Since you're day trading, the general approach I go with is to brainstorm what could happen and eliminate the unlikely situations (e.g an asset is unlikely to shoot up if there's a big sell wall that's larger than the daily trading volume). From then on you can balance your desired profit and risk; thinking of worst case and best case scenarios - people who go all in generally are not accounting well for the worst case.
For example:
>DGB could win or lose the Citibank competition.
>Bitcoin can either chain split (-> likely crash in value) into Bitcoin Unlimited and Segwit2x on August 1st, or avoid a chain split entirely (-> either stability or bull run because smooth upgrade).
From then on, your research would be assessing which outcome is more likely. E.g for the latter it would be the % of the network signaling for Segwit2x, the arguments behind anti-/pro- Segwit, and incentives behind different scalability proposals.
It doesn't sound like much, but its more sane than just trying to predict the future, plain gambling, or the buying the latest hype on the news or 4chan.