Hi,
This video explains the use of Montecarlo very well. https://www.youtube.com/watch?v=9S1Gslmgxcs .

Personally I use only "noise" (Randomize history data box) and variable shuffling (Randomize indicator parameters) with max 200 runs. If most of the lines is not above 0, probably it is a sign that you have bad strategy. Also if your original backtest is above all other lines, it means that probably you have overfitted strategy. For me I get almost half good looking systems fail these two Montecarlo test.

I made some experiment using montecarlo this way. It showed me some improvement on strategy portfolio if tested on unseen data chunk compared with not filtered strategies.

P.s. Mr. Popov, you can delete above link if you think it is spamming

All montecarlo tests randomizes something, and you can choose what you want to randomize for a purpose to find where your strategy breaks. The main thing what you want from these tests is that your strategies to be with profitable distribution (above 0). What my link talks about. Not much to explain, because all check boxes is self explanatory. Actually he shows and explains some in his youtube channels video, watch them all Definitely he shows montecarlo somewhere here https://www.youtube.com/playlist?list=P … ChI_XaESZP