[pcabc]

Are you certain your data is that of those stocks in the S&P500 at that period back in 2000 -2001?

The breadth data, except for the percent above the 50 and 200 day EMAs is for the NYSE.  For the percent plots it is based on on a few stocks within my database that I have price data for in that time-frame.  So some care is needed but I suspect those couple of plots are not far off what they would be.

Another thing to maybe play around with is RelativeStrength of Value vs Growth or Growth vs Bonds or MidCap vs LargeCap and see if that can bring down that BUY score during those times.

Sorry, I don't follow what you mean with respect to 'RelativeStrength of Value vs Growth' or 'Growth vs Bonds'.  The data is largely based on the NYSE and comprises of several of the breadth measures in the book and a few others I have added over time.

The simplest way to remove a lot of those buy indications was to remove buy signals made when the index price is below a falling 150 day moving average.  Basically do not buy in stage 4.  It is an interesting point as so far I have deliberately based the signal on the breath signals excluding the index behavior.  Given the index price action is so clear I did not think it necessary.  I'll have to ponder that one.

The challenge is we can tailor make our scores to fit 2001, 2008 & 2020 and then not be fool proof for the next down turn. Anyway its all part of finding something that works the majority of the time, so it is working alot better in 2020 and thats important.

Exactly.  A month or so before the down turn I found downward cross in moving averages (edit) of the advance / decline line, that occurred before the 1987 crash.  But this time that indication was much too slow.  One needs to be careful.