Quote from: iamnotback on July 19, 2016, 09:42:46 PM

Quote from: jl777 on July 19, 2016, 09:01:53 PM

https://steemd.com/distribution has the raw data.

I had explained upthread that data is insufficient and ambiguous w.r.t. to the half-life of the attrition rate:

Quote from: iamnotback on July 16, 2016, 04:36:00 PM

Quote from: Zer0Sum on July 16, 2016, 04:05:26 PM

The 10,000 number is accurate based on 7 day account activity from the steemd blockchain.

https://steemd.com/distribution

If the 3000 signups per day claim is correct, then they are seeing at least 60% abandoned accounts rate, which also is approximately corroborated by the loss of actives of newbies from all history to recent 7 days.

Only 28% of the weeks active are active on any given day. So the majority are not revisiting every day. Given the evidence of a high rate of attrition, it is possible that instead of gaining a set of sticky users which is 40% of the signups, instead what might be happening is after so many days, nearly 100% abandon the site and the new signups are replacing the lost people (mathematically plausible if the rate of signups has been increasing which would also mean the attrition rate is much higher than 60% when comparing recent 7 days to all history). Meaning that it is possible the site has no stickiness. We need see charts on users by age of account and activity in order to answer this question. Also to more accurately see how the rates might be changing over time.

The last 24 hours stats might be skewed by the downtime due to the hack.

I just look at latest 24 hours data 4,821 accounts and weekly 13,734 accounts

if those number keep going up then it is not at equilibrium yet

of course it is noisy data and to get a really accurate idea, the blockchain would need to be analyzed. however many people make more than one account, and recent hack incidents surely slowed some things down