Oct 24 2016
Quite interesting. To add to TJones' comments, have you looked at what the
ZRR rate would be if you dropped the double quotes from properly quoted ZRR
queries? Aka simulate the automatic removal of double quotes if no (or few)
results were found.
May 12 2016
My apologies for the delay.
May 11 2016
SERPs above look very promising.
Mobile is likely what you should be optimizing towards. Laptops are going to disappear quickly. Mobile queries can be more difficult due to issues like poor typing, voice recognition queries skew towards full sentences eg, "What is a gubbins?", and a (maybe) more local intent. Local intent ex: "ferry building" when standing in SF should bring up [[San Francisco Ferry Building]].
Apr 13 2016
Apr 12 2016
re: "One possibility is to use nDCG, but this requires having a measure of relevancy from 0..3 for the results."
Would it be hard to make a test system to verify the contents of the index?
Apr 8 2016
Another significance tester - http://www.evanmiller.org/ab-testing/chi-squared.html#!20716/30299;15273/21551@95
Similar bayesian calculator - http://developers.lyst.com/bayesian-calculator/