18.5% drop in players in only 8 weeks

[Striker44]

Wrong. Wrong. Wrong. That is not how you measure growth. Growth is a trend metric, and as such you do not measure it by taking two points in exclusion to the rest of the set. Doing so has a high likelihood of giving you an erroneous result. The only thing you can determine from noting that a more recent point is higher than a previous point is that a growth period had happened between the two points. You can not conclude that it is still happening or that it had not reversed at some point between the two points.

To accurately determine the trend of a data set, you must look at the entire set.

As a practical example, consider this chart of the stock market which shows the 1929 crash(Cropped for readability). If I were to look only at the points I have marked in red, using your proposed methodology, I might conclude that the market was in a period of growth at the latter point. When in actuality, the second marked point is towards the beginning of a massive decline.

Actually, right, right, right. Sadly, what you're describing is exactly the utterly flawed approach used by the negative echo chamber around here, where they cherry pick completely arbitrary data points, ignore the context, and then try to act as if it means something. "Hey! There were more people playing this game at Point A (when ShB first launched) than at Point B (a random spot in a late-patch lull for EW). The game is dying!" It's worthless.

Meanwhile, the person you were responding to (Ren) provided a thorough, valid, thought-out analysis using precisely the methodology you claim to want. He looked for trends in entire data sets and consistently compared similar points in time across multiple expansion (i.e. looking at the population precisely after Patch X.2 launched, or right before X.4 launched, etc.), and the data over and over shows steady growth from expansion to expansion, with EW being the most populated one yet at every comparable point along the way.

So, your commentary in general makes a great point, but I'm utterly confused why you're directing it at the person who is actually doing the valid comparison you apparently want, while ignoring that the doomsayers are exactly the people using the invalid approach you show is flawed.

And to put another way: if the stock market drops from a high of 50,000, to 20,000, and people say it's a bull market because it was higher than 18,000 thirty years ago... One would question their sanity. Yet this is what many are arguing.

Agreed. It's why we might question the sanity (or perhaps honesty?) of the doomsayers. Ironically, yourself as Exhibit A, given that your "Steam analysis" does exactly the thing here you say would lead people to question one's sanity, just in reverse. The stock market here is cyclical, and at the same point in the last three cycles, it's gone from 10,000 to 20,000 to 40,000. But you act as if it's a bear market because it reached 50,000 at its peak two cycles ago (while ignoring that it peaked at 80,000 last cycle).