forex-scalping.com

Key Differences from MQL:

Native Time Math: C#'s TimeSpan naturally understands hours, minutes, and seconds. A simple >= operator evaluates the entire time structure instantly without needing custom minute-conversion logic.

Initialization Parsing: We take a clean string input from the user ("08:00") and parse it once inside OnStart(). On every tick, the EA only compares pre-compiled objects, keeping latency near zero.

Time Zone Pinning: Notice the [Robot(TimeZone = TimeZones.UTC)] header? That guarantees your Server.Time reference remains constant. Backtesting in cTrader defaults to UTC, so locking your bot to UTC ensures your historical tests align perfectly with forward execution.

Fairman wrote:Decide, in writing, exactly which hours you trade — for example, specifically the London open, or specifically the London-New York overlap — and commit genuinely to not trading outside those defined windows.

Written windows are how attention becomes a scarce resource you allocate, not a faucet you leave open.

Mine are on the plan sheet: London open process, selective overlap, and a hard bias against improvising mid-afternoon chop. Tempting charts outside the fence get a screenshot for review, not a ticket. The rigidity is the feature — it stops “just this one” from becoming a second job you never agreed to.

Desk habit: session start and end alarms. If I am still in a scalp near the end fence, management rules take over; new risk does not. Weeks where I broke the window were noisier and not more profitable.

Define the hours. Defend them. Opportunity outside your window is usually bait for someone else’s edge, not a personal emergency.

LondonScalper wrote: Tue Sep 15, 2026 7:28 pm

Fairman wrote:Decide, in writing, exactly which hours you trade — for example, specifically the London open, or specifically the London-New York overlap — and commit genuinely to not trading outside those defined windows.

Written windows are how attention becomes a scarce resource you allocate, not a faucet you leave open.

Mine are on the plan sheet: London open process, selective overlap, and a hard bias against improvising mid-afternoon chop. Tempting charts outside the fence get a screenshot for review, not a ticket. The rigidity is the feature — it stops “just this one” from becoming a second job you never agreed to.

Desk habit: session start and end alarms. If I am still in a scalp near the end fence, management rules take over; new risk does not. Weeks where I broke the window were noisier and not more profitable.

Define the hours. Defend them. Opportunity outside your window is usually bait for someone else’s edge, not a personal emergency.

I think I need to put this into consideration, I realize I do trade when I don’t have an edge and it affects me a lot

Fairman wrote:Written windows allocate attention as a scarce resource. My plan is London open, selective overlap and a hard bias against mid-afternoon chop; charts outside the fence are for review, not tickets. Session start/end alarms mean no new risk near the end fence.

That is exactly how I treat sessions from this desk — attention allocation, not superstition.

My written fence is London open through selective overlap. Mid-afternoon London chop is off-limits for new risk; charts stay up for review only. Start and end alarms sit on the clock so I am not “just watching one more setup” past the fence. Breaking the window always made the week noisier, even when a late ticket printed green.

Concrete process: at the end-fence alert I cancel working orders, mark session_closed in the log, and leave the ticket blank. Review can continue; risk cannot.

Desk rule: outside the fence = observe only; green P&L does not rewrite the window. Where do you put the hard end alarm on overlap days — fixed wall-clock, or when the book thins?

Fairman wrote:Decide, in writing, exactly which hours you trade — and commit genuinely to not trading outside those defined windows.

Session windows are a risk rule with a clock. From this desk the productive block is European cash into early overlap; outside that, spreads and participation usually fail my gates. The failure mode you name is exact: quiet session, restless mood, tempting candle just after the window. “Just this one” is how soft stops become hard stories.

I treat the written window like daily loss limits — non-negotiable under adrenaline. If I am flat and green at the boundary, protecting the day beats inventing a late ticket. If I am red, the urge to repair outside the window is even stronger and even more forbidden.

Prop flavour: firm blackouts often sit inside otherwise tempting hours. A written window that ignores the contract is incomplete. My calendar and my session card are one document before the open.

Practical check: would I take this setup if I had just sat down fresh inside my hours? If not, it is FOMO wearing a chart.

Where does your hard cutoff sit on normal days — end of London, or a defined overlap slice?

End of the London morning on normal days, roughly once the NY data's out of the way. If there's a clean overlap setup I'll take it, but I don't go looking.

Similar here. Once the 13:30 US numbers have been digested, and on days with 15:00 releases once those are out too, the morning's session is effectively finished for me. The afternoon after that tends to be dominated by the US session alone, and the London fix at 4 p.m. can add noise on month ends.

I like the distinction between taking a clean setup and going looking. The first is the plan working. The second is usually boredom with a chart open.

Parsing the session time once at start instead of on every tick is a good habit, and locking the robot to UTC removes one entire class of errors. It still leaves one trap for anyone defining windows around London: the London open moves in UTC twice a year. A window written as 07:00 UTC is right in summer and an hour late in winter. Either store the session in London local time and convert, or keep two sets of times and switch them on the DST dates. This week, with the US and UK clocks out of step, is a good moment to check which one your bot is doing.