@Smee What I'm kinda envisioning (which is probably not the best or most efficient way to do it) is the program, when opened, would open the text file, read it, regex each line to store the item names, credit average, and current standard deviation, then API the forums for threads posted since the prog was last opened, read them, search them for item names, anytime an item name is found, find the next credit value number (or none if it hits another item first), check if it's an outlier (farther than 2.5 SD away from mean), if so, throw it out, if not, add it to the average and recalculate accordingly.