Here's my lifetime earnings from roulette over 20 years since I turned 21. I've been to the casino at least once a month, usually twice or thrice. I tracked my earnings for the month (blue and left axis), and my cumulative earnings over time (orange and right axis). I have a bachelor's in math and a PhD in physics. So I understand the math behind the gamblers fallacy, but yet was able to consistently over 20 years earn money at a constant rate. This constant rate of growth of earnings is my constant betting strategy.

The longest recorded streak at a casino in roulette is 32 reds about 80 years ago. Hasn't happened since. At a casino on any normal day, the longest streaks I've seen have been 8, which are expected to occur extremely rarely. On exceptional days I'll see a streak of 12. Once I saw a streak of 13 but that was an anomaly.

My strategy was to bet the minimum on whatever I wanted, I didn't care. I was waiting for a pattern. Once I saw a pattern of six in a row, I upped my bet to about 100 on the opposite color. If I saw a streak of 8 I bet 500 on the opposite color. If it hit 9 in a row I let it be and just go back to waiting for another pattern.

I'd repeat this process for at least three hours throughout a night, usually waiting for about five if I'm on a hot streak. So, /sci/ how does your gamblers fallacy explain this?