This might be the effect of Yobit signature campaign.
The traffic of bitcointalk was declining until a week ago. Is this due to Yobit signature campaign?
Correlation is not causation. Volume has been chugging along and any small spikes are most likely just a higher granularity in the graph. If you look at the entries other than those few days then you see that they've just made large overarching lines.
I completely understand you. A website rank does depend on traffic of other websites too.
But, when a website rank increases 500 in a day, can't we say that " The website traffic has been increased." ?
I think it's unlikely that traffic of 500 websites decreases at the same time.
It's not completely accurate, since traffic could remain stagnant or even fall, yet rank could still increase. All you can sufficiently say is that their rank increases when their rank increases: a tautology.
No matter how you want to redefine it, it will eventually crawl back to being
rank increase |=| rank increase