Guides and perspectives on trading automation.
Ideas, practical walkthroughs, and updates from PineConnector.
An R-multiple divides a trade's result by its initial risk, 1R. See the formula, worked examples, how to log R when stops move, expectancy in R and where TradingView, MT5 and PineConnector record the pieces.
Separate the value that sets a drawdown floor from the value checked for a breach. Compare static, trailing and daily rules, then apply FTMO's published daily reset to a worked example.
Support and resistance becomes testable only when every level has a reproducible definition. Define pivots, prior-day levels and round numbers, then write touches and breaks as closed-bar conditions.
Calculate position size from a loss budget, stop distance and pip value. Compare sizing methods, broker rounding and the documented TradingView and PineConnector volume options.
See what TradingView changed in Pine Script v6, which v5 code breaks or silently behaves differently, how the Pine Editor converter works and why existing alerts keep their old snapshot.
Understand ta.percentrank() with an explicit five-value comparison set, tie handling and a short Pine Script sketch. Separate relative rank, percentile values and alert events.
Learn exactly when ta.crossover(), ta.crossunder() and ta.cross() return true. A five-bar example shows touches, one-bar events and why realtime crossovers can disappear before close.
Understand how Pine strategy commands create simulated orders and how their fills generate alert events. Compare reversal sizing, exit reservations and broker-state limits.
Read higher-timeframe data with explicit timing. Compare lookahead modes, gaps and the documented previous-bar pattern using an hourly worked example.
A pivot has an origin bar and a later confirmation bar. See how ta.pivothigh() and ta.pivotlow() report values, why backward plots can mislead, and how alerts preserve the actual detection time.
Learn when nz() substitutes a numeric fallback and how na() and fixnan() differ. Worked sequences show why missing, zero and a carried-forward value need separate meanings.
Clamp values with math.min() and math.max(), including checks for missing inputs and reversed bounds. Worked examples separate clamping, rescaling and broker volume validation.
Separate confirmed bars from realtime updates, last bars and new bars. A worked price sequence shows why a temporary condition can trigger an intrabar alert but disappear before the close.
Compare alertcondition() and alert() by script type, message source and frequency. See why a true condition, a running alert and a broker order are separate events.
The payoff ratio is average win ÷ average loss. See a worked example, how it combines with win rate into expectancy and profit factor, the win/loss ratio naming mix-up and where TradingView and MT5 report it.
Overfitting tailors trading rules to a particular historical sample. Learn how candidate counts, parameter sensitivity and independent tests expose weaknesses in the selection process.
Out-of-sample testing checks frozen trading rules on observations excluded from their selection. Learn how to preserve a holdout, interpret its trade count and distinguish a historical simulation from broker execution.
What market, limit, stop and stop-limit orders do in MetaTrader 5, which side of the spread triggers each, and how long a pending order lasts. Includes a worked EURUSD example and the pending commands PineConnector documents.
Netting keeps one position per symbol; hedging permits separate positions from opening orders. See how closing deals, stop inheritance and reversal alerts change the result.
Compare MetaQuotes’ Windows support notes for builds 5320 and 6090, PineConnector’s separate EA requirements, and the checks needed before moving an older PC or VPS setup.
Understand the MT5 tester report by separating cash outcomes, trade averages and balance or equity drawdown. Worked examples explain common field mismatches.
Compare the five MT5 Strategy Tester modelling modes, their data assumptions and event streams. Learn where recorded ticks fall back to generated data and why tester delays and historical Forward tests are separate checks.
FOK, IOC, Return and BOC decide what happens to volume that cannot fill at once. See which MT5 execution modes allow each, how to read a symbol's filling modes, and what unsupported filling mode means.
MT5 build 6060 added MCP support, an AI Assistant with its own trading setting, passkey sign-in and two new account access levels. See what MetaQuotes documents, what it leaves open for EA-based automation, and what to check.
Builds 6140, 6180 and 6230 let the MT5 AI Assistant read Strategy Tester reports and logs, set up optimisation and manage programs on charts. Here is what each change means beside an EA acting on TradingView alerts, with a checklist.
