php - Normalize merchant weights based on long term performance (mysql data) -

php - Normalize merchant weights based on long term performance (mysql data) -

i have database of retail merchants. each record has following:

id | merchant | clicks | sales | conversion rate | epc

the problem have in situations new merchant has 2 clicks , 1 sale gives new retailer 50% conversion rate , high epc, "better" merchant has 200,000 clicks , 40,000 sales. looking way come number between 0 , 2 or 1 , 2 can computed based on above info , used give more weight retailers improve long term performance (higher number of clicks , sales). sql query based on info given ideal, can in php.

let me provide actual numbers. next info sorted number of sales , accurately shows our top merchants. there still few problems ordering since can see click nordstrom worth twice of click zappos.

merchant |clicks |sales|conversion|epc 6pm |982903 |21846|2.22 |0.10 zappos.com |1594711|13741|0.86 |0.09 nordstrom.com|564169 |12028|2.13 |0.18 shopbop.com |249714 |3910 |1.57 |0.16

php mysql statistics