AnomalyDetection

Any client side tracking solution is vulnerable or exposed is a better word to this. It's difficult for me to say it's a security issue because really by design everything about a javascript based tracker is public and really even server side trackers are not immune if someone decided to inflate your numbers or mess with referring traffic information it's all based on what the client sends you. I think in this case google maybe able to add some sort of machine learning to indicate in the result sets that certain links/visitors appear to be either bots or explicit spammers. Perhaps someone could even create a third party tool to do the analysis against a GA account using https://github.com/twitter/AnomalyDetection

Even the original twitter blog seems to have gone out of its way to make this seem more complex then it is... Decomposition of time series is done with STL (stl function in stats package) and this is the first part of what they call "Seasonal Hybrid ESD (S-H-ESD)" (sounds impressive right?) which then apparently just involves taking the max absolute difference from the detrended sample mean in terms of standard deviations, remove it and repeat until you have your collection of x outliers. If they wanted to this could be explained in a few sentences, and the underlying code is really simple [0], but for whatever reason it's been written up as advanced analytics — as if decomposing a time series is a major challenge. [0] https://github.com/twitter/AnomalyDetection/blob/master/R/de...