The issue isn't with pessimism. The issue is with seeking out sample sizes that support the narrative rather than building a narrative around the data. Your order of operations is backwards. My response with various sample sizes that showed an excellent offense weren't meant to prove you wrong. They were meant to demonstrate that statistics, sufficiently tortured, will confess to anything.
This is one of the oldest faux pas on this site. It's silly that it's something people still do.
I hear you. But sometimes small sample sizes indicate CURRENT trends. Which may matter greatly.
None of us really knows how big a sample size is the 'correct' size. I mean, many players have one great season over a pretty lousy career - or vice-versa. Was that one season enough data to conclude that the player was good or bad?
So all I did was look at how the team has been hitting most recently and what I saw over a week was not encouraging to me. It has "felt" like their bats have gone quiet. So I looked up the data over the past week and it supported that feeling. Yes if you go back far enough you'll see the Red Sox hitting well, especially over the course of the season.
I'm not asking the question of whether player X is a good player, or whether the Red Sox are a good hitting team or not. And if that's the question you're asking, then a small sample size is not helpful. But if you want to ask the question of how well player X is performing now (not literally this very second, but very recently), then a small sample size can be instructive.
Take Jackie Bradley Jr as an example. He goes through major swings of hot and cold. On the whole, he's a pretty darned good offensive player (much to our great joy), and when he's hot, the dude rakes. But when he's cold....man it's painful. So if you're asking whether JBJ is a good hitter, of course a small sample doesn't tell the story. But if you ask whether, in today's game, you're confident that he's going to hit well, isn't it reasonable to look at how he's done LATELY, as opposed to how he was hitting two months ago?
So I think it's all about the question you're asking which helps you determine the relative value of a small sample size.