn't help much. Can you post your raw sample data that includes all map drops? IIQ is linear on items and maps. Note that the map boss has a chance of an extra map, which can bias the data if you don't take it into account.

Our concern is that people will read the thread and think that IIQ doesn't help (when it significantly does, as your data should show when you look at all the drops) and then receive fewer maps, further increasing the complaints about drop rate.


I sincerely hope I do not sound like an asshole.

If you are worried about misinformation spreading, why don't you release the correct information?



GGG will never release map drop stats because people will freak out when they dont hit the margin similar to how they never released 6 link odds despite it obviously being around 800.
IGN: Arlianth
Check out my LA build: 1782214
Last edited by Nephalim on Feb 16, 2016, 10:04:58 PM
"
Nephalim wrote:

GGG will never release map drop stats because people will freak out when they dont hit the margin similar to how they never released 6 link odds despite it obviously being around 800.


This isn't a map drop rate thread though. It is a map tier sustainability cost effectiveness thread.



(Tin foil hat personal opinion)
Spoiler
I would guess IIQ has zero influence over which map tier drops, only how often any map drops. When people find out, they will adopt the "alch and go" method for rolling maps. Suddenly, a HUGE currency sink vanishes.
<3 Free Tibet <3
"
Chris wrote:
This thread has been linked to us a lot of times today and we're worried that it might be spreading misinformation.

You said "Not all map drops were being included in the research data", and then conclude that higher IIQ doesn't help much. Can you post your raw sample data that includes all map drops? IIQ is linear on items and maps. Note that the map boss has a chance of an extra map, which can bias the data if you don't take it into account.

Our concern is that people will read the thread and think that IIQ doesn't help (when it significantly does, as your data should show when you look at all the drops) and then receive fewer maps, further increasing the complaints about drop rate.


Not being a statistician, I really don`t care about how iiq increases the chance for low tier maps. If a large amount of map runs shows that iiq in only a small amount increases drop rates for higher tier maps, well that is a problem.

But I am somewhat biased as I find the map drop rate disgusting.

Also, if there was a significant effect, you would not have to make this post as everyone would say it was contrary to their experience.
''It was nice but not really great'' in memory of rondanashu 13.4.2016
Last edited by rondanashu on Feb 16, 2016, 10:36:28 PM
"
Weißenberg wrote:
- high-tier maps (T12+) were excluded from the research due to insufficient data consisting of low IIQ% map runs
- Not all map drops were being included in the research data.


In order for this to have any foundation at all, you need to include all maps that drop, regardless of tier, otherwise all you're doing here is fueling the tinfoil hat parade.

You could just as feasibly run an 100%+ IIQ map with the way you're currently isolating data and record 0 map drops, when in actuality you dropped 10 maps that were T6 or lower. But then the next map, for argument's sake lets say it was mirrored so literally everything else was identical, you could drop an additional 10 maps but they were all T10 or higher, so you record them.

This does nothing other than bias an already jaded community, in part because it completely ignores all the various other variables that go into map sustain. (Pack size, Zana mods, buying maps, selling maps, nevermind the fact that you're ignoring T12+ to begin with, which is PRECISELY where IIQ starts to play a significantly heavier role in pool sustain.
Jul 27, 2011 - Sept 30, 2018.
I would like a record with corruption included seems to be helping me a lot in staying between tier9-12.
"
conall88 wrote:
good idea, but sadly your sample size is statistically insignificant, and without this significance, any conclusions you draw are gonna be pretty inaccurate.


you should probably take a college stats course


a sample of 400 is pretty good for margin of error

ht tps://onlinecourses.science.psu.edu/stat100/node/17
"
Chris wrote:
This thread has been linked to us a lot of times today and we're worried that it might be spreading misinformation.

Why do you care? GGG had NEVER tried to offer correct information, isn't the whole point about us getting the wrong information, being part of GGG policy? Like that famous case of how scour orbs interact with master crafting, which was intentionally kept incorrect and deceived thousands of players for a year because GGG "didn't want to correct misinformation".

I can say outright that you are wrong. We can spread as much misinformation as we want. Your job is to keep silent and ban anyone trying to spread the truth, that is what your boss wants. (Edit: I see that it was Chris himself I was replying to. It seems Chris was being schizophrenic and didn't remember what his views are about how he approved of player misinformation a few months back.)

Disagree? Then release the data officially. Because telling us "you are wrong" when you want us to speculate and guess without any facts, is meaningless. That's what happens when you keep things hidden, we come up with conclusions you don't like. If you don't enjoy being misrepresented then offer us something in return.
Last edited by VallenChaosValiant on Feb 16, 2016, 11:42:12 PM
"
"
Chris wrote:

Your job is to keep silent and ban anyone trying to spread the truth, that is what your boss wants.


"
"
Chris wrote:
your boss

Think about that.

Report Forum Post

Report Account:

Report Type

Additional Info