Goes to show you how difficult and demanding moderation truly is, and why popular sites/services routinely fail to do it effectively. It's a quite proactive process.
That's an incredibly good point. I would think you'd need a pretty modest operation when it comes to that for a site like lichess, and the fact that it is the biggest expense really puts into focus the kind of scale the problem likely is at the bigger social media sites.
Nothing is modest at Lichess scale, we have a few million users, several thousand of them post in forums daily and chat with each other.
Also, we're a chess website with more than one million games per day, and some people like to cheat. We're classifying "fair play and anti-cheating" as "moderation" in our expenses because the two are intertwined.
The saltiness is much, much less than in some games (shooters, mobas) but a portion of every playerbase thinks they are the worst one even if they are not.
Extremely good point, I don't really game outside of online chess so I can't say that I have a very wide sample pool. TBH the chat I used to see when my brother was playing League blew any chess experience I've had out of the water. I'd almost use a word beyond "salty" for that
I've never ever looked at the chat. And knowing that it costs money, and quite a lot, to moderate that, I'll never report anyone. It's just vanity. I don't care what you write to me. I just beat you in the game. Just cry it out. There you go.
There is all kinds of player bad behavior that needs to be manually verified - cheating is the obvious one, but there's also smurfing, excessive sandbagging, clock abuse (e.g. letting your clock tick down to zero in a loss, just to be annoying).
