A TradingView research and review routine: repeatable inputs

Assign a deliverable to each stage

A plan defines the rules; a checklist tests whether a decision meets them. A routine specifies when those checks happen and how unfinished work passes to the next session. Set the stages relative to your own trading window. For a continuously traded market, choose a session boundary and keep its timezone explicit.

Before the session: validate the inputs

Open the saved TradingView chart and compare it with the last session record. Check the exchange-qualified symbol, data feed, chart type, interval, regular or extended session, displayed timezone, and indicator inputs. TradingView exposes delayed-data information in the chart status line; a paid platform plan does not automatically include every exchange data subscription.

• Record the observation time and last completed bar. Distinguish a market that is closed from a chart whose updates are delayed or unavailable.
• Confirm that session settings match the research baseline. TradingView aligns intraday bars to trading sessions, so changed session coverage can change the bars you compare.
• Check relevant scheduled events against the exchange, issuer, or official release calendar and apply your existing event rule. Record the event timezone.
• Review carried positions and pending orders in the execution account. A clean chart does not prove that the account has no exposure.
• If the feed or settings differ, document the discrepancy and pause new entries that depend on those inputs. Manage existing exposure using the prewritten contingency process.

During the session: preserve decisions as they happen

For each observed setup, record the timestamp, rule version, rule checks, intended action, and what actually happened. Capture a chart image before later bars reveal the outcome when practical. An alert is an observation to investigate; an order submission and an executed fill need separate records.

• Use a short record such as: session ID; version; symbol; signal time and timezone; conditions passed or failed; order ID; fill or cancellation; reason for any deviation.
• Record skipped eligible setups separately from sessions with no eligible setup. Neither should be invented as an executed trade.
• If using script alerts, keep their creation context in the log. TradingView saves the script, inputs, symbol, and timeframe when the alert is created; later chart changes do not update that running alert.
• Queue research ideas for the weekly review. If a fault requires stopping the current version, record the stop time and reason instead of silently changing its parameters.

After the session: reconcile, then annotate

Match the journal to the account or simulator records before interpreting results. Check entry and exit fills, quantities, partial closes, fees, remaining positions, and pending orders. Mark whether each record is live, paper, or historical research; do not combine those environments into one performance claim.

• Retain the original plan note. Add a separate review note describing observed evidence, rule adherence, and unresolved questions.
• Annotate the chart with the signal, decision, and fill timestamps. Label retrospective annotations so they are not mistaken for notes made before the outcome.
• Label each executed trade compliant, deviated, or unknown under a rule definition written in advance. Missing evidence is unknown, not an automatic pass.
• Close the session with a carry-forward note: open exposure, unreconciled records, and the first check required next time. A no-trade session still gets a reason.

Worked review: a profitable deviation is still a deviation

This fictional week uses one unchanged version, V1, and four fully closed USD trades. Net results already include all assumed costs. Compliance means every required rule check passed and the trade followed that version. There are no unknown classifications in this example.

• Execution adherence is 3 compliant trades / 4 executed trades = 75%. Two skipped eligible setups and one session with no setup are tracked separately and do not enter this denominator.
• Net result is $18 - $12 + $8 + $14 = $28. Compliant trades contribute $14; the deviation contributes $14. Profit does not change the recorded rule classification.
• This sample does not show that deviations help or that adherence causes better returns. Four observations, differing opportunities, and selection effects prevent that conclusion.
• For a week with zero executions, report adherence as not applicable, not 0% or 100%. With unknown records, show their count and record completeness before interpreting adherence.

Weekly review: freeze the baseline and name the next test

Reconcile the week, group records by strategy version and trading environment, and inspect repeated deviations separately from losing trades that followed the rules. Decide whether the evidence supports retaining the baseline or designing a new test. A weekly review is a scheduling choice; one week is not automatically enough evidence to change parameters.

• Freeze the active version for the next observation period: source or indicator version, inputs, symbol universe, timeframe, session, entry and exit rules, and cost assumptions. Save its start date and review date.
• For example, keep V1 with a lookback of 20 while testing a V2 candidate with 30 in a separate research record. These are illustrative settings, not recommendations. State the test period and evaluation measure before comparing outcomes.
• Use a separate evaluation period before adopting a candidate selected from historical results. Keep V1 records intact and record when V2 becomes active. A changed market or sample can explain an apparent improvement.
• If an adopted version changes a script alert, delete the old alert and recreate it from the intended new context. Verify the new condition and notification settings.
• A freeze never requires continuing with broken data or a defective implementation. Suspend the affected process, record the exception, and use a new version identifier for corrected behavior.

The minimum routine when time is limited

Reduce research breadth before removing the records needed to reconstruct a decision. No fixed number of minutes suits every market, order frequency, or holding period. Use these completion conditions instead of a timer.

• If essential checks cannot be completed, defer new entries that require them. Do not treat limited time as proof that a shorter strategy is suitable.
• Use Pineify Trading Journal to enter trades, preserve plan and review notes, and associate trades with setups and tags. Give each strategy version a distinct setup name or tag, then filter records for comparison.
• Keep a session with no trade in a diary or session note. This workflow uses deliberate record entry; it does not assume automatic synchronization with TradingView or automatic enforcement of a parameter freeze.