That is true if you have arbitrary knowledge that says that at some point in the future, Bitcoin is certain to crash. If you're able to associate some timefram

That is true if you have arbitrary knowledge that says that at some point in the future, Bitcoin is certain to crash. If you're able to associate some timeframe to the idea that with p = 1 Bitcoin will be at half the value it currently is, the Kelly criterion suggests you should be quite a bit of the farm. Applying the Kelly criterion in supposed p = 1 situations (e.g. "Like it or not, [the market will crash]") is pretty invalid.