Okay, so, looking at individual animal captures, we have around a 12% Confidence interval with our 50 captures.
Paissa (which is at 92 captures it looks like) is at a 10% confidence interval. So if you were to capture 100 of everything, we'd have a 10% Confidence interval. To get that down to a 5% We'd need to do ~400 captures.
That being said, I don't think we NEED to do that. +/- 12% is a pretty big range (it means that the correct value for the capture percentage could be wrong by up to 12% in either direction) But all we REALLY care about for the individual animals is seeing what the common trend is (are all rarity 1s the same? Rarity 2s? Rarity 3s?) Once we do that, we can collapse the data and suddenly we HAVE our >400 captures (800 captures, actually)
Like, looking at all the rarity 3s, they all seem to have around the same success % of 50%. Morbol seedling is absolutely an outlier there, but since it's a singular outlier it seems unlikely that this is due to a trait of the morbol seedling and much more likely that it's just statistically unlikely (We are going for a 95% confidence interval here, so we're actually expecting 1/20th of our animals to be weird and not right.)
Same is true of our rarity 2s and 1s, and our rarity 0s. So despite us not being able to really trust our individual animal capture rates, we've still managed to generate statistically significant evidence for the hypothesis that the % chance of success is based off of Rarity and no other factors (Smalls/mediums/larges don't have the same grouping, even inside rarity tiers, and there does not appear to be any kind of pattern around leavings or the kind of weather/time that they spawn either.)