Nice to have you hear and listen to my Friday evening ramble. I've been racking up an environmental debt getting ChatGPT to explain typescript and react to me. After two weeks of "Explain this to me" and "how can I do this", I'm now at the point where I'm sort of comfortable to try it myself and to ask "Give me 5 suggestions for doing X".
Finally fixed that the right graph loads. This made me realize how few I know about react, async and sync requests and the flow of events. Maybe time to draw a diagram of the flow? In hope that it will also help with the next bug: Duplicate charts! The old chart isn't updating. Instead a new chart is added below it 😠.
Epiphany struck on my Sunday morning run. The quickest way to test if the metrics I'm working on are actually useful for people is a google sheet 🤦I'll have to update it every week, but at least I can get to test the idea iteratively instead of building endlessly alone.
Beyond being stupid and spending too long looking at lines of code without making progress, I've read an interesting piece on community contests. The tldr is that contests and games don't have to lead to ghost-town likes communities, if done well. Contests can be broad to gain more members, or deep to increase member's involvement with each other. I'll say the naming of the game (contest, giveaways vs challenge) doesn't matter. I've seen this play out among scientists: Fight over how something is called, just to get more citations. People don't care, and it doesn't help your cause.
What you need to remember about having games for your community, is why you are doing them in the first place. Do you want to gain more members or do you want to strengthen the bond between members? If there's no friction for outsiders to participate, like with those low-key quests that gets you to earn eXPerience points for retweeting the main account, you'll get engagement (aka posts) and member growth. Your daily active users increase dramatically, and will crash as quick.
Nouns 40 request for a qDAU dashboard for the nouns channel caught my interests. The challenge is that qDAU sounds like a nice objective metric, but that's deceptive. Quality is in the eye of the beholder. An observable feature of qDAU is their interaction patterns. No one likes to interact with someone when the conversations are gonna be superficial and shallow. The first step is to map interaction patterns. I'll say what is needed isn't a network graph, but member profiles like in FarcasterUserStats.com by welter.eth, or a dashboard showing number of casts, number of replies, number of reactions, and whose getting most often mentioned.
Leaving you with the question Is reputation a scam? - thanks Peth and DAOcember for creating the event. Act III is starting this week. 👇
