Use Phase 1 data to estimate Phase 2 success probability without fake precision. Learn how to combine strategy expectancy, setup quality, drawdown, opportunity rate, execution costs, market regime and behavioral stability into conservative probability ranges and scenario models.
Akash Mane is the Founder and CEO of Prop Firm Bridge, where he leads the company’s vision, platform growth, and long term strategic direction. He oversees operations across research, marketing, content systems, SEO, and product positioning while driving the platform’s mission of becoming a trusted authority in the prop firm industry. At Prop Firm Bridge, Akash plays a direct role in shaping educational frameworks, comparison systems, and trader focused resources designed to help users make informed decisions with transparency and confidence. His work focuses on building scalable organic growth systems, improving platform authority, and strengthening trust through accurate, structured, and search optimized content. In addition to leadership responsibilities, he actively manages growth strategy, social media marketing, search visibility, and brand development to expand the platform’s reach across global trading audiences.
Manoj Gholap is responsible for content accuracy, compliance, and factual integrity at Prop Firm Bridge. He acts as the final verification layer for all published content, ensuring that prop firm reviews, rules, and comparisons are clear, accurate, and aligned with transparency standards. Manoj plays a key role in maintaining trust and credibility across the platform.
Phase 1 gives traders something they did not have before the evaluation started: a live sample from the exact account environment. It contains real spreads, real platform behavior, real rule pressure, real entries, real mistakes and real emotional reactions. That information can make Phase 2 planning much stronger. The danger begins when the trader turns a small successful sample into an exact prediction.
A Phase 1 pass does not justify statements such as “I now have an 82% chance of passing Phase 2.” The sample may be too small, the market regime can change, the second target and rules can differ, and the future outcome sequence remains uncertain. The correct use of Phase 1 data is to update a probability range and identify the variables most likely to help or hurt the second stage.
This guide shows how to use Phase 1 as evidence without pretending it is a crystal ball. The framework combines broader strategy data with the live evaluation sample, then adjusts for setup quality, risk stability, drawdown path, execution cost, market regime, opportunity frequency and behavioral drift. The result is not one magical percentage. It is a set of realistic fast, base and stress scenarios that tell the trader whether the Phase 2 plan is robust.
Quick answer: Use Phase 1 data to update—not replace—your broader strategy expectations. Start with long-run win/loss and payoff ranges. Add Phase 1 live execution data, A-grade setup percentage, maximum drawdown, opportunity frequency, slippage, rule errors, overtrading or undertrading, and current market-regime match. Build optimistic, base and stress assumptions and estimate whether the account reaches the target before a failure boundary under each. Never treat a small Phase 1 win rate as a precise Phase 2 pass probability.
Written by Akash Mane, Founder and CEO of Prop Firm Bridge. This guide focuses on conservative probability updating rather than fake statistical precision.
Fact checked by Manoj Gholap. Account rules and market conditions vary. All examples are educational and should be rebuilt from the trader’s own tested data and exact current Phase 2 account.
For deeper survival modeling, see Phase 1 vs. Phase 2 Risk of Ruin Calculations. For live-data transfer, see How to Use Phase 1 Track Record to Optimize Phase 2 Performance.
Forecasting improves when new evidence is relevant, but one stage is still only one sample from an uncertain process.
The first-stage sample uses the same platform family, similar account rules, the trader’s current skill level and recent market conditions. That makes it highly relevant operationally. If the trader repeatedly made position-size errors, those errors matter for Phase 2. If live slippage was larger than backtest assumptions, that matters immediately.
This relevance makes Phase 1 useful even when the number of trades is small. Operational evidence does not require hundreds of trades to identify a repeated mistake. If the trader violated the session boundary three times in ten trades, the behavior deserves attention even though statistical expectancy is still uncertain.
The strongest Phase 1 evidence is therefore often about execution and behavior rather than exact market probabilities.
The stage ended when the trader reached the profit objective. That means the sample length was partly determined by success. A large final winner can stop the sample immediately. The equity path is not a random fixed-length experiment.
This creates selection effects. The observed Phase 1 win rate or average winner can be unusually favorable because the account happened to finish during a strong sequence. Another Phase 2 sample can start with losses and still belong to the same strategy distribution.
Do not treat the pass condition as proof that the recent statistics became permanent.
