Most traders who keep a journal record their trades, but few can answer which setup actually pays, which strategy is degrading, or which mistake costs them the most. Trade tagging closes that gap by attaching structured labels to each execution so results can be filtered and measured. Done poorly, it produces a folder of inconsistent notes. Done well, it turns broker history into evidence. This guide covers how to build separate tag groups for setup, strategy, and mistake, how to keep them consistent, and how to analyze results by tag. The examples assume US stock and ETF day trading. SpiceLog is a trading journal and analytics tool built for US stock day traders, with particular depth for DAS Trader Pro users, and the examples here reflect how tags are structured inside it.
Table of Contents
- Quick Takeaways
- Why Setup, Strategy, and Mistake Need Separate Tags
- Defining Setup Tags You Can Audit
- Strategy Tags and What a Strategy Review Requires
- Mistake Tags: Building a Taxonomy You Can Count
- Importing Fills and Tagging Without Re-Entry
- Comparing Tagging Methods
- Analyzing Results by Tag
- Frequently Asked Questions
- References
Quick Takeaways
| Key Insight | Explanation |
|---|---|
| Separate setup, strategy, and mistake into distinct tag groups | A flat tag list cannot isolate whether the setup has edge or whether execution failed. |
| Define each setup by observable entry conditions | If two reviewers would tag the same trade differently, the tag cannot support analysis. |
| Keep mistake tags as a two-level taxonomy | A category plus a specific error lets counts roll up while preserving the dominant subtype. |
| Tag at the end of the session, not weeks later | Recall bias rewrites the reason for an entry, and impulsive trades get reclassified as planned. |
| Version strategies and setups when their rules change | Editing a definition in place makes earlier trades measured against a rulebook you no longer follow. |
| Read tag results in risk units and check sample size first | Raw dollar totals reflect position size, and thin samples produce false edges and false failures. |
| Use a controlled list, not free text | Variants such as chase and chased split one behavior into several tags with half the sample each. |
Why Setup, Strategy, and Mistake Need Separate Tags
A single trade carries three independent attributes. The setup is the pattern that triggered entry, such as a first pullback after a gap or a retest of a prior-day high. The strategy is the rule set that governed the trade, including position sizing, stop placement, and exit logic. The mistake tag records whether execution deviated from that rule set. Each attribute can vary without the others.
That independence is the reason to keep them apart. A trade can use a valid setup under a valid strategy and still be a mistake, for example by holding past a defined stop or doubling size after a loss. A losing trade with a mistake tag says nothing about whether the setup itself has edge. If all three share one flat list, you cannot filter for setup works, execution fails, which is one of the most useful diagnoses a journal should produce.
Where the Dimensions Overlap
Overlap happens at the boundaries. A strategy can be defined as a collection of setups, and one setup can appear under more than one strategy. In practice, keep setup as the entry trigger and strategy as the management and sizing framework. If a tag answers why you entered, it belongs to setup. If it answers which rules applied after entry, it belongs to strategy. A tag that answers whether you followed the rules belongs to mistake.
Why One Flat Tag List Fails
Flat lists grow by accretion. After a few months, a journal might hold tags like ORB, momentum, oversized, revenge, and news, all at the same level. Filtering by momentum returns trades that mix a valid setup with an impulsive exit, so the statistic describes neither. Separate groups force every trade to be labeled along each axis. The cost is a few seconds per trade. The benefit is that results can be decomposed instead of averaged into noise.
Defining Setup Tags You Can Audit
A setup tag is only useful if two people reading the journal would apply it to the same trades. That requires entry criteria written as observable conditions: price relative to a defined level, a volume or relative-strength condition that can be checked on the chart, or a time window. Vague labels such as good breakout or strong trend fail this test, because they depend on how the trader felt when tagging.
Writing Entry Criteria as Observable Conditions
Write each setup as a short checklist of conditions that exist before the order is placed. A first pullback setup, for example, might require a stock to hold above the opening range after a gap, a pullback on contracting volume, and a reclaim of the pullback high on a one-minute chart. Each condition should be verifiable from the chart at the time of entry. If a condition requires interpretation after the fact, such as looked extended, it will drift over time and corrupt the sample.
Keep the checklist short. Three to five conditions is enough for most day trading setups. Longer lists are rarely applied consistently, and a setup that requires a dozen simultaneous conditions often accumulates too few trades to evaluate.
