So I've been gathering data on AeroPress Pourover style brews (this recipe: https://youtu.be/njxTlITJ6Ds), and I noticed something really interesting (which I mentioned in the video): the grind size has a significant impact on the resulting extraction yield, when dose and brewing ratio is kept under constant across different beans with different roasts and processes, albeit the water temperature ranged from 60 to 70C. The data is scarce, sure, so I decided to test out the formula from the linear regression ("before" graph), which predicts 28% extraction yield by grinding coffee at 1.0 on ZP6S.

It did not hit 28%! It only reached 23% EY, which is still pretty high, but I was wondering what might be the reason for it. My first thought was to inspect the puck, and surely enough, it had several channels (see if you can find dark spots):

So channeling was surely an issue. Another possibility is that I used Colombian beans, which may have less solubles in them. But that's just a guess.

What if I ground coarser? I moved the grind setting from 1.0 to 2.0, and the resulting brew hit 21.3% EY. The coffee was definitely more evenly extracted (the puck still had a dark spot in the center though):

After adding these results, the correlation still holds up ("after" graph), with a lesser slope. It's not really that surprising that grind size matters, but it is still interesting to see how much it can affect and has affected the extraction!

Of course, this experiment is kind of flawed, because the pressing time differed between these brews; I will be repeating this experiment in a more controlled manner at some point.