Phase 1 vs Phase 2 failure rates explained without fake industry statistics. Learn why denominators matter, what current public datasets can and cannot prove, how stage-selection bias changes the numbers, and how to calculate your own Phase 1 and Phase 2 failure probabilities.
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.
Searches for “Phase 1 vs. Phase 2 failure rate” often produce confident percentages. One article says almost everyone fails Phase 1. Another says Phase 2 is where the real failures happen. A third gives an exact pass rate without explaining whether it counts accounts, people, first attempts, repeat attempts, challenge starts, Phase 2 entrants or funded outcomes.
The most important conclusion in 2026 is therefore not one dramatic percentage. There is no independently audited failure rate that covers the entire retail prop firm industry and cleanly compares every Phase 1 with every Phase 2. Public datasets exist, and some providers publish their own statistics, but the samples, products, definitions and denominators differ. A percentage can be mathematically correct for one dataset and still be misleading when presented as an industry-wide truth.
This article treats the title as a statistics lesson rather than a scare headline. It explains why Phase 1 often shows a lower pass rate among all starting accounts, why Phase 2 can show a higher conditional pass rate among traders who already survived Phase 1, why that does not automatically mean Phase 2 is “easier,” and how traders can build a personal stage-by-stage failure model using their own attempts, rule breaches, drawdown and process data.
Quick answer: There is no trustworthy universal statistic proving that Phase 1 or Phase 2 always has the higher failure rate across the entire prop firm industry. Phase 1 often filters a much larger starting population, while Phase 2 includes only traders/accounts that already passed the first stage. That selection effect can make the conditional Phase 2 pass rate look higher. To compare stages honestly, define the denominator, time period, account model, whether repeat attempts count, what “failure” means and whether the statistic measures accounts or unique traders. Use public data as a benchmark, not a guarantee.
Written by Akash Mane, Founder and CEO of Prop Firm Bridge. This guide focuses on statistical interpretation, denominator discipline and trader-level failure analysis instead of unsupported industry-wide claims.
Fact checked by Manoj Gholap. Prop firm datasets and rules vary by provider and time period. Any public benchmark should be interpreted within its published sample and definitions.
For the mathematical side of account survival, see the Phase 1 vs. Phase 2 risk-of-ruin guide. For behavioral stage differences, use the Phase 1 vs. Phase 2 skill comparison.
The industry contains different products, data systems and definitions. A single number would require a consistent population and methodology that does not currently exist across every provider.
Some two-step accounts use a larger Phase 1 target and a smaller Phase 2 target. Others use different minimum-day requirements, drawdown mechanics, news rules, consistency conditions or time limits. A failure rate from one structure cannot automatically describe another.
Even two accounts that both call themselves “2-Step” can expose traders to different failure probabilities because the maximum loss floor, daily-loss formula and target distance are different.
A statistical comparison must therefore define the product before it defines the percentage.
A transparent provider can publish meaningful statistics about its own evaluations. A data platform can publish a multi-account snapshot. Those numbers are useful, but they describe the accounts actually observed.
They do not automatically include every retail prop firm, every account type, every country or every strategy. The sample can be large and still not be universal.
Size improves precision inside a sample; it does not magically expand the population represented.
One trader can buy five challenges. If four fail and one passes, an account-level dataset shows a 20% pass rate for that trader’s attempts. A unique-trader statistic can say that the person eventually reached the next stage.
Both statements are true, but they answer different questions.
Whenever a pass or failure rate is shown, ask whether the unit is accounts, attempts, unique people or funded accounts.
A first-attempt pass rate can be low while the percentage of people who eventually pass after several attempts is much higher. If these measures are mixed, readers can believe two contradictory statistics.
Phase 1 is especially sensitive to this because every new account begins there. Phase 2 only appears after a first-stage pass.
A fair comparison needs a clear rule for how repeat attempts are counted.
Rules, product design, trader population and market conditions change. A 2023 dataset can behave differently from a 2026 dataset.
Even within one year, a provider can change target, drawdown or minimum-day conditions. Aggregating before and after a rule change can hide the effect.
Current statistics should include a date range and product version where possible.
Some accounts fail through daily loss, some through maximum drawdown, some through prohibited behavior, and some simply stop trading or expire under inactivity/time rules. Other accounts are abandoned by the trader without a formal market-loss breach.
Counting all of these as “failure” can be appropriate for completion statistics, but it does not explain why the account ended.
The reason category matters when comparing phases.
Akash's research lens: I do not trust a Phase 1 or Phase 2 percentage until I know the population, product, time period, unit of counting and exact failure definition.
