Feels like greek to me and/but here is the entire use case and you guys are probably smarter than me and can maybe even think up a better solution for it. Appreciate the help.
In the initial post I called it a "Priority score" (thought it would be easier to conceptualize) but it´s really a "Popularity score" and is used for photo cards used by the end users of the app (multi tenant where each tenant has their cards which are then used by the end users). The popularity score should reflect how popular each card is and is based on 2 parameters: card view and card click BUT it also has to factor in active days (as the administrator/tenant of the app can decide on when the card should be active). The card view is based on how many times the card was viewed, the card click is based on how many times links in the card was clicked and the active days is based on how many days the card has been active.
The current calculation I am using is something I found in a forum online and basically is:
Card.StatPopularityScore=Card.StatViewed+Card.StatClicks
Card.StatPopularityScore= Card.StatPopularityScore-(Card.StatDaysActive*0.5)
Source: https://stackoverflow.com/questions/11189452/creating-a-formula-to-determine-popularity
"... I am thinking that younger the uploads are it should be more popular, for example. image 1 was uploaded 2 months ago and has 65 likes. Image 2 was uploaded 2 days ago and has 65 likes. Image 2 should rank higher because it is newer. How could i do this?
...rating each image by considering different parameters. For Example: Image1 is having 65 likes and it is 60 days old(2months), Now to rate this image you add 1 point for each like and substract .5 for each day. In that way ratingOfImage1 = 65-(60.5) =35 , same way ratingOfImage2 = 65-(2.5) = 64 (as image2 is 2day old and having 65 likes). clearly image2 is having more rating so it is more popular. ..."
This gives me different StatPopularityScore based on any of the 3 factors and/but to add to this, I need in some way to be able to get these scores converted to a scale between 0-100 for a particular tenant (is used in the presentation of the score on a color based gauge which can only have values 0-100) and should of course hold in all scenarios and also over time. Regarding MIN and MAX, I could get this from looking at all the cards a particular tenant has and then getting the MIN from the card with the card with the lowest score and MAX from the card with the highest score (or is this incorrect?, again my math and thinking skills in this area is limited and add some stress and no sleep to that as well 🙁 ).
Will the Rocketman calculation work in this scenario? Is there a better way of doing this?
