Game is turning into a glass cannon only builds

"

AdRonZh3Ro wrote:

Aside from the conversation being derailed(again):

Taking steam charts as clear evidence of anything is simply utilizing really, really bad information because multiple sources state that steam accounts for 60% of the player base, fluctuation of course. Acknowledging you're ignoring 40% of viable data into any serious study would automatically make that study instead a joke. That applies here. Considering how it's not possible to calculate how many players even know Maven exists, least of all having fought her, it's really laughable how any accurate conclusion can be taken from that data, at all. Data treatment to show patterns must come from 100% of available data, or data treatment becomes data manipulation.

Huge spikes and equally huge dips are built into the financial model of the game's release cycle. PoE isn't about retention day to day or week to week, but league to league. As long as each league release sees a huge spike and that spike reflects financial support, the dips could be literally zero and it doesn't really matter. Conversely, there could be no dip at all but if financial contributions themselves became near zero past the first few weeks, then, technically, it's the exact same situation.

"

DarthSki44 wrote:

At this point I'm speechless at the obtuseness. It's at an epic level.

Even if we remove all the data and simply go from an emotional, unscientific, purely speculative guess work standpoint, and I ask a simple question:

"

AdRonZh3Ro wrote:

Acknowledging you're ignoring 40% of viable data into any serious study would automatically make that study instead a joke

Data treatment to show patterns must come from 100% of available data, or data treatment becomes data manipulation

"

feike wrote:

that 20% came from where? I would guess from your rear, and you want to talk unscientific?

"

DarthSki44 wrote:

"

feike wrote:

that 20% came from where? I would guess from your rear, and you want to talk unscientific?

It's actually less than that, the achievement for opening a map and entering is 16%. This isn't a trust me bro situation. You can verify. I was just rounding.

"

DarthSki44 wrote:

Well at least I'm basing my conclusions on actual data

"

DarthSki44 wrote:

over 80% not making it to maps seems bad :/ though

"

feike wrote:

Actually, you dont need 100% of the data to reach a conclusion, its the concept of sample and most statistics are made with small groups and extending conclusions to larger groups. Unless theres reason to think the average steam player and the average client player differ significantly, you CAN take just the steam players and extend the conclusion to every player

"

DarthSki44 wrote:

I'm not aware of some data that indicates players from the PoE launcher exhibit behaviors that are different than those of the Steam launcher? I'd actually be interested in that if you can back that up.

"

AdRonZh3Ro wrote:

"

feike wrote:

Actually, you dont need 100% of the data to reach a conclusion, its the concept of sample and most statistics are made with small groups and extending conclusions to larger groups. Unless theres reason to think the average steam player and the average client player differ significantly, you CAN take just the steam players and extend the conclusion to every player

"

DarthSki44 wrote:

I'm not aware of some data that indicates players from the PoE launcher exhibit behaviors that are different than those of the Steam launcher? I'd actually be interested in that if you can back that up.

There's no need to. You can't attempt to assert what behavior is exhibited if data to corroborate what that behavior is doesn't exists. The only affirmation you can possible make is confirmation bias.

The only fact is that there's inaccessible and/or unverifiable data. It's impossible to take any accurate affirmations from that. You can conclude what you want, but it's guess work at best.

"

DarthSki44 wrote:

I mean sampling is absolutely valid methodology in statistical analysis. It's done all the time...in fact it's almost entirely how you model. Not sure where you are going with this in general.

"

AdRonZh3Ro wrote:

"

DarthSki44 wrote:

I mean sampling is absolutely valid methodology in statistical analysis. It's done all the time...in fact it's almost entirely how you model. Not sure where you are going with this in general.

Oh, sure, when we are talking about material samples, but when we talk about human sampling and psychology, that has so much bias involved it might as well be useless.

"

DarthSki44 wrote:

I mean sampling is absolutely valid methodology in statistical analysis. It's done all the time...in fact it's almost entirely how you model. Not sure where you are going with this in general.

"

AdRonZh3Ro wrote:

Ever heard of people swallowing 8 spiders a year while they sleep on average? Yeah. LoL.

"

DarthSki44 wrote:

Uhhh...what about polls, surveys, census, analysis on expected behaviors, and much more.

Again this happens all the time. It's truly the only reasonable way to forecast. 100% is often impossible, and frankly not really needed.

Edit: But damn we are super sidetracked now :(

"

Nomancs wrote:

"

AdRonZh3Ro wrote:

Ever heard of people swallowing 8 spiders a year while they sleep on average? Yeah. LoL.

My neighbor and his dog have 3 legs on average ;) Man, this is a real problem!