Book insight: How to Lie with Statistics by Darrell Huff is useful because percentages can be technically correct while still misleading when the denominator or sample is hidden. Page: varies by edition.
The denominator is the group at the bottom of the fraction. It is the most important part of a pass-rate comparison.
Every two-step account generally begins in Phase 1. That means the Phase 1 population can include experienced traders, beginners, impulsive buyers, strategy testers, inactive accounts and repeat purchasers.
This broad population naturally contains many different levels of preparation. A raw Phase 1 pass rate therefore reflects both evaluation difficulty and the composition of people/accounts entering.
It should not be interpreted as the probability that a well-prepared individual trader must fail.
To enter Phase 2, the trader/account has already passed a filter. The population is selected. It contains accounts that reached the first target without violating the rules.
This selection can produce a higher conditional pass rate in Phase 2 even if the second stage still requires serious discipline.
The denominator changed from “everyone who started” to “those who already demonstrated enough to pass Phase 1.”
Suppose 20 out of 100 accounts reach Phase 2, and 10 of those 20 pass Phase 2. The Phase 2 conditional pass rate is 50% among Phase 2 entrants. But only 10% of the original 100 reached full completion.
If someone says “half of traders pass Phase 2,” they may be describing the conditional number. If someone says “only ten percent complete both phases,” they may describe the original cohort.
Both can be true. The denominator explains the apparent contradiction.
A trader who fails nine accounts and passes the tenth can count as a successful unique trader in an “ever passed” statistic. Account-level economics see ten attempts with one success.
Neither measure is inherently better. The first answers whether people eventually succeed; the second answers how often individual attempts succeed.
Readers need the label before drawing conclusions.
Payout statistics often start from funded accounts, not from challenge starts. A high payout rate among funded accounts can coexist with a low percentage of all initial challenge starts reaching a payout.
Do not compare a funded-stage percentage directly with a Phase 1 pass percentage unless the populations are aligned.
Funnels should preserve the same original cohort or clearly show conditional stages.
“Out of exactly which accounts or traders?” This question forces the source to reveal the denominator.
If the answer is unclear, treat the number as marketing or anecdotal rather than decision-grade evidence.
Statistical literacy protects traders from both exaggerated optimism and exaggerated fear.
Akash's research lens: My first question about every pass-rate statistic is not “What is the percentage?” It is “Percentage of what?”
Book insight: The Art of Statistics by David Spiegelhalter is useful because interpreting data correctly begins with understanding what is being measured and how uncertainty enters the result. Page: varies by edition.
Raw Phase 1 data can look harsh because the first stage is the widest part of the funnel.
Some people begin after months of testing. Others buy an account immediately after seeing a promotion or payout post. Some know the drawdown rules; others learn them after the account starts.
The first stage therefore absorbs many failures that come from preparation, rule misunderstanding and strategy mismatch.
Phase 2 automatically excludes accounts that failed those issues badly enough to never progress.
When Phase 1 has a larger target than Phase 2, the account may need more net favorable R or more trades to complete, assuming similar risk.
More trades can mean more opportunities for a losing streak, execution error or behavioral drift before completion.
This does not prove Phase 1 is universally harder, but target distance can affect the length of the path.
The trader may still be learning order entry, server time, commission, spread behavior and dashboard calculations.
Operational mistakes are more likely when the environment is new.
By Phase 2, many of those basic frictions are familiar.
Daily-loss calculations, maximum drawdown, news windows, holding conditions and minimum days can be misunderstood before the trader has practical experience.
An account can fail even with profitable market calls if the rules are violated.
Phase 2 entrants have at least demonstrated enough rule compliance to survive one stage.
A larger target can encourage traders to increase risk or trade more frequently because the distance feels large.
Some accounts fail not because the strategy lacked edge but because the risk wrapper became too aggressive.
The first-stage target therefore interacts with behavior as well as mathematics.
The percentage includes strategy, risk, rules, preparation, abandonment and trader behavior.
It cannot tell you what portion failed because they were “bad traders.”
Stage completion is a system outcome, not a pure intelligence or skill score.
Akash's research lens: Phase 1 is the broadest filter. Its raw failure rate mixes trader skill with preparation, product fit, rule knowledge, inactivity and risk behavior.
Book insight: Thinking in Systems by Donella Meadows is useful because outcomes emerge from many interacting parts rather than one simple cause. Page: varies by edition.
A higher Phase 2 pass rate among entrants can arise from selection and a smaller target, but the stage can still create unique risks.
They already demonstrated a minimum level of rule compliance, risk survival and target generation in Phase 1.
This makes the group different from the original challenge population.
Comparing raw percentages without acknowledging selection is statistically weak.
