Kelly makes sense when you can reliably quantify the odds of a bet (like in a card game for example). While kelly sizing will give you the fastest way to compound, it comes with a heavy price. In addition to the nerves of steel you need to face the required volatiliy, the penalty for misjudging and ending up to the right of the actual optimal size is basically ruin. My sugestion is to not even come close to half kelly. At least this way you will survive long enough to be able to judge if the odds you think you have are the odds you actually have. A common theme with trading strategies is that they lose their edge as more people become aware of them, so the right kelly size of yesterday is the overbet of tomorrow.
Is the Kelly Criterion useful for avoiding total blowups in writing iron condors? If so, how to calculate it? The online calculators as me for odds as in 10 to 1. If my shallowest delta is -95, would that be 20 to one in the odds calculator or one to 20?