A trend strategy can complete Phase 1 during persistent direction and enter Phase 2 during a range. The relevant win rate, payoff and opportunity frequency can change.
This means a Phase 2 forecast must include a regime-match variable. A strong Phase 1 sample has less predictive value when the current environment is materially different from the environment that produced it.
The strategy can remain valid overall while the active regime is temporarily absent.
Some programs keep the same drawdown and timing conditions. Others change targets, minimum days or other variables. A smaller target can shorten the required path; a new consistency condition can lengthen it.
Forecasting must model the actual Phase 2 target and failure boundaries. The fact that Phase 1 was passed under one geometry does not automatically describe the next stage.
Probability belongs to the current account structure.
A trader might build a spreadsheet that outputs 73.4%. The decimal makes the answer feel scientific. If the input win probability is uncertain by ten percentage points and the average payoff changes with regime, the decimal is false precision.
Use ranges: perhaps favorable, base and stressed. The exact numbers should come from the trader’s own data. The goal is to know whether Phase 2 remains viable when assumptions are worse than hoped.
Robust ranges are more useful than a fragile point estimate.
If the model says the current risk is too fragile under stress, reduce risk before trading. If the model says the market regime is inactive, use observation mode. Do not use a high estimated pass probability as a reason to enlarge the next trade.
A success estimate describes the whole path. It does not increase the probability of the next individual setup.
Forecasting should improve the account plan, not become an entry signal.
A calibrated trader knows when confidence should rise and when uncertainty remains large. Phase 1 success can increase confidence in operational ability. It may modestly increase confidence in live strategy fit. It should not eliminate uncertainty about the next sequence.
Good forecasting therefore produces better questions: Is current R too large? Is opportunity frequency realistic? Is the strategy still in regime? Are behavioral errors improving?
These questions make Phase 2 more professional even if no exact probability is ever calculated.
Akash's research lens: I use Phase 1 to update confidence in specific variables. I never let one successful sample become a universal probability claim.
Book insight: The Signal and the Noise by Nate Silver is useful because forecasting improves through calibration, updated evidence and humility about uncertainty. Page: varies by edition.
The cleanest forecasting approach begins with what the trader believed before Phase 1, then updates that belief with live evidence.
Gather backtest, forward-test and prior live data using the same strategy rules. Remove data from clearly different strategy versions where possible. The goal is to estimate ranges for win probability, average winner, average loss, opportunity frequency and losing streaks.
This broader dataset is the prior because it existed before the current evaluation. It provides stability when Phase 1 contains only a handful of trades.
The prior should be relevant, not merely large. Thousands of trades from an old strategy can be less useful than hundreds from the current system.
If the strategy behaves differently in trend, range, expansion or compression, keep separate statistics. A single blended win rate can hide conditional behavior.
Phase 2 forecasting can then weight the regime that currently exists. If the market is in a state where the strategy historically underperforms, the probability range should become more conservative or the strategy can move into observation mode.
Conditional data is usually more useful than one universal average.
Do not use one exact historical percentage. Build a conservative range that reflects sample uncertainty and regime variability.
The stress case can use a lower win rate than the historical average. If the account survives only when the optimistic rate is used, the plan is fragile.
Probability modeling is most valuable when it is slightly pessimistic.
Theoretical targets can differ from actual outcomes because of partial exits, slippage, early closes and management. Use net realized R.
Keep a distribution rather than only the average. A trend strategy can have many small losses and a few large winners. A high-win-rate strategy can have smaller average winners.
The shape of the distribution affects path risk.
Estimate valid setups per week or session. Include quiet and active regimes.
This helps forecast how long Phase 2 could take without forcing activity.
A strategy can be profitable and still be a poor fit for an account with restrictive timing rules.
Record the longest observed sequence and stress beyond it. The future can exceed the historical maximum.
This input is essential for selecting R and estimating the probability of touching a failure boundary before the target.
Survival depth should be comfortably larger than the observed bad path.
This prevents hindsight. If the Phase 1 pass was unusually strong, the trader can compare it with the original expectation rather than rewriting the expectation after success.
A prior written in advance gives the update process integrity.
Phase 1 should move the forecast proportionally rather than replacing it emotionally.
Akash's research lens: My Phase 2 forecast begins with the broader strategy sample. Phase 1 is an update layer, not the entire statistical foundation.