When the Phase 2 objective is lower, fewer net favorable outcomes may be needed at the same risk.
A shorter path can reduce exposure to losing sequences.
However, minimum-day rules or behavior changes can offset that advantage.
The trader usually knows how to place orders, calculate size and read the dashboard better than on Day 1 of Phase 1.
This learning can improve the conditional success rate.
Experience is a real difference between the two populations.
Recent success can create overconfidence, target impatience or fear of losing progress.
These pressures can increase risk or reduce participation even when the mathematical target is smaller.
A higher aggregate pass rate does not mean the stage has no behavioral difficulty.
Traders can believe the smaller target deserves a safer or faster system and abandon the method that passed Phase 1.
This breaks comparability and can create failure despite a favorable stage structure.
Conditional statistics cannot tell you which process each individual used.
Your own strategy frequency, risk, psychology and account rules determine the path.
Population statistics can provide context, but they do not replace personal risk-of-ruin and process analysis.
The correct question is how your edge fits the current Phase 2 constraints.
Akash's research lens: A higher Phase 2 conditional pass rate can be perfectly real while telling me very little about whether my next Phase 2 attempt will feel easy.
Book insight: The Signal and the Noise by Nate Silver is useful because probability needs context, conditioning and humility about what a dataset can actually predict. Page: varies by edition.
One failure-rate number hides very different endings.
The account ends because intraday equity or loss exceeds a defined limit.
This can happen from one oversized trade, several correlated positions or repeated losses.
It is a risk-path failure, not necessarily evidence that the underlying strategy lacks edge.
The account falls below the total loss floor.
The path can be gradual or sudden.
This category is especially important when comparing static and trailing structures.
The trader can be profitable and still violate a formal condition.
News, copy trading, platform behavior, position-size caps or other program rules can matter.
These failures should be separated from market-loss failures when analyzing skill.
Some accounts stop because the trader does not continue, loses interest, changes strategy or misses an inactivity condition.
Counting abandonment as failure is logical for funnel completion, but it has a different cause from drawdown.
Phase 1 may contain more abandonment because more casual attempts begin there.
Where a maximum duration exists, the account can fail without breaching drawdown.
Other programs may have no maximum time limit.
Mixing these products in one statistic can distort phase comparisons.
A trader can fail one account, learn, and pass another. The account failed; the person’s longer learning process did not necessarily fail.
This is another reason to keep attempt-level and unique-trader statistics separate.
Good analysis asks what unit is being judged.
Akash's research lens: My failure dashboard records the reason, not just a red yes/no. The fix for abandonment is different from the fix for oversized risk.
Book insight: Black Box Thinking by Matthew Syed is useful because improvement depends on classifying failure accurately instead of treating every bad outcome as the same problem. Page: varies by edition.
Phase 2 statistics describe survivors from Phase 1. That creates a built-in selection effect.
They passed the first stage. That achievement is itself a filter.
Traders who consistently oversize, ignore rules or abandon accounts are less likely to enter the Phase 2 sample.
This can improve the average characteristics of the second-stage population.
Not every Phase 1 pass reflects perfect skill. Some accounts can pass through a favorable sequence or one large winner.
Therefore Phase 2 entrants are selected by outcome plus process, not by process alone.
The survivor population is stronger on average but not perfectly screened.
If weaker or unprepared attempts are removed in Phase 1, the remaining accounts can naturally pass Phase 2 at a higher rate.
This does not prove the formal rules are easier.
It proves the population changed.
Every Phase 2 trader has recent success. That common experience can create confidence or pressure that Phase 1 starters did not all share.
The selected population therefore faces a new behavioral environment.
Statistics and psychology interact.
The cleanest display begins with one cohort of Phase 1 starts and shows how many reach Phase 2, how many complete Phase 2 and how many reach the next stage.
Conditional percentages can be shown alongside the funnel.
This lets readers understand both overall and stage-specific success.
It is a reason to interpret data correctly.
A conditional Phase 2 pass rate is useful when labeled accurately.
The problem begins only when it is presented as though the denominator never changed.
Akash's research lens: Phase 2 data describes a filtered population. I never compare it with Phase 1 as if both groups were drawn randomly from the same starting pool.
Book insight: The Book of Why by Judea Pearl and Dana Mackenzie is useful because observed differences between groups do not automatically reveal the cause of those differences. Page: varies by edition.
A percentage is less useful than the structural variables that produced it.
A larger target can require more net favorable R, all else equal.
That can increase time exposed to variance.
Do not assume the exact relationship is linear because payoff distributions differ.
If both phases use the same daily limit, the trader faces similar intraday boundaries.
