PainChek General Discussion, page-15208

That's literally my idea Soo far no one has done it. Either they have done it and keep it to themselves or are lazy to do it. Both outcomes are likely. I am in the lazy camp and not too fussed. This is a long term hold.

You can get excited here and there when a spike goes up, but if you hold you get the rewards. The download spikes if anything is a sign to buy more when considering many variables. You need to apply correlation on the "spike strength" to workout if the app moves up in download band.

Here's my framework to do it;
1. Get the 3month/6month (if you pay for it) app rankings
2. Create a database of the top 200 apps in the app category. In this case it's the medical section.
3. Manually fill out the number of downloads per app. If you are a programmer then parse that, I guess.
4. Extract App ranking movements of the 200 apps for the same time period as well.
5. Make Painchek the base case (no.of downloads + app ranking movements) for the analysis
6. Find patterns

• One pattern would be to count the number of upward spikes and correlate with actual beds in the quarterly
• Another pattern is to categorise the top 200 apps in bands of download (i.e. 0-1k, 1k-10k, 10k-100k, 100k - 1M). My assumption is that highest band = highest ranking and the bands go descending order. But you have to do the grunt work to prove that's true. Then look at the app ranking movements between bands. If you see an app that was previously in the 10k-100k band move above the highest ranking app in that band. Then you can make the assumption that the app now has more than 100k downloads. You then wait to see if the app can maintain their ranking to be certain.
• So, for instance if you see Painchek go from ranking 43 to 1 then something major must have happened. I.e. Look at the downloads of the number 1 app. The next key important thing to understand is that if the ranking goes down from 1 to 100. Then you could say that no large growth in new downloads. However if the app maintain ranking in top 10 then it means that have moved to an upper band.

7) Update data daily/weekly/quarterly (You now see why this is a lot of work)

There could be more ways to slice and dice the analysis. That's how I would tackle it. I am being a lazy satoshi and not doing it myself but for anyone who are too agitated that's how you can attempt it. It won't be perfect but atleast it's a framework. There are probably better ways to do it, but I haven't seen a framework.

Good luck.