Book insight: Thinking in Bets by Annie Duke is useful because beliefs should update as evidence arrives rather than flip completely after one outcome. Page: varies by edition.
A winning stage can still contain poor decisions. Setup quality tells the trader whether the sample actually represents the strategy they intend to repeat.
Use A, B and off-plan or another simple system. The grade should be based on regime, location, trigger, invalidation, reward room and session rules.
If a large winner was B-grade, it should not strengthen confidence in the A-grade system. If an A-grade trade lost, it should not weaken confidence solely because of the outcome.
Decision quality and result must be separated.
Divide A-grade trades by total trades. A high percentage suggests the Phase 1 sample closely represents the intended strategy. A low percentage means the pass may contain substantial behavioral noise.
Phase 2 forecasting should weight clean strategy trades more heavily than improvisations.
A passed stage with poor setup quality can deserve a conservative forecast.
Count trades that violated a rule but made money. These are especially dangerous because they can be remembered as proof of skill.
Remove them from the confidence update. Keep them in the behavioral-error update.
Phase 2 should not repeat a mistake simply because Phase 1 rewarded it.
Count valid A-grade losses. If the trader executed them correctly, they provide evidence of discipline under uncertainty.
This can improve confidence in the process even while P&L decreases.
Good forecasting values execution evidence as well as financial outcome.
A Phase 1 pass can hide undertrading. Count A-grade opportunities that were not taken and the reason.
If fear caused many skips, Phase 2 can be vulnerable to even stronger undertrading near funding.
Opportunity capture belongs in the success forecast.
Behavior can change with outcome. If Phase 1 off-plan trades cluster after losses, Phase 2 drawdown can activate the same pattern.
If they cluster after wins, Phase 1 success itself can increase Phase 2 risk.
Conditional error rates are more useful than one total error number.
Do not build a fake mathematical multiplier unless enough data exists. Conceptually, high-quality execution should increase confidence that Phase 1 reflects the repeatable edge, while low-quality execution should reduce it.
The trader can use qualitative labels such as strong, medium and weak evidence.
Forecast quality improves when the sample is filtered by decision quality.
Akash's research lens: I care more about how much of Phase 1 came from the real strategy than about the headline win rate of the stage.
Book insight: Black Box Thinking by Matthew Syed is useful because improvement requires separating process quality from flattering outcomes. Page: varies by edition.
Expectancy is useful, but a short live sample can make it swing wildly.
A simplified expectancy is win probability multiplied by average win minus loss probability multiplied by average loss. Use net R after costs.
Do this for the broader strategy sample and separately for Phase 1. Do not merge blindly.
The difference between the two can reveal execution or regime effects.
If Phase 1 winners were consistently smaller than expected, early exits or execution can be reducing the edge.
Phase 2 forecast should use the realized number unless the behavior is clearly corrected.
Do not assume theoretical targets will suddenly be achieved.
If losses exceed -1R because of slippage or stop movement, the account is more fragile than the backtest suggests.
Update the stress model with the live loss distribution.
Operational reality should improve the forecast.
If one trade produced most of Phase 1 profit, the average winner can be distorted upward.
Run the forecast with and without that trade. Ask whether the strategy remains viable if another outlier does not appear in Phase 2.
This sensitivity test reduces dependence on one memorable outcome.
Average can be misleading when returns are highly skewed. Record median winner and the range of outcomes.
Trend-following systems especially can rely on rare large wins.
The forecast should reflect the actual payoff shape.
A five-win, one-loss Phase 1 sample looks excellent but contains very little information about the true probability.
Use the broader prior and let the live sample adjust confidence modestly.
Small samples deserve wide uncertainty ranges.
If live execution is dramatically worse than historical expectations across multiple dimensions, investigate. If it is broadly consistent, confidence in account fit can increase.
The goal is not to discover a new permanent expectancy in ten trades.
It is to verify that live conditions have not destroyed the edge.
Akash's research lens: Phase 1 expectancy is a diagnostic. The larger strategy sample still carries most of the statistical weight unless the live sample becomes substantial.
Book insight: The Art of Statistics by David Spiegelhalter is useful because averages require context, sample size and distribution awareness. Page: varies by edition.
Success probability is partly the probability of surviving the path long enough to reach the target.
Measure the deepest peak-to-trough account decline and the maximum intraday stress if available.
