I recently posted on the topic of historical market investigations and how those can aid one's feel for markets. I find they can also provide longer-term context and ideas for short-term traders.
Here's a simple example. VIX has been making 20-day lows, but SPY has not been making 20-day highs. In other words, volatility is coming out of option pricing, but it is perhaps more due to stocks topping out and trading in a range than trending higher. One might think that such a pattern would have bearish implications going forward.
Looking back to 2000, we find that when VIX is making 20-day lows, but SPY is not making 20-day highs, the next 20 days in SPY average a loss of -.72% (76 up, 100 down). For the remainder of the sample, the average 20-day return in SPY has been -.16% (1215 up, 957 down). Indeed, when volatility comes out of the market and it's not due to price strength, intermediate-term returns have been subnormal.
Such information can provide background for the daytrader looking for reversals of strength going forward; it also can frame useful swing trades. For those testing historical patterns, the logic underlying this particular pattern might inspire investigations at other, shorter time frames. ;-)