LuckyBancho data shows that FF14 is in decline

[Noumenon]

Not all players make their character achievements public.

You are correct. There is therefore a sample bias in the LB data. The rate of decline however seems to be similar to the overall decline (around 20%) according to LB's complete count (anyone who has a Lodestone change in general). So the total decline count trend is similar for the latest quarter.

The question is: do the proportions of the colors of the bar generalize to the overall population? I'm inclined to say absolutely not. Players who opt in to make the achievements public are of course more dedicated than the average player. So this data shows that the more dedicated players are more likely to "terminate" in Endwalker.

However, do the change in proportions of the colors of the bar over time generalize to the overall population? I don't see much reason to say no. For this to not be the case, we would need to violate the assumption that there is a sharp shift in the "innate" motivation or playstyle of players that is going in very different directions with a large magnitude between the two groups. While this could certainly be the case, I think it would take an extremely massive widening of the gulf between the two groups for us to not conclude that the number of terminations among the wider population is also increasing.

So what? It's a 10 year old game, which is much longer than most games remain relevant.

But does it have to be that way? No.

I’ve given you three examples

Already discussed the pandemic long ago in page 2. I've also already remarked that the WoW influx was a fluke. Your criticisms are already addressed in other comments and I am not going to repeat myself.

It's also quite obvious you didn't bother reading the OP anyways. The primary point of the OP post was about the change in distribution of "terminated" players over time. This has nothing to do with whether EW was the conclusion of a story arc. You are just whiplashing at the obvious story of player counts going down when I was exploring the details of the graph: the different proportions of types of players and how they're changing across expansions.

You've shown zero understanding of data scientific concepts, which is why I'm not bothering to refute you. You don't write like a statistician. Certainly my remark on indicator functions went completely over your head (and it's quite an elementary mathematical concept too). I'm not going to bother replying anymore unless you show aptitude.