Winners under 0.6R, losers to -1.0R
The detector is a pair of habits, not a single click. Manually closing winners well before TP1 (realized R under 0.6R) while letting losers travel to the full stop inverts expectancy. Prospect theory predicted this decades ago: humans lock green to avoid regret and hold red to avoid crystallizing pain. On a prop book that pattern needs an unsustainable 72%+ win rate just to break even after commissions.
Worked tape: average winner +0.45R ($225) and average loser -1.00R (-$500). Fifty trades at a strong 60% win rate is 30 wins and 20 losses: $6,750 minus $10,000 equals -$3,250. The win rate looked healthy in the journal. The R multiple did not.
Why a 60% win rate still bleeds a 10% max-loss floor
Phase targets on 2-Step books are 8% to 10%. A -$3,250 hole on a $100,000 account is 3.25% of progress that must be re-earned with the same broken payoff. If overtrading (more than 10 trades/day) is how you try to make it back, the sample of early scratches grows. The 10% static max loss on an FTMO 2-Step is a slow bleed, not a single explosion, which is why this leak is under-diagnosed.
activate Venasri free on your MT5 account and let Manual Trade Guard own the exits. Partial volume at TP1 (1.5R), stop to breakeven plus buffer, runner to TP2/TP3 (3.0R-4.0R) on structural swings. That distribution is how a 54% engine still compounds.
Take the close button away from the amygdala
Connect the account on MT5 ZeroMQ Bridge Setup. Mobile closes from the iPhone app are the classic disposition click: you see green, you flatten the winner, you leave the loser. The server-side guard is what still trails the runner you would have killed.
Measure MFE, not how the close felt
Maximum Favorable Excursion on scratched winners tells you whether TP1 was reachable. If MFE regularly prints 1.5R and you exit at 0.4R, the location was fine and the close was the leak. If MFE never reaches 0.8R, the take-profit is too far and you have a strategy problem, not a disposition problem. Do not mix those diagnoses.
