The discussion revolves around customizing the Pageviews plot. BryanV wants to make two changes: hiding the “crawlers” bar, which is often uninteresting and causes the vertical scale to become unreadable, and aggregating the year view by week to better track growth over longer periods.
Codinghorror and joffreyjaffeux suggest exporting the data to graph it as desired, with joffreyjaffeux noting that an update now allows hiding dimensions by clicking on the legend. BryanV explains that, due to time constraints, he prefers the changes to be available out of the box. Stephen agrees that daily values might not be the most useful, making trend identification difficult. Mittineague proposes graphing the data exponentially instead of linearly after exporting it.
Wondering if there is any simple way to make some changes to the Pageviews plot:
Can the bar for “crawlers” be turned off? It is mostly uninteresting to begin with, but then the periodic spikes blow out the vertical scale and make the rest of the plot un-readable.
Can the year view be aggregated by week? We would like to look at growth over longer term, but we have a heavy dip on weekends so a weekly binning on the “year” scale would make that much easier to discern
Why don’t you export the data and graph it however you like in the tool of your choice? 
I pushed an update few days ago which lets you hide a dimension by clicking on the legend, this will solve your first bullet.
For the other as jeff said if you have any specific grouping requirement you better export it.
I’m not sure there’s any value on showing daily values like this, at a glance it’s impossible to tie spikes to dates.
We’re not using a mass spectrometer where the individual spikes identify individual molecules. It’s nice to see and investigate blips, but it doesn’t make trend identification (i.e. growth) easy.
Maybe after exporting the data graphing it exponentially instead of linearly is what you want?
Truthfully? Between work-work and OSS-work I am overcommitted and just want to have it available out of the box. Given that daily bars on this scale are not really useful to begin with, I hoped perhaps that our interests would be aligned on this point.
