I just started using Twitter about a month ago after having an account since 2009 and I'm really enjoying it (despite seeing a number of tweets about Twitter being broken). I've almost entirely stopped using Facebook (outside of events/messaging) and strictly used Twitter. It's a fun place to share programming thoughts and discuss things with people.
For the most part I've been able to share links and chime into discussions with people whom's work I'm interested in and keep the discussion around economics, mathematics, statistics, and computer science. I follow roughly the same rule as you - I look at the tweet frequency and then dig into actual tweets. I've actually made a couple friends in the Elixir community through Twitter which is actually fun.
The last thing I would recommend is to use lists. @patrickc's reading list is probably one of the best things I follow, as well as @pmarca's press list. You can actually check out the lists I follow as well.
https://twitter.com/udayrsingh/lists
I think for the most part, like you Max, if you try to keep it small - it works great.
I'm following you. Better yet, I'm following who you follow.
https://chrome.google.com/webstore/detail/hackybird/ddlhmpom...
I think more people would like Twitter if they didn't have to spend 30h+ curating content until it becomes valuable
First thought was, great feature, but I had the same issue with batch editing, then I tried using lists through Tweetdeck and it is much easier to create lists, remove & add profiles to each list, Twitter might actually become useful for me with this.
if people start spamming or their signal/noise ratio goes down I unfollow them quite quickly.
Yeah, that's my greatest wish too - something like average tweets per hour/peak tweets per hour/average tweets per day/peak tweets per day stats.
Went from following 1500 to only 400. Almost no politics now and high granularity & outstanding signal to noise ratio. An insight-machine that I would not want to miss.
Selection criteria I use: - don't pay too much attention to profile, better to check actual tweets. Plenty of smart/impressive sounding tech/science people who talk about their kids, sports and politics all day - sane tweet frequency: If they've tweeted 20k times, there's a good chance it is just noise/spam - select people who follow few people, anything above 2k is suspicious ("I'll follow you so you follow me")
Here's the list of people I follow - the further down you go, the longer they have survived my aggressive filtering process (another design flaw: it takes minutes of scrolling to get to the good stuff):
https://twitter.com/mxschumacher/following