once again it's a cumulative chart aka "vanity metric" we don't use this kind of metrics to judge the virality of a product. Look at the new accounts/day (non cumulative): if you had a "y = ax +b" curve with a !=0 then you'll have a 'real' exponential curve on your cumulative charts
You are viewing a single comment's thread from:
I perfectly get your point, let me post you an extra chart, below is the monthly linear (non-cumulative) chart, you can see it shows an exponential growth as well. The growth is really exponential, not just linear (y=ax+b) we have something that looks more like (y= exp(ax) with a>1/2).
Of course for a monthly chart where you take into account stats prior to the 4th of july you'll have this kind of curve, as most of the new users came from the month of july... What happen before 4th of july is what i would call a pre-launch. So it should not be taken into account.
And as right now we are in august, having a monthly chart where only july count, doesn't provide any sufficient data imho to conclude that Steemit has reached escape velocity.
For me when on the account creation per day chart (https://steemle.com/charts.php) the curve will go up to the north then it will be trully viral (not counting once again DAUs or retention rate)
Total accounts doesn't say anything about abandoned accounts no longer in use, which I documented to be in the realm of 85% abandonment rate.
Also plotting on a non-logarithmic chart makes a constant rate of increase look exponential which is mathematically incorrect. See my other comment on this page for more information.
It's not exponential. It's just 2 linear lines in a row. If you want more accurate stats go to my profile @calamus056 and watch the 3 part statistics articles i released.
@anonymint, abandonment, which is difficult to accurately measure anyway, can be almost entirely ignored as a growth statistic if you compare active accounts over time. Daily active has been growing significantly and is now close to 6000 from being closer to 4000 not long ago (1-2 weeks IIRC, but I don't have hard numbers). I'm not sure about the weekly-active numbers as I haven't paid attention to it.
I agree with the earlier comments that: 1) constant growth is not exponential; and 2) there isn't enough data to conclude much yet, except that there is steady, but still kind of slow growth, even after factoring in abandonment.