As we become more profitable, should we scale the number of contracts, or keep the position sizing relatively constant? This is the question that Dr. Data (Michael Rechenthin, PhD) wanted to solve.
So how does a data scientist approach a problem? He used a statistical approach -- the Monte Carlo simulation. The Monte Carlo method is when the problem involves a high degree of randomness and is very difficult when it is nearly impossible to use other "traditional" approaches.
He setup the experiment by comparing two groups "No Scaling of Positions" versus "Scaling of Positions." Results were found that on average, traders that scaled contract size as success occurred grew the account size at a faster rate. But with this came increased volatility of account returns.
A table was shown that broke down the results in a great amount of detail.