Compare it with planned risk. Was the drawdown normal for the strategy or amplified by behavior?
Phase 2 should not assume the smoother Phase 1 path will repeat.
Use valid trades only when measuring strategy variance, and separately record behavioral losses.
Compare with broader history.
If Phase 1 experienced only two consecutive losses while history shows eight, the Phase 2 stress case should use the larger number.
Translate personal drawdown room into normal R. If Phase 2 can survive only four full losses and the strategy can experience eight, risk is too large.
Adjust R until the account has comfortable survival depth.
This is one of the most important probability levers the trader controls.
Losses can arrive in one session rather than evenly across days. Determine how many full-risk attempts the personal daily stop allows.
A strategy with many trades per day needs a daily portfolio model, not only a total drawdown model.
Phase 2 forecast should include first-boundary failure risk.
Several positions can lose together. Model a scenario where the largest correlated cluster hits stops at the same time.
If the account approaches the hard limit, reduce simultaneous risk.
Per-trade probability is not enough for portfolio survival.
Make the average loss slightly worse in the stress scenario. This captures execution uncertainty.
Do not rely on the stop price being exact in every event.
Robust forecasts include room for friction.
Run the conceptual model under optimistic, base and stress input sets. The exact probability can be difficult to estimate, but the direction of risk changes is still useful.
If reducing R dramatically improves survival under stress while preserving a realistic target path, the risk change is rational.
Forecasting should lead to safer planning.
Akash's research lens: My most important Phase 2 probability input is not the Phase 1 win rate. It is whether the account can survive the losing sequence the strategy is capable of producing.
Book insight: Against the Gods by Peter L. Bernstein is useful because survival depends on planning for adverse paths, not only expected outcomes. Page: varies by edition.
A strategy can have strong expectancy but too few opportunities for a particular account structure. Forecasting needs a time dimension.
Record all valid setups, not only those taken. This creates a live opportunity-frequency estimate.
Compare with broader history.
Do not assume a fast Phase 1 means the normal rate is high.
Record median and longest normal gaps. This helps build Phase 2 duration scenarios.
A low-frequency strategy should expect quiet days.
The forecast should not punish normal patience.
At the chosen money R, estimate how many net strategy units are required. Do not turn that number into a trade quota.
Combine it with opportunity frequency to estimate a broad time range.
Scenario planning is more honest than a deadline.
Use the favorable end of historical opportunity and outcome ranges.
The trader can see what quick completion might look like without increasing risk.
Fast should mean favorable sequence, not aggressive behavior.
Use median opportunity rate, average payoff range and normal losing events.
This becomes the central planning case.
The account should be comfortable if the base path takes longer than Phase 1.
Include a losing streak, fewer opportunities and normal waiting periods.
Check whether the account still has enough time and psychological tolerance.
Slow scenarios reduce the temptation to force progress when reality is quiet.
A target can be reached before the earliest completion condition. Add minimum days, profitable days, inactivity or maximum duration where they apply.
Use current official rules.
Probability of financial target completion is not always probability of formal stage completion.
Akash's research lens: I forecast Phase 2 in R and opportunity ranges, then add the account’s real timing rules. I never promise a date from a small Phase 1 sample.
Book insight: The Goal by Eliyahu M. Goldratt is useful because completion depends on whichever constraint remains active, not only on financial progress. Page: varies by edition.
The difference between backtest probability and live probability often appears in execution.
Record typical and stressed spreads during the actual trading window.
High-frequency strategies can lose meaningful expectancy through repeated spread cost.
Use live values in the Phase 2 forecast.
Convert commission to a fraction of planned risk. This makes cost comparable across account sizes.
If each trade pays 0.05R in friction, twenty trades consume one R before directional outcome.
Frequency and cost interact.
Compare planned entry and exit with actual fill. Separate normal and fast-market conditions.
Use a slightly worse stress assumption for Phase 2.
Do not model every stop as exactly -1R if live history says otherwise.
Wrong lot size, duplicate orders, wrong symbol or platform mistakes are not market variance. They are operational risk.
If Phase 1 contained such errors, Phase 2 success probability improves most by removing them through checklists and technical setup.
Preventable errors deserve more attention than tiny changes in estimated win rate.
