Now (from August 1, 2015) we show data according new pageview definition using Pageview API. The main differences are:
This website provides a graphical interface for aggregated pageviews statistics for English Wikipedia articles. Our data aggregates number of pageviews or hits of a certain article as well as all its redirects and some variants.
Wikipedia articles use specific conventions for naming titles so please be sure to enter the article titles exactly as they appear on Wikipedia. To avoid typing errors, we suggest using Autocomplete feature.
Use your mouse to move the sliders at the bottom left and/or right corners to get your desired date range.
The moving average in this case is a simple moving average, which is the unweighted mean of the previous n datum points. The default number of datum points or value of n is 7. To change the default value enter your desired number into the square in the left bottom corner and reload the page.
Our website's data is based on the number of daily visits to English Wikipedia articles: Wikipedia pages and all redirects to them.
Visits to Wikipedia pages have several patterns: daily, seasonal, etc.
In order to eliminate influences of these patterns on articles' pageviews we use the normalized
data calculated as a number of daily visits (or hits) to a certain article divided by a number of Wiki Main Page hits for the same day multiplied by a maximum number of Wiki Main Page hits.
((Daily page views/Main page views) * max Main page).
This scale shows the same data as other graphs for a certain article but here each value is divided by a yearly maximum value of hits for this article. It makes easier to compare data fluctuation for articles with different number of pageviews.
Over the years many very useful sites with visualization tools were developed. To mention a few:
The original data integrity is severely damaged by site outages and underreporting. Sometimes all the data for a certain article was missing for several days. We tried to correct this, at least partially, by substituting the missing data with data from previous hour(s)/day(s).