> I also thought it would be interesting to remove the time aspect and simply look at these Series A rounds as a progression of deals, regardless of when they took place. This graph looks at the 10 deal moving average of Series A round size.

That's not how moving averages work. You can't just take an unordered univariate and smooth it however you want.

You argue in your first chart that time is a relevant factor; therefore, you must smooth as a time series for that analysis to make any sense. (which wouldn't make sense in this case anyways, due to the nonconsistancy in timing of seed funding.)

Also, the annotations in the first chart make absolutely no sense since they are not relevant in any way. (unless you are asserting that correlation of these events implies the causation of the increase, which is very bad logic unless you have proof otherwise)