But note that first you need a positive expectancy system, i.e. one that wins more than it loses. Otherwise you're mathematically destined to eventually lose, no matter what position size you use (Kelly returns a negative value).
Hence the first step is to improve the accuracy of your entries and exits to the point where you have high positive expectancy. Only then should you start looking at optimizing your position size (and addressing emotional issues like fear or greed).
Next, you must decide how consequential the loss of your entire account is. If your starting deposit is small enough that you can shrug off a margin call, then you can be "aggressive" in your risk taking; however, if safeguarding your current equity is a priority, then you need to be "conservative" (in which case you might want to also consider your risk of ruin/drawdown). The parameters "aggressive" and "conservative" are quantified in the Kelly article.
When you have a large enough track record to statistically establish your win rate and average payoff (RR), the Kelly criterion can help you to estimate your optimal position size. But you must also size small enough to stay well within your emotional tolerance (your 'comfort zone'). As the saying goes, scared money never wins.
