nanasplitter

So, it has come to my attention that the Growth/loss estimation program uses a linear regression to determine the data for the graph displayed. While in most cases this is perfectly acceptable, I would much prefer if you switched to an exponential regression so users could get a more accurate set of data and understand which items are going to flip, and which items are going to flop. A generic, variable example of this follows: Item x has prize y. Price y fluctuates greatly , and at a rate that has a pattern that can be calculated with an exponential regression system. However, because the current system uses a LINEAR regression system, that pattern can not be predicted, and a massive potential profit is missed out on. Example B: Item A has price B. Price B Has been steadily decreasing but has spiked suddenly over the last few days. Someone checks the projected growth of item A and it is projected to continue growing at the current rate. However, after the item is purchased, it's price drops down to it's original, much lower, cost. A massive loss of RSGP is experienced due to this incorrect statistical calculation. Long story short, a linear regression system does NOT cut it for market predictions. An exponential prediction is necessary when predicting things that are constantly fluctuating. P.S.: If this is in the wrong section, I apologize. This is my first post and I didn't see where else it should go.