I love to program. I love EVE. How can I use these facts to waste an entire weekend?
I decided to create a market analysis tool in Matlab. I started by toying with Eve-Central APIs and eventually figured out how to import and decode the data from an xml file in Matlab. I whipped up a quick gui so I wouldn’t have to stare at an ugly console (might as well just use the Eve-Central website! :D)
The program lets you type a few letters in the upper left and hit search to get a list of market items. I downloaded a look-up table with the IDs for each item so they can be matched to the names seen in the list. When you select an item, the ID is found and the Eve-Central site is queried for current market data in Jita. I was very pleased that this process was had no noticeable delay for a single item.
Raw aggregate data from Eve-Central is displayed on the left. My own calculations are shown on the right, with profit margin for station trading. But scrolling through lists of items isn’t a very good way to find trades! I added the find trades button to scan the item list for things with high margins and good volume.
That’s where I hit a bit of a block. Eve-Central doesn’t have data on the volume moved of market items, only the volume of orders listed. Unfortunately that’s not very useful information, since items with higher profit margins tend to move less and thus are not always very good trades. Without knowing how much of the product actually moves, it’s very difficult to determine if a trade is good or not.
Nevertheless, I had come this far. Pressing the “Find Good Trades!” button scans the list for items with profit margins above a certain threshold. Hoping that I could get some use out of the volume data available, I decided items with an equal quantity of buy and sell orders would be more likely to be good trades and created a ratio between the two to test.
Of course I was wrong. The analysis returns a list of terrible trades with good profit margins. But hey! I entertained myself for a day and now I can easily check prices in Jita without pulling up a browser.