Question 1

What is Roast My Trades?

Accepted Answer

Roast My Trades is an AI-powered trade review tool for beginner traders. You upload a batch of closed trades and the AI reviews the quality of your decision-making process — not just whether you made or lost money. Every trade is rated Good, OK, or Bad based on how well you followed disciplined trading principles.

Question 2

Who is Roast My Trades for?

Accepted Answer

It's built for retail traders who are still learning — people trading stocks, spread bets, or CFDs with real or simulated money. If you've been trading for less than three years and suspect your process has more room to grow than your picks do, this tool is for you.

Question 3

Is this financial advice?

Accepted Answer

No. Roast My Trades reviews your trading process, not your portfolio. It does not make recommendations about what to buy, when to enter the market, or what positions to hold. Think of it as a coach reviewing your technique — not a broker advising your account.

Question 4

How is this different from a regular trade journal?

Accepted Answer

A trade journal records what you did. Roast My Trades tells you what your decisions reveal about your trading psychology. Where a journal shows you your history, this tool identifies the behavioural patterns — panic selling, cutting winners early, holding losers — that are silently eroding your returns.

Question 5

How does the AI analyse my trades?

Accepted Answer

Each trade is reviewed against key process criteria: entry timing, position size relative to your account, holding behaviour, and exit quality. The AI also retrieves historical market context for the dates involved — earnings announcements, CPI releases, central bank decisions, index performance — so your trade is assessed against what the market was actually doing at the time.

Question 6

What is a "basket"?

Accepted Answer

A basket is a group of trades submitted together for analysis — either a fresh upload or any selection from your trade library. The AI analyses each trade individually and also looks across the whole batch for recurring patterns — things like whether you consistently cut winners early or hold losers through support levels. A basket costs the same single analysis credit as one trade, however many trades you include.

Question 7

What does 'process quality' mean?

Accepted Answer

Process quality measures how well your trading decisions align with sound, repeatable risk management principles — regardless of whether the trade made money. A trade where you entered at a logical level, sized correctly, respected your stop, and exited at your target is a Good process trade even if it ended as a small loss. A trade where you entered on impulse, doubled your normal size, and got lucky is a Bad process trade even if it ended in a big win.

Question 8

Why focus on process rather than P&L?

Accepted Answer

Because short-term P&L is heavily influenced by luck. You can follow a terrible process and make money; you can follow a perfect process and lose. Over hundreds of trades, process is the only reliable predictor of long-term success. Optimising for process rather than outcome is how professional traders think — and what separates consistent performance from gambling.

Question 9

How long does an analysis take?

Accepted Answer

Most analyses complete within a couple of minutes. Larger baskets can take a few minutes longer, because the AI researches real market context for every trade date before writing its feedback. A progress overlay shows each step while you wait.

Question 10

What is the process score?

Accepted Answer

Alongside the Good/OK/Bad rating, every analysed trade gets a 0–100 process score judging decision quality against a rubric — not the outcome. A losing trade with a logical entry, correct sizing, and a respected stop can score 85; a lucky win on an oversized impulse entry scores low. Your average process score also drives the discipline score on your dashboard.

Question 11

Can I tell the AI what my plan was?

Accepted Answer

Yes. Manual entry and CSV import both accept an optional stop-loss, take-profit target, and a free-text "why you took it" note per trade. When a stop is present, the app computes your planned risk and realised R-multiple, and the AI grades how well you stuck to your own plan — a disciplined stop-out is credited, and an exit long before your own target is called out.

Question 12

What's the difference between Professional and Roast modes?

Accepted Answer

Both modes run the same underlying analysis and produce the same trade ratings. Professional mode delivers feedback as structured, metric-led commentary: win rate, average risk-to-reward, specific action points — plus, for OK and Bad trades, a counterfactual breaking down why it went wrong, what you could have done instead, and what that alternative was worth. Basket analyses also get 2–3 prioritised improvements with estimated P&L impact. Roast mode delivers the same insights with humour and exaggeration — designed to make your mistakes memorable enough that you actually change your behaviour.

Question 13

How do I upload my trades?

Accepted Answer

You can import a file exported from your broker's platform, or enter trades one at a time manually. The standard importer handles CSV files with the most common column naming conventions from major UK and US brokers, plus MetaTrader 4 and MetaTrader 5 account history statements exactly as MT exports them, and a downloadable CSV template is provided if you prefer to fill one in yourself. There is also an example gallery of pre-analysed trades and baskets so you can browse full analyses — including both tones — before uploading anything of your own.

Question 14

Which file formats are supported?