Naming Conventions and Collisions
Use short, stable codes and avoid names that differ only by a modifier. Pullback and Pullback strong will be applied inconsistently, and the analysis will split one behavior into two tags. Prefer one code per setup, with any variant recorded as a separate attribute if it matters for the analysis. Review the setup list at a fixed interval and merge codes that are used interchangeably. A setup code applied to nearly every trade is no longer a setup. It is a default, and it should be redefined or retired.
Strategy Tags and What a Strategy Review Requires
A strategy tag identifies the rulebook that governed the trade after entry. For a day trader, that usually covers position sizing, stop placement, profit targets, and the time window in which the strategy is permitted to trade. Because strategy governs trade management, its results are most informative when measured in risk units rather than raw profit and loss. Two strategies with identical win rates can differ substantially in expectancy if one takes larger losses or has wider targets.
Strategy as the Rulebook Container
Define each strategy in a single document that lists its setups, sizing rule, maximum loss per trade, and exit conditions. The strategy tag should point to that document, not to memory of it. If you change a stop rule mid-quarter, create a new strategy tag rather than editing the old one, because otherwise the historical data is measured against rules you no longer follow. Versioning is cheap. Reconstructing which trades used which rule is not.
Running a Strategy Review from Tagged Data
A trading strategy review should answer four questions for each strategy tag: what the average result per trade is in risk units, what the distribution of outcomes looks like, how holding time differs between winners and losers, and whether performance varies with position size. Filter by strategy, then by setup, to see whether the aggregate result is carried by one setup or spread across several. In our experience, a strategy's weakness often sits in a single setup, and the aggregate hides it. Review at the setup level before changing the rulebook.
Mistake Tags: Building a Taxonomy You Can Count
A mistake tag records a deviation from the rules, and it has to be countable. Free-text notes such as bad entry or got greedy cannot be aggregated, and their wording drifts from one week to the next. A controlled list of error types, each with a fixed definition, produces a count per error that you can track over time and compare across strategies.
Two-Level Mistake Categories
Use two levels: a category and a specific error. A community discussion on trading journals suggests defining the top-level category first and then nesting specific errors beneath it, for instance Mistake, Entry, Chased under a broader entry category. This structure lets you roll counts up to entry errors for a monthly summary while still seeing which subtype dominates. Keep the leaf codes few enough to remember, typically a handful per category, so tagging stays fast during the session.
A mistake tag is only useful once it can be applied the same way on a Tuesday and on a Friday.
Separating Mistakes from Bad Outcomes
A mistake is a deviation from your rules, not a losing trade. A trade that followed every rule and lost money should carry no mistake tag. Conversely, a winning trade that broke a sizing rule still carries one. If you tag by outcome, the mistake category becomes a proxy for losses, and you lose the ability to identify behaviors that happen to succeed in the short term. Tag the deviation independently of the P&L.
Pro tip: Record mistake tags as soon as you close the trade, before you review the day's total. Tags applied after a red session skew toward the outcome, and you will over-count errors on losing days.
Importing Fills and Tagging Without Re-Entry
Manual entry is where most tagging systems fail. A trader who logs fills by hand skips trades on busy days, and the skipped trades are often the ones with the most mistakes. Importing execution data directly from the broker closes that gap, and tags are then applied to a complete record. For US stock and ETF day traders, the practical question is which import path is reliable for their platform.
Broker Imports as the Starting Point
SpiceLog imports trade fills from DAS Trader Pro, Schwab, Interactive Brokers, thinkorswim, and PropReports, and accepts generic CSV files for other sources. For a tag-based workflow, the main benefit is that fills from several brokerages can sit in one view, so a trader who executes across accounts can apply one tag vocabulary to all of them. Traders using DAS Trader Pro should confirm that their fill export carries timestamps precise enough for the time-of-day cuts described later. The trading journal is free.
Tagging in Batches Within the Session Day
Once fills are imported, tag in batches at the end of each session rather than trade by trade. Batch tagging is less disruptive during the session, but it should happen the same day. Waiting until the weekend introduces recall bias, and you will reconstruct a plausible reason for entries that were actually impulsive. A short fixed routine works best: review the session's fills, apply setup and strategy tags, then add mistake tags while the rules are still fresh.