Scalping and automation can be sensitive to execution environment. A system that relies on fast entry can degrade if latency or slippage changes.
Phase 2 should reverify the fresh account environment before normal risk.
Operational continuity is part of the forecast.
Longer holds can accumulate financing cost. Include it in net payoff.
Do not assume Phase 1 cost remains identical if rates or instrument conditions change.
Small recurring costs can matter near a tight target or drawdown.
Forecast with realized economics, not chart-only outcomes.
If the strategy remains comfortably positive under realistic costs, confidence increases.
If small cost changes erase the edge, the account-strategy fit is fragile.
Akash's research lens: Live Phase 1 data is most valuable when it tells me what my backtest could not: what execution actually costs in this environment.
Book insight: Market Wizards by Jack D. Schwager is useful because real trading results depend on execution and risk, not theoretical signals alone. Page: varies by edition.
The strongest Phase 1 sample becomes less predictive when the market environment changes materially.
Use the strategy’s own definitions for trend, range, volatility expansion, compression or another state.
Tag every trade and setup.
This shows which conditions produced the live results.
Use the same method. Do not change the regime definition because the recent outcome changed.
Compare current volatility, stop distance, spread, opportunity frequency and directional persistence.
Stable definitions make the comparison meaningful.
If Phase 2 begins in conditions similar to Phase 1 and broader historical data supports the edge there, the live sample is more relevant.
This does not guarantee another favorable sequence.
It simply makes the Phase 1 evidence more applicable.
If the system passed during a strong trend and Phase 2 is range-bound, the first-stage outcome should receive less forecasting weight.
Use broader regime-specific data.
Observation mode can be appropriate if the strategy is inactive.
Volatility and drawdown can justify immediate size changes. Core entry logic should change only after stronger evidence.
This keeps the forecast responsive without overfitting.
Risk is the fast adaptation layer.
A regime change can reduce or increase valid setups. Adjust time-to-target scenarios.
Do not hold Phase 2 to the Phase 1 calendar.
Forecasting must reflect current market opportunity.
The environment can change during the stage. Update the forecast when the inputs materially change.
A probability estimate is not a one-time number.
Good forecasting is a process of continuous calibration.
Akash's research lens: Phase 1 data predicts best when Phase 2 looks like the environment that produced it. I reduce its weight as the regime moves away.
Book insight: Thinking in Systems by Donella Meadows is useful because the same process can produce different outputs when the surrounding system state changes. Page: varies by edition.
Many evaluation failures are not pure strategy failures. The forecast should include the trader.
Record any breach, almost-breach or misunderstanding. A trader who repeatedly approached the daily-loss line had a fragile process even if the account survived.
Phase 2 forecast should improve only after a preventive control exists.
Lucky survival is not strong evidence.
Count B-grade trades, repeated re-entries and session extensions.
If these increase after losses or quiet days, include behavioral risk in the stress scenario.
A good strategy can fail through poor frequency control.
Count valid A-grade setups skipped from fear.
Phase 2 funding proximity can increase this behavior.
Low opportunity capture can make the target path much longer than the strategy data suggests.
Compare planned and actual R. Look for increases after wins, losses or near-target progress.
Unplanned size variation increases path risk.
Stable risk should raise confidence in repeatability.
Track early profit-taking, moved stops and held losers.
These behaviors change average win and loss, so the strategy forecast should use realized behavior until corrected.
Phase 2 planning must model the trader who actually trades, not the ideal version.
Record whether preparation, session boundaries and journaling were followed.
A stable routine reduces operational uncertainty.
Phase 1 should make Phase 2 easier through habit.
Rate process stability as strong, medium or weak. Keep it separate from market expectancy.
A strong edge with weak behavior can still have a poor evaluation outlook.
Forecasts become more useful when market and human risks are both visible.
Akash's research lens: My Phase 2 forecast has two engines: strategy edge and trader behavior. A strong first engine cannot fully compensate for a broken second one.
Book insight: The Daily Trading Coach by Brett Steenbarger is useful because repeated behavior patterns can materially change how a trading method is actually executed. Page: varies by edition.
Scenario modeling is the safest way to use probability without false precision.
Use the favorable but plausible end of win probability, payoff, opportunity frequency and execution cost ranges.
Do not increase risk in the optimistic case. The scenario describes what happens if the sequence is favorable under the same plan.
This gives an upper planning range.
