This. Ron Kohavi 1) has some excellent resources on this 2). There is a lot of noise in data, that is very often misattributed to 'findings' in the context of A/B testing.
Replication of A/B tests should be much more common in the CRO industry, it can lead to surprising yet sobering insights into real effects.
Thanks for bringing this up. Just to clarify, the base version of GA provides an event-based model. GA App+Web is a feature of GA that's aimed at sharing/consolidating info between websites and mobile devices. We're focused on the web side of things, as we've insofar only written a web analytics article. We suspect GA becomes a bigger player when we take a look at mobile analytics, but we haven't fully delved into that yet.
information was scarce indeed, till this day i still feel a little sorry i was not born 10 years later. Since when i learned (gw-)basic from the 3 books our public library had to offer i was full of enthusiasm, but after that my development practically stalled due to lack of information. Until the internet came around, but that was 10 years later and i was already in my early 20s :|
A small bias towards more variety was introduced to prevent 80% of locations being in the US. Repeating locations are possible but should be increasingly rare.
I'll further tweak the randomization patterns for the better. And fix the address language indeed.
One neat feature/bug is that if you have multiple monitors the screens seem like they are pointed to different angles of the same exact place. I suspect the randomness algo is based on JS timestamps.
That's great and all, but a real screensaver provides a cheap way to lock a computer, and is functional and can be beautiful as well. This is beautiful, would love a screensaver of this to lock my computer when I step away from it.
Random street view locations are retrieved from a database (on-the-fly lookup is too slow); France had not so much records; that has changed over the last 5 minutes though. Try again :)
