I saw a very interesting stat last week and I'll talk about that first this morning. The stat was at marketwatch.com and you can see the article here. I've been looking hard at this and have a few comments to add in the current instance.
I would say first that I have not checked whether the statistic itself is correct and am assuming that the bare facts that these series have only occurred at these times since 1970 are correct. I think that is a reasonable assumption however, so am proceeding on that assumption unless I receive strong evidence to the contrary.
I have no issue with the basic conclusion with the article, that a series like this is found at or near the end of a decline, that it means that extensive further downside this year is unlikely, and that, with less confidence, the odds favor SPX being higher in a year than it was at the 1930 level reached at the close on 17th February, which was the last of this three day series in 2016, are pretty good.