An apparatus and method of training helps separate emotions from decisions. A user account with single masked login allows selective access to a live account with actual funds and a demo account with false funds. A program randomly causes trading in the live and/or demo accounts according to a predetermined percentage, allowing the user to place trades/bets but without letting the user know which account is active. The present method further determines if trades/bets on the live and demo accounts are not offset/flat, and potentially forces the user to continue trading on the same accessed account. When the user logs out, the present method syncs the demo account to the live account, generates data concerning differences between live and demo accounts, and gives feedback to at least one of the user and a system controller.