10.000.000 bets equal running on Primedice for 10 months. Only simulation can do this.
Simulations are unrealistic, you're a professor you should know better if you posted something like this you should also post your own opinion not asking for others opinion, so people doubted you for posting they thought you are on something bad like selling script or borrowing money, people did not see the point of you posting this, because you are not telling us anything, very unlikely for a professor, you left us all speculating without giving your own reason for posting this.
One of the signs that it is unrealistic is the only 22 losing streak in a 10,000,000 bet. Many dice players got more than that in just thousands of bets. So with a real dice game, I think it will be more than 22 losing streaks.
The reason why I got a super profit this time is because I increase the base bet proportional to the balabce size. So it magnify by a geometric progression. I decreased the risk by reducing the profit rate. But by geometric progression, when it survives long enough the long run, an amazingly impossivle result came out.
Is this some kind of martingale strategy? Since you stated the bolded part, so I got to think that you are doing a martingale method on that simulation.