We lost 70% of our player base half a year in

[Rolder50]

Horsebiscuits.

"We interviewed 10% and that means that the other 90% think the same way"

No. Not even close.

If the 10% was randomized then that is more then sufficient information to draw conclusions from. It's not like polls survey literally the entire population...

[AlyaNimfa]

If the 10% was randomized then that is more then sufficient information to draw conclusions from. It's not like polls survey literally the entire population...

Multiplatform game, not two platforms. Also a minority population on Steam.

I'm sad to see these hysteric threads infecting here after seeing it elsewhere, where their games are still around. People need to touch grass.

[Komarimono]

For the longest time, Steam didn't even know when I ran the game until this recent change. The fact I'm forced to now is disappointing.

[VelKallor]

It's not like polls survey literally the entire population...

"We surveyed 300 people out of a population of 10 million

Every time I see these polls that say "85% of the population agree" i just sit and laugh myself hoarse..see. NOT ONCE in 50 years have I or anyone I know ever been polled...soo..they dont speak for me as I was never asked.

There are lies.

Damned lies.

And statistics.

[anhaato]

"We surveyed 300 people out of a population of 10 million

Every time I see these polls that say "85% of the population agree" i just sit and laugh myself hoarse..see. NOT ONCE in 50 years have I or anyone I know ever been polled...soo..they dont speak for me as I was never asked.

There are lies.

Damned lies.

And statistics.

Just out of curiosity, what is an acceptable sample size to you?

[Toutatis]

So the Op takes 6 months of the Steam numbers which are incomplete to begin with, choses the release of an expansion in December as the starting point to draw a conclusion to the effect that the “game is dying”? Lol.

It is normal to have a huge peak when something major and new is released. Many go through the story and that’s it. If you look further back , the (incomplete) Steam numbers show an overall increase over the years. It’s pretty hard to draw any conclusions from those numbers as the steam version also had issues. It is doubtful SE would be adding new worlds if the game was dying...

Just out of curiosity, what is an acceptable sample size to you?

It’s not just the size of the sample that counts. While I am sure that there has been a peak followed by a decrease of the number of players across all platforms/purchase sites in the past six months due to the release of the expansion in December, we don’t know how representative the Steam sample is of the overall game population (though the poster above you clearly does not understand anything to polling methodology, including randomization and weighting, and statistics).

[VelKallor]

Just out of curiosity, what is an acceptable sample size to you?

Realistically speaking?

If the survey is to include a population of over 100,000, then an acceptable sample would be at least ONE HALF. Telling me that I am in the group that agrees with some idiotic claim after a sample of 300 people in a population of 300,000 ( the local area I am in ) is absolutely bizarre.

"We sampled a group of 300 people and have concluded that 85% of a population of 300,000 agree with it" is to put it bluntly, idiotic.

Ask 5000 people and you get 5000 DIFFERENT answers.

its like these political polls you see that apparently agree with some asinine government proposal....leading me to ask who did they ask, the members of that particular political party? We never hear who they talked to , or their political affiliation either.

Then again we never know if the questions themselves were loaded to get a predetermined result.

Throw out all the statistics you like, they are all garbage.

Ive never been polled ONCE in over fifty years. No one I know has, no one in my family has been.....so no i dont agree with these "polls"...you want my opinions?

ASK ME.

Most statistics are utter GIGO.

(though the poster above you clearly does not understand anything to polling methodology, including randomization and weighting, and statistics).

Wrong.

Worked for a statistician on several projects, and am very familiar with the entire system. I dont care what "polling methodology" you use, any survey that gets a sample of 10% and then concludes that those results accurately affects the other 90% is...........delusional.

He thought the "science" was a steaming load of fwe too.

[Raoabolic]

Horsebiscuits.

"We interviewed 10% and that means that the other 90% think the same way"

No. Not even close.

Don't worry about the naysayers, they are just pessimists that notice basic patterns. :^)

[DPZ2]

Statistically speaking (and yes, I do have a degree in a field directly related to statistics: biometry) the only thing one may infer from the Steam numbers is that the average concurrent number of Steam players who play FFXIV at any given time has dropped from a high at Expansion Release to the current low at the current moment.

One cannot use statistics to assert that the observed drop is due to any speculative argument of the OP. The conclusion that "you WILL lose the remaining playerbase" based on Steam numbers is ludicrous ...

And, yes, mmo-population utilizes a methodology that is as speculative as all get-out. I use it as a sort of "buzz counter". The less "buzz" a game gets, the fewer people are interested in talking about that game, which correlates, but does not confirm, the amount of players in the game. If you know of anything closer to reality across a wide swath of games (and no, Steam numbers don't count for anything that is cross-platform), bring it up. I still trust their numbers are closer to reality than the OP's speculation of a 70% drop across all platforms for this game.

[Biometry: the application of statistical analysis to biological data. Back in the days when I got my degrees, the general path was to go work for a company doing biomedical research, mostly in the pharmaceutical industry. I spent the decade following graduation working as a Research Fellow for a University's Health Informatics department that was operating a research resource for the National Institutes of Health.]

[Shougun]

"We surveyed 300 people out of a population of 10 million

Every time I see these polls that say "85% of the population agree" i just sit and laugh myself hoarse..see. NOT ONCE in 50 years have I or anyone I know ever been polled...soo..they dont speak for me as I was never asked.

There are lies.

Damned lies.

And statistics.

You do not need that many people to have VERY accurate statistics.



The biggest issue is having biased samples, biased questions (leading to biased results), and of course either really cleverly worded or lying statistics (like you can tie unrelated things together and make them appear strongly related, those fun stories about how you drink coffee and so you're a psycho or whatever lol).

A fun one that isn't lying is good stuff like you could have a 5,000% increase but if you're going from .005 to .25.. It could still be relatively meaningless to people (particularly if you were using smaller % like "50" or "100" lol, then combined with purposeful wording like 50% larger, but forgot to mention there is 30% more air and the tube was already tiny and the box is huge, etc, :3).

Forums for example, if they were a good sample, would very easily represent the game. Of course due to their nature they can't be a perfect sample cause it's all voluntary / away from the game and there are specific biases to why you'd be here then, like to complain because most people don't post to state what they like (data, imo, still VERY useful but you couldn't let the forum, reddit, twitter, etc, blindly lead you).

Steam drop off in and of itself wouldn't help you too much. Like you'd have to see if this drop off happens before, as we know people come and go to games as content is hot / whatever reason (big game comes out, like Elden Ring, etc). If majority players are on steam then the drop would be a bit O_o;; but I assume that's not true (have not bothered to look it up, this is the best form of statistics I tell ya.. lol), so then you'd have to contend with what that means (if that introduces issues, like 'can you pool steam data into all the other data and it doesn't introduce flaws?' perhaps steam does represent the game, maybe it doesn't - I don't think we have the data to really determine it, and even if it did again it could be 'mmo as expected').

If we wanted to determine the health of the game I feel an easier way to do this is just to wait for their financial reports, rather than guess at something that could be business as usual or misleading us given some interesting relationship we're not seeing because we don't have a good full picture.