Couple requests:
1) Could you break this down by time slot & template?
2) Could you plot this against account & overall game creation rates?

The analysis of distinct clans and personal stats will probably a topic for the next episode (probably 1 week before the end of CW season 22, when I will want to promote another Prime recruitment thread :P ).

But here are some fancy charts on breakdowns and quick thoughts.

Games per slot type
Had to do this and next chart on a weekly basis, because daily data makes the charts an unreadable mess.



- So basically it is the 8 and 12 hour slots, which has rise the most since the death of CW in December 2021.
- Other slots dont seem to change that greatly


Templates breakdown



Maybe the SECB has some dynamic? Besides that, dont really see anything behind the noise.

CW games as percentage of total Warzone games
We can estimate the % of total games by the game_id deltas. Although there may be issues, with that methodic, it is still something.
The "0.004" value in Y labels may be also read as 0.4%.



Now that's interesting:
- The CW share of games has risen to the same levels as during the first 1-3 seasons when we saw the peak absolute participation.
- The dynamic is a little bit different than that in absolute numbers (clearing the summer season effect probably)

---

⬛️⬛️🟩🟩⬛️🟩🟩
⬛️🟩🟩🟩⬛️🟩🟩🟩
🟩🟩🟩🟩🟩🟩🟩🟩🟩
🟩⬜️⬛️⬜️🟩⬜️⬛️⬜️🟩
🟩🟩🟩🟩🟩🟩🟩🟩
🟩🟥🟥🟥🟥🟥🟥🟥🟥
🟩🟩🟥🟥🟥🟥🟥🟥
🟩🟩🟩🟩🟩🟩🟩🟩

Damn, I'm getting hot and bothered by all these charts. Thanks for compiling and breaking down the data available, 5S! ●

Thank you 5s! btw, to get the graphs to look less ugly, you can just do scatterplots and add linear regressions (if they fit well). That way we can spot trends rather than a heartbeat monitor. The only trend I can make out is an apparent shift from SE1W to cumbomb.

I believe cubic splines are better (they keep the concavity or convexity of the function they approximate, assuming the function isn't too wild).