When I first began trading, I invested a small amount of $50 and made a profit of $20 in a slow run. This was my "aha" moment as a beginner. Because I am a self-learner, even a single penny profit is meaningful to me.
Every 5-6 years I have gotten an "aha" moment, as gotten far better than 44 years ago. Now it has gotten much easier of knowing what works for me, losing percentages have dropped under 5% for day trading and option trades. Holy Grails I thought I never find, have 6 of them, put decades into your skills pays off.
When I first started trading forex, I had no idea what I was getting myself into. At first, I had a lot of failures. But over time, I improved my trading techniques, continued to practise on my demo account, and learned from my failures, until finally I entered a trade and profited. So there was my "aha!" moment.
There have been a few but the most profound was when I learned how to use the Fibs, certain indicators and oscillators. The first few years were pretty rough and I lost a lot of money but here 19 years later I'm doing well. People that trade their "gut" usually don't last long in the market.
Aha! => 2 months later => Nope. Aha! => 2 months later => Nope. Aha! => 2 months later => Nope. Aha! => 2 months later => Nope. Aha! => 2 months later => Nope. etc. Still fun, though.
Who said an Algo is not reading a chart? This is a misunderstanding and makes my point even more clear. Thanks. Algo and chart reading are not exclusive To think so is tunnel vision, stuck in the box, can't see out, vision. The point is, people use single trade concepts when looking at the chart. An Algo can look at a chart in hundreds of way in a second. Then they can compare them and make decisions. For instance, on can make a system that follows a trend, and then has a sub system that looks for counter trends. Then each time it wants to put on a trend trade, it "consults" the sub system. Consistently, reliably, quickly. Humans cannot process that many evaluations per second over an entire day. They just look at their "next" trade, like a human from 1960 then concentrate on their orders and see it this next trade works out. An Algo considers 1000s of trades per seconds while a human considers a few or none per second. But bottom line IMO, it was an epiphany to realize, I can evaluate many setups, counter indicators each tick, and then make a decision. So many people get invested in their setup evaluations, then put on the trade, even if there were, or will be better ones, or the data has changed, BECAUSE they can only deal with a single trade evaluation at a time.
The big boys including, hedge funds use algorithms in their trading. If they have determined a stock is headed higher for various reasons----that is good enough for me. You see, I am just along for the ride. They can do the hard work and research for me. My job is merely to go along with the major trend. It does not need to be more complicated than that. Chances are pretty good I will make monies as the hedge funds, make their monies. It is all good for everyone, me included.
The combination of using a longer time frame (15 min not 5) and scale ins/outs gave me some degree of control over price and added confidence. Becoming more mechanical with exits checked the greed. Limiting funds in the account to what I actually needed to trade helped me to avoid doing something stupid. Walking away when you aren't in sync with the market preserves mental capital.
Not needing to "deal" with them, work for you. That is great. But times are a changing. Perhaps your view of Algos is incomplete. There are so many types of Algos it is hard to categorize them. It is so sophisticated that for decades now Algos battle Algos. OTOH, a retail trader can do Heuristic Algos with the simple scripting languages in MC, TS or NT. There was a quote in a book cira 2019, from the head of Algo at GS. It was something to the effect: "Eventually, 'Algo Trading' will no longer be called that. It will simply be called 'Trading'". IMO there are two sins for day trading. Thinking and Trading single ticket methods, like 1960. Using manual interfaces, because you can't write scripts, yet. Best of luck.