I'm sharing an email exchange I had earlier today in hopes that it will help others to find ways to get exactly what they want when using the site to scan for stocks. A friend sent me this:
I was looking through your hammer scans on SwingTradeBot to see how Zillow stacked up on the grading curve. I was surprised Zillow (Z) only got a B. I looked at the A rated puppies and felt like Zillow (Z) should be higher rated. Then I realized what I really want is a modified hammer scan. I want hammers that come at the end of a period of selling, even better if it is an oversold hammer or a hammer coming off a high. The preceding selling has to be at least three straight down days. Did you already read my mind and create such a thing, or do I need to put a suggestion in the idea box? :)
Here's my answer:
So I think you're misinterpreting what the grades mean. They're essentially relative strength ratings. So, in general, the A-rated stocks are going to be closer to their 52-week high than B-rated ones...
I've written the hammer scan such that it takes into account the shape of the candlestick and if the stock made a 5-day low. The 5-day low thing was my attempt to encode "after a downtrend". I know it's not nearly the best way to do that but determining what's a downtrend is a pretty fuzzy thing to code. Plus I wanted to err on the side of being too liberal rather than exclude some potentially good signals.
You could achieve the things you describe by using the Combo Scan. You could combine the "Down 3 Days in a Row" scan with the Hammer scan. You could even add the "Oversold Stochastic" scan into the mix to require all 3 things to be true.
Alternatively, you could just use the hammer scan, sort the results by Grade and don't look at the A-rated stocks because they're not likely to meet your oversold criteria.
I get questions like this pretty frequently and a high percentage of what people ask me for can be solved by combining scans. One of my goals with the site was to try best to keep things simple but still effective. So I came up wiith scans that I thought would cover 80% of important/popular technical signals. The Combo Scan enables people to add some more customization into the mix without me having to add to the already too long list of scans. (I am considering removing some.)