Twitter API - randomization worth it?

Just starting out with twitter and building out my own solution which uses their api.

Been building in some randomization stuff to make the follow & tweets look more 'random' ... but then I thought ...

they know I'm using their api, why should they give a fuck about when I follow/unfollow/tweet etc as long as I stay within their rate limits?

e.g. I've been running a scraper thru their api over the last few days and scraped over a million accounts withoutany bans/trouble at all

any thoughts on this?

I know that randomization is good and required even you use the API based on my experience. Just think like that you didn't like the design of Twitter web-site and mobile app. So, you've decided to use or develop your own Twitter wrapper application based on some customizations. But you would still use it manually to follow//unfollow or post tweets. Maybe, you are up to make your service public in soon completely conforming with regulations. These are only the alpha/beta tests.

Anyway, twitter api regulations already say that any automation of follow/unfollow is restricted. If you don't spam too much then tweeting is okay. I don't know anything about scraping. But it must be safe also. So, Twitter wants us to make sth to be used with manual follows with their api. This is why I think randomization is required to simulate real humans. I even make my accounts sleep 6-8 hours in a day.

Besides, you can easily implement a random scheme considering that you have knowledge of developing your own bot via the API. Why bother and take risks?

@kctnky

Good points, well put. Thanks for the different perspective.

More randomization for me then!

Using their API in time intervals as a way to mimic human behaviors is the best way, as long as you don't start spamming and going over the API limits all the time they will not care.