3 Comments
User's avatar
Someone or other's avatar

Interesting. Looks like 2% is something of a sweet spot in that game: https://imgur.com/a/U0DYZUC Max consecutive losses was only 13, but max drawdown was a harrowing 80%. https://imgur.com/a/S50Q2Qn Still... 53,000% gains in 1,000 trades... wish this was more representative of the real market!

Expand full comment
Danny Merkel's avatar

An 80% drawdown doesn't sound like a sweet spot to me...

Expand full comment
Someone or other's avatar

Look at the final total. Even with a one-time 80% drawdown, $1000 ultimately turned into $500,000 in 1000 trades.

I tried it again, next time a consistent 2% bet wound up to over $200,000 in 1000 trades. (Sorry, didn't look at the max drawdown on that one.)

Just tried it again, this time 60% max drawdown, but over $546,000 balance on the 1000th trade: https://imgur.com/a/XrfNqRQ

I tried a couple times with 1%, 1.5%, and 3% and never got anything like the same result, not even within an order of magnitude.

Not really fair to say an electronic mockup is a proper test, though, since "random" algorithms never really are. They're complex enough to look random but if you happen to tap them just the right way, sometimes huge periodicity or other exploitable effects can appear. That could be this. A consistent 2% bet on this simulator might have different results than a consistent 2% bet doing it with a real life, well-shuffled deck of cards because of some subtle non-random factor in the "randomness" algorithm. But try it yourself... so far, for me, 2% has consistently produced huge drawdowns along the way, but nonetheless ended up several tens of thousands of percent at the 1000th trade, and none of the other values I tried have.

Expand full comment