I appreciate your saying this, because I really don't want you to feel like I'm picking on you personally. In this case, all I looked at was your reply to the GP. The GP comment was too irritable to be a good HN contribution, but reducing it to "USDS is bad because you personally don't want to work there" seemed like an obvious violation of the plausibility guideline (which is basically just the principle of charity). So I replied with the guideline, as I often do [1, 2]. It's really a shallow and mechanical response, which frankly most HN moderation responses are, and which I suppose most things are that one repeats so often.
For us most moderation interactions are stateless, because we do thousands of them, which sandblasts the brain [3] and scrubs it of previous state. That creates an asymmetry: because individual users interact with moderators only rarely, they're more likely to remember the interactions. Plus there's the authority thing: it sucks to be reprimanded by authority, no matter how mildly, even on a trivial playground like an internet forum, so inevitably that impresses itself on the memory, usually with a lot of extra torque. None of us is immune from that psychology. We try to mitigate it but we're not perfect, plus we can't mitigate what we're not conscious of.
Then there's the quantity issue: even if we never slept, we could never moderate every violation of the site guidelines or even read all the comments to find them all. So every regular reader of HN is going to run across comments that should have been moderated but weren't. It's irresistible to give that an interpretation—nobody looks at it and says to themselves, "Ah, randomness". Instead they see confirmation of whatever bias they fear the mods are governed by—usually a political or ideological bias, sometimes a personal one, occasionally something on the long tail of weird.
In reality, all we see is a random sample of comments plus the ones that readers bring to our attention by flagging and/or emailing. That leaves a large corpus of unmoderated material that provides way more than enough sourdough-starter for every perception that is out there to feed on. The natural human response is to construct a story about what happened: what the moderators did, why we did it, what we were thinking and feeling when we did it. These stories are basically all made up, because making up stories is what we all do all day; the brain is a compulsive curve-fitter, and there is almost never more than a handful of beans for data.
I can tell you that, to the extent that I recognize your username, I associate it with someone who has improved their HN commenting style over the years—which FWIW is a high-praise bucket in my mental hashtable. But if you really want to be relieved of the issue that "I don't see you making this kind of correction to other posters", the surest way would be to read backward through https://news.ycombinator.com/threads?id=dang until you can stomach no more of it.
1. https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...
2. https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...
3. https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...