An evaluation is not really a trading test. It is a behavior test with a profit target attached.
Read the rules of any funded futures account and you will notice they barely care whether you are a good analyst. They care whether you keep your risk capped, whether you stop when you are told to stop, whether you show up on ordinary days, and whether your results came from a process rather than from one lucky morning. Those are all questions about self-control.
Which is convenient, because self-control is the one thing software has in unlimited supply.
Evaluations are lost to behavior far more often than to strategy. A machine removes the behavioral failures entirely and adds a small set of new, mechanical ones (timing, size, news windows) that you can eliminate with a checklist before the first trade. That trade is a very good one.
Strip the marketing away and every evaluation asks four questions.
Can you reach the target without blowing the floor? The profit target is the easy half. The maximum loss limit is the half that ends accounts.
Can you cap a single day? Daily loss limits exist because one uncontrolled session is how most retail accounts die.
Did the result come from a process? That is what consistency rules and minimum trading days are for. One enormous day does not prove anything, and firms know it.
Will you obey a rule you find inconvenient? Flat-by-close deadlines, position limits, news restrictions. Every one of them is a test of whether you do what the contract says on a day when you would rather not.
All of these are laid out individually in every funded account rule explained. Read that before this.
A system like the Algomyth engine trades one window a day, at 9:30 AM ET, and is asleep the rest of the session. That shape lines up with what an evaluation rewards almost by accident.
None of that makes an evaluation easy. It removes a specific set of failures: the ones that have nothing to do with whether the strategy works.
Automation swaps human failure modes for mechanical ones. These four are where they show up.
News restrictions. Some firms prohibit holding positions across high-impact scheduled releases. That matters here, because 8:30 AM ET is exactly when US economic data lands and the system trades an hour later, at the 9:30 AM ET cash open, as explained in how it works. That clears the release itself, but restricted windows vary by firm and some are wide. If a firm's news window reaches the opening bell, that firm is incompatible with this system. Settle it before you pay for anything.
Flat-by-close deadlines. Rarely a problem for a morning system, which is finished long before the deadline. It becomes one when a person adds a manual trade in the same account, or when the platform is left running unattended in a state nobody checked. Confirm the account is flat at the end of each session for the first week, until you have seen it happen with your own eyes.
Consistency rules. The awkward one. If a rule caps any single day at, say, 30% of total profit, one exceptional session can leave you unable to pass until enough ordinary days accumulate around it. A system that occasionally produces a large winner will meet this. Nothing is lost, but the timeline stretches, and people who did not expect it start doing something clever. Do not do something clever. Keep trading normally until the ratio evens out.
Contract size and scaling plans. The system trades the size you configured. It does not know your account has a scaling plan and will not read the rulebook for you. Setting size correctly is a manual step before session one. Standard sizing is 50k at one contract, 100k up to two, 250k up to five. Those are ceilings, not targets.
The full backtest, the worst drawdown and the losing months, plus a calculator for what each account size implies.
Uneventful, mostly, and slower than you expect. The system takes its window, wins some and loses others. It puts its target at three times the risk, so it breaks even at a 25% hit rate and is not built to be right most of the time. The money comes from the ratio, not from being right.
That means losing runs are ordinary. Four or five in a row is arithmetic on a system shaped like that, not a malfunction, and it is the point at which most people intervene and ruin an account that was fine. The worst drawdown in the whole record, January 2022 to September 2026, is $5,273, which is larger than the allowance on a typical 50k evaluation. Read that sentence twice, because it tells you honestly that accounts do get consumed.
Hypothetical performance. Those figures come from a backtest: a simulation over historical data, with no real money at risk. Simulated results are prepared with hindsight and cannot fully reflect real execution, slippage or liquidity. Past performance, real or simulated, does not guarantee future results.
It will happen, and the correct response is administrative rather than emotional. Pull the account's trade report and answer one question: did it fail on a rule or on a drawdown?
A rule failure is a configuration problem you can fix, whether it was the wrong size, a news window you did not check, a manual trade or a missed deadline. A drawdown failure is the model working as designed: the worst stretch in the published backtest is $5,273 deep, more than a typical 50k evaluation allows, so a repeat of it ends the account and the fee paid for it. Accounts are consumable. That is priced in from the start, and it is why the result comes from a bankroll of accounts rather than one account you must protect forever.
An evaluation you failed for a reason you can name is cheap. One you failed for a reason you cannot name is expensive twice.
Many do, and several allow trade copying across multiple accounts as well, but the policies differ and they change. Never assume from a forum post or a screenshot. Find the written rule on the firm's own terms page before buying, and if the wording is ambiguous, ask them directly and keep the answer. This is the single check most worth doing before you spend anything.
One, at the smallest size the plan allows. Running several in parallel is how the model works once a system has proven itself to you, but in the first month it just multiplies the cost of anything you have not learned yet. One account teaches the same lessons at the lowest tuition.
No, and the temptation is strong. Stopping near the target usually collides with a minimum trading days rule or a consistency cap, and it hands the decision back to exactly the instinct automation was meant to replace. Let it finish the way it started. There is a fuller list of these self-inflicted failures in the ten mistakes that ruin a good system.
The account converts to a funded account with its own rulebook, which is often gentler on drawdown and stricter around payouts. Read it as a separate document, because it is one. How the money actually reaches you (profit split, cadence, per-payout caps) is covered in withdrawing profits from a funded account.
Everything we write comes back to one idea: verify, then decide. The live terminal exists for exactly that.
Unlock Algomyth Live