I counted the prior events where the move down was more than 4 points (there are smaller events on the chart, but they are basically background noise given the nature of what the VIX measures). I also excluded the multi-year spike events that aren't comparable to the month we just completed. The resulting odds are 6 in 7 chances that we see 4+ months of VIX decline. The failed event is March 2001. For patterns I look back up to two candles, but most are the immediate predecessor.
Caveat - we start March at 13.34. Only the April 1994 candle ended that low. However, the failed instance in 2001 had the candle end at 21.57 which is higher than the open of the red candle we just completed. In other words, the failed pattern occurred when the VIX was still high after the move.