Cheers!

Without running the game many times, I’d argue that it should be at least possible to calculate the ratio based on the risk you’re willing to take e.g. set the ratio to a value that with 90% chance you don’t loose all your money. (there’s probably an analytical solution here or by running a monte carlo simulation). Agree, great follow up topic. Thanks for sharing.

Thanks Dries.
That’s a smart comment. And indeed, you are right!
You can basically “hack” the game and take advantage of restarting it. (If it was real money, of course, it wouldn’t be that easy. ;-))
If you can win big on the first bet that can take you much further on the long term because of the compounding effect.

Tomi

hey Matt,

Thanks a lot! And as of your question: my intended “message” with the game was, that all positive EV bets can be good.
So… If I were the one betting I’d put all positive EV bets. As a rule of thumb (to minimize the role of chance) I’d bet with 1/10 of my money all the time. But I think that’d be a more conservative strategy and far from optimal.

I’m sure that there’s a far more optimal betting strategy that can either be calculated using stats+math… or if there are too many unknown variables in it, using machine learning (an “AI player”) to figure that out. Only I never had the time for that. : )

Would be a nice second hobby project!

Tomi

PS. another commenter suggested a pretty nice “hack” that solves the question for this particular game. See below/above.

Really enjoyed this refresher on EV. I played with your application and wonder what’s the best strategy to both minimize the risk of ending up with 0 and maximizing the outcome e.g. what ratio of money do you allocate on the favorable bets?

Best, Matt

Nevertheless, nice introduction to people not familiar with the concept.

(PS: my current high score is 1317700908)

Thanks! Cheers! Tomi