Scouting is usually seen as the most important role as a League of Legends analyst. I pride myself on catering exactly to coach's needs, and automating my solutions as much as possible to dynamically build on each games preparation resources.
Usually deemed the least important form of scouting, SoloQ can give a small insight into an opponents champion pool and match up preferences.
With Bayes, I am able to gain second by second insight into an enemy's habits in game. This could be Level Ones, Back Pathing, Warding Patterns, Jungle/Support Proximity and many others.
All automated, to get the information as fast as possible.
Automatically generated level one animations can give a quick view into an opponents habits in the first 90s of a game.
It's also important to know what your opponent is doing before the game, as well as within. I use the Leaguepedia API to give insight to a teams drafting habits such as popular blinds picks per role, preferred counters and more.