Ideally, it would be a combination of exponential and linear components that correspond to the factors that change when blogging. The number of followers means nothing if they are not active. So would would be a better model? It looks like linear models are not very reliable since they fluctuate greatly with post frequency and diligence of followers. That, however seems extraordinarily high. Using 33.33 as the coefficient for a linear model, I would have an increase of 1000 in a month, 12000 in a year. I was unfortunately not recording data at that time. This estimation does not seem very accurate considering I reached 500 followers 15 days ago, or an average increase of 33.33 per day since then. The longer term has me at 945 in 30 days and 3681 in a year. They estimate I will gain 58 this week, though the straight math is. (The past 6 days are a rather poor sample size since I’ve not been as active on this blog, which led to a lower current average.) For the past 6 days since I began taking data, I have gained 49 followers. Sadly, however, their estimation is based only on a linear increase of followers.
They also provide a ‘follower prediction’ for a week, month, and a year.
#Rss follower count tumblr code
I inserted followercounter’s code into my blog’s html to help keep track of my stats. I actually thought about creating my own plot with data I would manually enter every day, but had been lacking the energy to do so.
One thing that I find fascinating about tumblr is it’s viral and exponential growth capabilities. Taking data, looking at statistics, modelling growth. AKA boring things that interest me.Īs an engineer, I really enjoy data. This is mostly a massive analytic text post for my own information/posterity regarding the exponential and mathematical nature of how one might better keep data, model, and predict the number of tumblr followers one might gain over a period of time. Trying to come up with a follower formula.
0 kommentar(er)
