1. I typically start with the weights where I can enter my values. These weights are "trust" values. "How much do I trust certain sources?" These weights are all relative. If you trust Box Office Report 6 times more than Box Office Prophet then put a 6 for BO Report's weight and a 1 for BO Prophet's weight. If you don't trust Box Office Mojo at all, then enter a 0 for that weight and BO Mojo's data will be ignored. When you land on the page all of these weights are set to 1 by default. You can also click on the Moovee Picker banner in the upper left of the page.
2. The source of the weights are clickable. If the source is green, then the code has found values for it otherwise the source is blue and the values have not been found or the source is old and not during the game week.
3. Enter a weight below the expert's name and press TAB or leave the field. Changing this weight will trigger a recalculate of the page and the box office value. At the moment the URL will not change so you will not be able to share your page until you click the green button "Calculate Picks". This creates a weighted average of ALL the columns and fills in the numbers on the movie posters. Those are YOUR numbers.
4. If you choose to edit the numbers, the weights above will all go to 0 because you have now created custom values.
5. You can copy the URL for future reference if you want to edit your numbers later, but you don't have to. At the moment the URL will not change so you will not be able to share your page until you click the green button "Calculate Picks".
6. SHARE your picks on Twitter using the handy share buttons (with bonus or without). You will see a preview of your tweet and you will be able to edit it before you post.

• The weights that I currently use as a baseline and then maybe adjust from there. My Default Weights

Backlog (TODO list)

• Daily Historical Graphs
• Add URL strictly for Twitter image generation. (is the summary description cached on the Twitter side?)
• Rotten Tomatoes API?
• Dynamic help page to stay current with the UI.
• Find more data to aggregate. (Before the 10:00 am MST cutoff for picks.)
• Keep a history of data to figure out which "experts" are the best at what they do and then choose the weights accordingly. This would involve adding a database or file storage.
• Add a view for the simulation code which is already built. Asynchronous feedback would be needed here.
• If you're looking for what's BEEN DONE then look at the GitHub history (link above). You've probably noticed a few of these items have been REMOVED! Yay progress!