The results were a little different than I expected in that the mean is -3, but this is recalculated over the period of the graph to show outliers at +/- 1 StdDev as Buy/Sell signals with weaker signals in between. So the actual mean is not important and seems to be a result of long periods of low bearish sentiment offset by short periods of extremely high bearish sentiment.
Rather than looking for longer term indicators, I decided to set up a short term composite using indicators with the highest short term correlations discussed last week and have come with this for now.
The setup looks very similar to Nov of 2015 with a rapid rise in bearish sentiment over the last few days. Continued volatility over the next few days could very well set up a buy spike for May option exp before a larger decline.