Day 3 started with talk of the night before. The entertainment for the evening had been "trampoline dodgeball" (I'm not kidding) and although this had been a lot of fun, the main talk was of one of the GAAC's who had been rushed to hospital with a dislocated ankle! We were told that he is doing ok, so I wish him a speedy recovery.
After that interesting start to the day, the subject became very dry, with Hal Varian, Google's Chief Economist, taking the floor. Although the subject matter was intensely data focussed and of a serious nature, it was incredibly interesting. Google are developing more and more ways to "predict the present" i.e. use search data to know what the current actions certain populations are undertaking, several weeks before official statistics are available to the Government. This algorithmic prediction analysis (regression, confidence intervals etc) is a similar methodology to the way they work out the recently released predictions in "Insights for Search".
The rest of the day was then set aside for discussions on the use of the API interface (how others are using it, how it is changing, what we'd like to see in the future) and also what other GAAC's are up to generally.
The energy of the day slowly drained throughout the afternoon, as keen as people were to be there, 3 days of in depth discussion on Google tools can be tiring! I could easily have gone to bed early, but I was fortunate enough to meet up with some of the guys from Luna Metrics (John and Jonathan). After several Tequilas I realised that John Henson was responsible for creating the goal copy short cut that we have used many times, so I knew that I had met up with some real experts. I am sure we'll get the chance to meet up with them again next year, along with others, the aim being that by then, we will be able to give something back to the group.
Looking forward to it already.