If I may, a couple of things.
If after six weeks you are still interested in this approach, I suggest first that you begin to take a wider view of S&R. Picking a day at random to observe price movement in order to gain sensitivity to traders' behavior is one thing. But picking a day at random to develop a trading strategy is quite another. If you're going to get down to the strategy/tactics phase, you're going to make things far more difficult than they need to be by plopping yourself down in the middle of nowhere, particularly if you're going to rely on S&R. In order to succeed at this, you're going to have to develop a thorough understanding of trend, and to get that, you're going to have to back off to provide yourself with at least some context.
Therefore, I suggest that once you've selected your day you zoom out of it to see the entire day, using perhaps a 5m or 15m bar. Then include the next day and the day after that. Determine from those three days where the levels are that price regularly turns back. Determine also the trading ranges, if any (the two may be the same). If three aren't enough, do four or five or more. But look for the important S&R levels (if price happens to be trending during that period, there likely won't be any and you'll have to choose another set of days). Once you have them, project them forward into the day you're going to trade via replay rather than just start with no more than a brief sample of premkt data. You will find separating the important from the trivial to be much easier.
Second, now that you've "played" with this for a while, stop trading and go back to the observation stage, incorporating what you've learned about price movement into what you will now observe. Doing so will help you avoid sliding into the mechanization trap and perpetual tinkering. And what may have puzzled you before may now seem perfectly clear.