Accepted Answer

The standard (free, rule-based) importer takes CSV files in two shapes: complete-trade format (one row per closed position, with entry and exit data in the same row) and ledger format (individual transaction rows matched into round-trip trades). Common column names from IG, Trading 212, Spreadex, Hargreaves Lansdown, and similar platforms are recognised automatically — including optional Stop Loss, Take Profit, and Notes columns. MetaTrader 4 and MetaTrader 5 statements are read directly too: export your Account History as a report and upload the .htm, .html or .xlsx file as-is. Your stop and target come across with each trade, cash entries like deposits and credits are skipped, and lot sizes are converted so the P&L matches what MT shows you. If your statement is awkward, an opt-in "Use AI to process upload" toggle sends the file to the AI instead, which can extract closed trades from CSV, Excel (.xlsx/.xls), HTML, and PDF broker statements.

Question 15

What markets and instruments are supported?

Accepted Answer

Any stock, ETF, index, or commodity traded as a direct equity, CFD, or spread bet. As long as you have entry price, exit price, position size, and trade dates, the AI can analyse the trade.

Question 16

Can I review trades that are still open?

Accepted Answer

Not directly — analysis requires closed trades with both entry and exit data. If your CSV includes open positions, you'll be prompted to either close them manually (by entering an exit price and date) or exclude them before submitting the batch.

Question 17

What behavioural patterns does it detect?

Accepted Answer

The AI detects seven key patterns common in beginner traders: panic sell (closing early out of fear before your plan triggered an exit), FOMO buy (entering late after a move, chasing price), revenge trading (aggressive trades to recover recent losses), holding losers (refusing to close a losing trade), cutting winners (closing profitable trades too soon), oversized position (trading too large for your account), and DCA sequence (averaging down without a pre-planned strategy).

Question 18

Does the AI have access to live market data?

Accepted Answer

The AI uses search grounding to retrieve historical market context relevant to your trade dates — earnings announcements, economic data releases, central bank decisions, and index performance. It does not have access to real-time prices or your live portfolio.

Question 19

How accurate is the AI analysis?

Accepted Answer

The AI is strong at identifying structural process problems — position sizing, holding patterns, entry timing relative to market events. It's most reliable for retail traders in major markets (US and UK equities, major indices, popular commodities). For highly niche or illiquid instruments, the market context retrieval is less comprehensive. Behavioural patterns are additionally grounded against the historical price path around each trade where available, and are labelled "possible patterns" in the app because they remain AI inferences about your intent.

Question 20

What are MFE, MAE, and capture ratio?

Accepted Answer

Where historical price data is available for your instrument, each trade gets deterministic exit-quality metrics: Maximum Favourable Excursion (the best the trade ever looked while open), Maximum Adverse Excursion (the worst it looked), and a capture ratio — the share of the best available move you actually banked. These are computed from real daily price data, not estimated by the AI, and they turn "you cut that winner early" from an opinion into a number.

Question 21

What is the Trading Archetype?

Accepted Answer

Once you have five or more analysed trades, the dashboard rolls your dominant behavioural patterns into a persona — The Paper Hands, The Revenge Addict, The Peak Buyer, The Bag Holder, The Averager, or The Disciplined One — with a discipline score, the pattern that drove the label, and one actionable rule to break the habit. It is computed from your existing analyses, so it costs no credits, and it is framed as a read of your patterns, not a verdict.

Question 22

Can I share my results?

Accepted Answer

Yes. Every analysed trade, basket, and archetype can generate a branded 1080×1080 share card — the rating, headline, and key stats on a single image. On mobile it opens the native share sheet; elsewhere it downloads as a PNG.

Question 23

Is my trading data private?

Accepted Answer

Yes. Your trade data is stored in your own account and is not shared with other users or used to train AI models.

Question 24

Do I need a Google account?

Accepted Answer

No. You can sign in with Google or create an account with an email address and password — both give you the same product.

Question 25

Is Roast My Trades free?

Accepted Answer

The free tier includes 3 analyses — enough to see what the AI catches in your trading. The Pro plan (£16.99/month, or £12.99/month billed annually) removes the limit entirely; every other feature is identical on both tiers. No credit card required to get started, and Pro can be cancelled anytime from the billing portal.

Question 26

How many trades can I include in one analysis basket?

Accepted Answer

There's no hard cap, and a basket costs the same single credit however many trades it contains. Batches of up to 30–50 trades work best — larger batches can be submitted, but per-trade feedback becomes more concise as the batch grows.

Frequently asked questions

About Roast My Trades

How the analysis works

Uploading trades

AI & market context

Privacy & account

Still have questions?