It’s just an incorrect presentation of the concept of small sample size. A singular event or even a game’s worth of them is not a small sample size. That’s like claiming a stock performed well on a day and saying, well, all indications are that’s a great stock, but, you know, small sample size! No. Stocks move up and down every day. One day’s performance only tells you how a stock performed on that day. Small sample size is something that seems to evince a trend and may very well indicate a trend, but lacks sufficient data to be relied on statistically. Small sample size is a pitcher throwing unusually well for a month or two of a new season, but you’re not sure if that pitcher is actually a very different (and better) pitcher from past years or if it is just a statistical blip, because, small sample size. Small sample size is an event that occurs infrequently enough that it is hard to build up sufficient data to assess. Like, maybe, “percentage of times a 2 point conversion is successful when the defensive team has 12 men on the field for the play.” That’s just an example off the top of my head, and maybe not a good one, but say that in the course of multiple seasons, you only have twenty occurrences of that precise scenario - you can hope to draw some sort of meaning from the results of those twenty plays, but, it’s only twenty plays - which is insufficient data - so, small sample size. Small sample size is trying to assess some aspect of the defensive performance of a backup catcher versus the starting catcher, but the backup only plays in 20 games a year, so your sample is not sufficiently robust. Or comparing games caught by the BUC w/ Pitcher X vs. games caught by the starter with Pitcher X. There you have an even smaller sample. I am very confident with where I stand in this conversation, this is not grabbing the last comment, this is just an answer that it has become apparent would need to be provided to move on from this weird digression.