Pro tip: Keep the tag vocabulary in the same place as your rulebook. Adding a new tag mid-week without updating the list is the most common source of inconsistent data.
Comparing Tagging Methods
Three approaches are common among active traders. They differ mainly in how much structure they impose on tags and how much work they require to maintain. The trade-off is consistency against flexibility. A free-form method captures nuance but cannot be aggregated. A rigid taxonomy aggregates cleanly but needs revision when strategies change. The grouped approach is the one that holds up as the number of setups grows, because its cost is paid up front and at each periodic review.
| Method | How It Works | Main Trade-off |
|---|---|---|
| Spreadsheet with free-text notes | Each trade is a row with a notes column, and filtering relies on keyword search. Many traders start with a trading journal template. | Notes cannot be counted reliably, and spelling variants split one behavior into several results. |
| Flat tag list in a journal | Trades carry any number of tags drawn from a single list with no grouping. | Tags on different axes mix together, so setup and mistake statistics cannot be separated. |
| Grouped taxonomy (setup, strategy, mistake) | Each trade receives one value per group from controlled lists, with mistake codes nested by category. | Requires upfront definitions and periodic maintenance, but supports decomposed analysis. |
| SpiceLog (ours) | Imports fills from supported brokers and applies grouped tags across consolidated accounts in one analytics view. Free. | Tag discipline still depends on the trader, and the tool reflects only the structure you define. |
SpiceLog is ours, so weigh the last row with that in mind. The grouped taxonomy is the method we built around, and the comparison above reflects that design choice rather than a neutral review of journals.
Analyzing Results by Tag
Tags pay off at the analysis stage. The most useful outputs are cross-tabulations: setup against mistake, strategy against time of day, and setup against planned risk. Read each cut in risk units per trade, not raw dollars, because raw dollars are dominated by position size.
Sample Size and Tag Stability
Small samples are the main source of false conclusions in tag analysis. A setup with a handful of trades can show a high win rate or a sharp loss by chance alone. Before acting on a tag's statistics, check how many trades it contains and whether the result holds across sub-periods. If a setup's apparent edge disappears when the sample is split by month, treat it as unproven. Retire rarely occurring tags, since they add clutter without enough data to judge.
Cross-Tabulating Setup Against Mistake
The most actionable view is setup against mistake. If a setup performs well on trades with no mistake tag but poorly on trades with an entry mistake, the setup has edge and execution is the problem. If it performs poorly even on clean trades, the setup definition needs to change. Only a grouped taxonomy produces this view cleanly, because a flat list mixes both cases into one average.
Before you assign size to a setup's trades, compute planned risk per trade with a position size calculator. That way the size tag reflects planned risk rather than whatever figure was entered at the time of the order.
Frequently Asked Questions
How many tags should one trade carry?
Assign one setup tag and one strategy tag per trade. Mistake tags can be multiple. If a trade fits two setups, record the trigger you acted on first, because two setup tags on the same trade double-count it in every setup statistic.
Should I tag trades I did not take?
Track them separately. Skipped setups are useful for checking whether your entry criteria are applied too strictly, but they have no fills and should not sit in the same statistics as executed trades. Keep them in a separate log that uses the same setup codes, so they can be compared without contaminating realized results.
What should I do when a setup definition changes?
Create a new setup code rather than redefining the old one. Historical trades were tagged against the earlier definition, and changing its meaning silently makes the time series incomparable. Record the change date so the analysis can be split at that point.
Is a spreadsheet enough for trade tagging?
A spreadsheet can work for a small number of trades if you enforce a controlled list with dropdown validation. It becomes fragile once you import from several brokerages, because reconciling fills by hand takes time and invites errors. A trading journal template is a reasonable starting point for building the habit before you move to automated imports.
How long should I wait before judging a tag?
There is no fixed duration. Judge a tag on its number of occurrences and on whether its behavior is consistent across sub-periods, not on the calendar. A tag with a thin sample across several months is still unproven. A tag with a sufficient sample over a shorter window can be assessed sooner, provided market conditions in that window were not unusual.
Which tag group has produced the most useful signal in your own records, and what did you change after you saw it?
References
Written by the team behind SpiceLog (https://www.spicelog.com/).
No comments:
Post a Comment