Use median or central assumptions from the larger sample, adjusted by relevant Phase 1 live data.
Include normal losses and costs.
This should be the main planning scenario.
Lower win probability, reduce average winner slightly, increase average loss/cost modestly, reduce opportunity rate and include a normal-to-severe losing streak.
Ask whether the account survives.
Stress is where risk-size weaknesses become visible.
Model what happens if the trader takes extra trades after losses or cuts average winners near the target.
This can show that behavioral control matters more than a small improvement in market win rate.
Use the result to prioritize preventive rules.
Use statistics from a less favorable market state or observation mode.
If the strategy is inactive, the correct success path may involve waiting rather than trading.
Probability of immediate progress can fall while probability of account survival rises.
Run the same cases at normal and reduced R conceptually. See how survival depth and expected time change.
A smaller R can make the stress case much more robust while only moderately slowing the base case.
This trade-off is often more useful than chasing an exact pass percentage.
The trader can classify the Phase 2 outlook as strong, acceptable, fragile or observation-required based on the scenario set.
If enough data and modeling skill exist, numerical ranges can be used, but uncertainty should remain explicit.
A label with honest assumptions can be more useful than a false 73.42% probability.
Akash's research lens: I never ask one scenario to predict the future. I ask whether the same Phase 2 plan survives several plausible futures.
Book insight: The Signal and the Noise by Nate Silver is useful because good forecasts are distributions of possible outcomes, not confident stories about one path. Page: varies by edition.
The forecast should live in a simple dashboard and update only when evidence changes materially.
Record broader win/loss and payoff ranges.
This is the statistical foundation.
Do not change it after every trade.
Record A-grade percentage, realized net R, execution costs, drawdown and errors.
This is the live update layer.
Weight clean process more than lucky outcomes.
Label strong match, partial match or mismatch.
Update when market variables change.
This controls how relevant Phase 1 is.
Show personal drawdown room in R and stressed losing-streak capacity.
This is the main account-survival variable.
Update after significant P&L changes.
Track valid setups per session and expected waiting time.
Use it to update fast/base/slow duration scenarios.
Do not force the rate.
Track off-plan trades, skipped setups, size drift, session extensions and rule errors.
Behavior can improve or weaken the forecast.
Keep the score independent from P&L.
Track net commission, spread and slippage in R.
Update when platform conditions change.
High friction can reduce the edge.
Show target, days, consistency and hard boundaries.
Formal completion requires all applicable conditions.
Do not forecast from target alone.
Choose strong, acceptable, fragile or observation-required based on all inputs.
Write the reason.
The label should change only when evidence changes.
One outcome rarely changes the true probability dramatically. Review after a meaningful block or when market/account state changes sharply.
This prevents forecast chasing.
The dashboard should create stability.
A “strong” outlook does not justify a weak setup. A “fragile” outlook does not mean every valid setup should be skipped.
The strategy and account risk gates still control execution.
Forecasting belongs above the trade level.
Compare scenarios with what actually happened. Identify which assumptions were most wrong.
This improves future calibration.
Forecast skill itself should be reviewed.
Akash's research lens: My dashboard updates slowly. If one win changes the Phase 2 forecast dramatically, the model is probably reacting to noise.
Book insight: Superforecasting by Philip Tetlock and Dan Gardner is useful because strong forecasts improve through frequent but proportionate updates and post-outcome calibration. Page: varies by edition.
The final framework converts Phase 1 into useful evidence while preserving uncertainty.
Use relevant historical data for win probability, payoff, losing streak, opportunity rate and costs.
Use ranges.
This is the starting belief.
Separate A-grade trades from B-grade and off-plan trades.
Do not allow profitable mistakes to strengthen the strategy forecast.
Behavior is analyzed separately.
Use realized commission, spread, slippage and platform behavior.
These are high-value live observations.
Update net expectancy assumptions.
Record overtrading, undertrading, size drift, exits and rule errors.
Create specific preventive controls.
Do not assume Phase 2 behavior will improve automatically.
Verify target, daily loss, maximum drawdown, days, consistency and timing rules.
Calculate personal drawdown and R survival depth.
The forecast belongs to this exact stage.
Compare it with the Phase 1 and broader strategy regimes.
Adjust the relevance of the live sample.
Observation is allowed when the edge is inactive.
