Probably don't need years of data but a longer history is a corrective to the last twenty trading days. But once you set conditions on a price move - up 30, 40, 50, 75 bps before the close - you need a lot of data to fill the hoppers.
I come from a quant sales background - DRI, BARRA, StarMine - and sold stock selection models for the last ten years - Haugen model, IQS. I daytrade futures and I'm trying to get a grip on the time of day patterns I've mentioned.
Again, different people see things differently!
I returned to trading not so long ago, after a few years of time wasting, and, it has taken me a while to get back into the swing of things - you would be amazed how much you forget when you stop, but recently, I have purposely forgotten a good bit of what I had learned, as it was costing me money, so, stopped using it right away.
The best way to "see" what you need to see, is to have money on, and watch it go into the red and green - there is no better way to get the brain to "cop on".
Of course, never risk too much when conditioning the brain, as if one is silly enough to think that one can make thousands, before one makes hundreds, then Dan Brown comes to mind!
J_S