Use valid setups per session and target in R.
Create fast, base and slow paths.
Do not create a deadline.
Vary win rate, payoff, costs and opportunity conservatively.
Check account survival.
The plan should remain viable beyond the optimistic case.
Model the impact of known errors.
Then design controls to reduce them.
Behavior is a probability input.
Select R that lets the account survive plausible adverse paths while keeping the target achievable.
Do not choose risk from the desired pass probability.
Risk controls the path, not the market outcome.
Review weekly or after meaningful regime/account changes.
One trade should rarely rewrite the model.
Calibration beats reaction.
Phase 1 can make the Phase 2 outlook stronger or weaker. It cannot remove uncertainty.
The best forecast is one that improves decisions while remaining honest about what cannot be known.
Use the data to plan, then trade the next valid setup as an independent uncertain event.
Akash's research lens: The best Phase 1-based forecast is not the one with the most precise percentage. It is the one that makes the Phase 2 risk plan robust when the future is less favorable than expected.
Book insight: Superforecasting by Philip Tetlock and Dan Gardner captures the central idea: confidence should move with evidence, but good forecasters remain willing to be wrong and update. Page: varies by edition.
It can improve the forecast but cannot guarantee the result. A small Phase 1 sample contains uncertainty, and market regime, rules and outcome sequence can change.
Usually not by itself. Use the broader tested strategy range as the foundation and let Phase 1 update it modestly, especially when the live sample is small.
Setup quality, realized spread/slippage, average win/loss in R, maximum drawdown, opportunity rate, skipped trades, size drift, rule errors and behavior after wins/losses are highly useful.
An exact number usually requires assumptions that are too uncertain for most retail samples. Use optimistic, base and stress ranges rather than fake precision.
No. Phase 1 can contain favorable opportunity and sequencing. Phase 2 can have a different regime or losing sequence. Use fast, normal and slow scenarios.
No. A smooth first-stage path may not include the strategy’s normal worst streak. Stress-test beyond the Phase 1 drawdown using broader history.
Not automatically. Choose risk from Phase 2 drawdown survival, stop distance, correlation and stress scenarios. Strong recent results do not improve the probability of the next trade.
Phase 1 data is more relevant when Phase 2 begins in similar conditions. If the regime changes materially, rely more on broader regime-specific strategy data.
Update after meaningful blocks of data or real changes in market/account state, not after every win or loss. One trade usually contains too little information to justify a major probability change.
Use it to choose a robust risk plan, realistic time scenarios and observation/reduced-risk states. Never use a high estimated probability as an excuse for a weak or oversized trade.
Final takeaway: Phase 1 is valuable evidence, but it is not destiny. Use it to learn how the strategy actually behaves under evaluation pressure, how much execution costs, how the account draws down and how you behave after wins and losses. Combine that evidence with a larger strategy history, current market regime and exact Phase 2 rules. Build ranges, stress the assumptions and choose a risk plan that survives when the future is worse than the first stage. The purpose of prediction is better preparation, not greater certainty.
Prop Firm Bridge's Evaluation Mastery Center is designed to help traders turn live evaluation data into better risk and process decisions without pretending uncertain markets can be predicted perfectly.
It can improve the forecast but cannot guarantee the result because the Phase 1 sample, future regime, rules and outcome sequence all contain uncertainty.
Usually not by itself. Use a broader tested strategy range as the foundation and let Phase 1 update it proportionally.
Setup quality, realized costs, average win/loss, drawdown, opportunity rate, skipped trades, size drift, rule errors and behavioral stability are especially useful.
Most retail samples do not justify exact precision. Use optimistic, base and stress probability ranges and make the assumptions visible.
No. Market opportunity and outcome sequence can change. Use fast, normal and slow scenarios rather than assuming the first-stage pace repeats.
No. Stress-test the account against broader historical losing streaks and a worse path than Phase 1 experienced.
Not automatically. Choose risk from Phase 2 drawdown survival, technical stops, correlation and stress scenarios rather than recent success.
Phase 1 data is more relevant when current conditions are similar. If regime changes materially, broader regime-specific data deserves more weight.
Update after meaningful data blocks or real market/account changes, not after every individual win or loss.
Use it for risk planning, scenario timing and observation/reduced-risk decisions, never as a reason to take a weak or oversized trade.