If they differ, failure probabilities can change.
Use actual formulas, not headline percentages alone.
Static, trailing and end-of-day structures create different path dependence.
The same risk per trade can have different survival depth.
Risk-of-ruin analysis should be phase specific.
If both phases require the same number of days, the smaller Phase 2 target can still be constrained by the calendar.
Additional required activity after target achievement can create more exposure.
Completion is not always target-only.
A deadline can force a different opportunity problem than unlimited time.
An inactivity rule can affect low-frequency systems.
Stage statistics from different time structures should not be pooled casually.
More operational restrictions create more ways to fail outside pure P&L.
A simple target-and-drawdown account and a complex consistency model are not statistically equivalent products.
Structure should be analyzed before percentages are ranked.
Akash's research lens: I compare phase geometry before phase percentages: target, daily floor, maximum floor, days, time and rule complexity.
Book insight: Thinking in Systems by Donella Meadows is useful because changing system constraints changes outcomes even when the people inside the system look similar. Page: varies by edition.
Public benchmarks can be valuable when their scope is respected.
Look for sample size, provider coverage, dates, account types and whether results are live, self-reported or estimated.
A large sample with vague methodology can be less useful than a smaller transparent sample.
Method determines meaning.
Does “Phase 2 pass rate” mean percentage of Phase 2 entrants or percentage of original challenge starts?
These can differ dramatically.
The source should label the denominator.
Some datasets count every challenge separately. Others group people.
This changes both pass and failure rates.
Do not compare unlike units.
A dataset containing one-step, two-step, futures and CFD programs can produce an overall number that hides very different structures.
For this question, isolate two-step Phase 1 and Phase 2 where possible.
Product segmentation matters.
When multiple credible datasets differ, show the variation and explain why.
Do not average them mechanically unless the populations and methods are compatible.
Uncertainty is more honest than false precision.
Rules and trader behavior change. A benchmark should carry a date.
For evergreen education, explain the method so the article remains useful even when the latest percentage changes.
Statistics should be refreshable, not permanently embedded as truth.
Akash's research lens: I use public pass-rate data as context, never as a prophecy. The methodology gets more attention than the headline number.
Book insight: The Signal and the Noise by Nate Silver is useful because useful forecasting requires separating real information from overconfident interpretation. Page: varies by edition.
Personal statistics can be more actionable than industry averages when enough attempts exist.
Decide whether each purchased account is one attempt. Keep resets separate if they create a new evaluation path.
Use the same rule for both phases.
Consistency makes the comparison meaningful.
Divide Phase 1 passes by Phase 1 starts under the chosen definition.
Also calculate the failure rate and abandonment rate separately.
Do not hide inactive attempts inside market failures.
Divide Phase 2 passes by Phase 2 starts.
This tells you how often you complete the second stage after reaching it.
It does not describe the probability of full completion from the original start.
Divide full two-step completions by Phase 1 starts.
This is the personal attempt-level probability of completing both stages under the sample.
Keep the sample size visible.
For failed accounts, classify daily loss, maximum drawdown, rule breach, abandonment, strategy drift, overtrading and other relevant categories.
These percentages reveal what actually needs fixing.
A low Phase 2 pass rate caused by fear requires a different solution from a low rate caused by oversized risk.
Ten attempts are not enough to believe the observed percentage is your permanent true rate.
As sample size grows, the estimate becomes more stable.
Treat early personal statistics as evidence with uncertainty, not identity.
Akash's research lens: My personal funnel has three rates: Phase 1 pass per start, Phase 2 pass per Phase 2 entry, and full completion per original start.
Book insight: The Art of Statistics by David Spiegelhalter is useful because data interpretation should keep uncertainty visible, especially with small samples. Page: varies by edition.
The failure rate is the start of the analysis, not the end.
Look for position-size concentration, repeated same-day attempts and correlated exposure.
If Phase 1 has more daily-loss failures, the larger target may be triggering urgency.
If Phase 2 has more, post-success overconfidence may be relevant.
These can indicate risk too large for the strategy’s losing streak or failure to reduce risk during drawdown.
Compare the account path in R.
The stage label matters less than the survival depth.
Count whether off-plan trades appear more in one phase.
Phase 1 drift can come from target distance; Phase 2 drift can come from finish-line pressure.
Different stories can create the same behavior.
Some traders never breach but stop because the account feels too slow or stressful.
That is a failure of process fit or motivation, not market-loss risk.
Time and strategy frequency should be reviewed.
Track news, holding, minimum-day, consistency or other operational errors.
These should decline in Phase 2 because the trader is more familiar.
If they rise, familiarity may have become carelessness.
Oversizing gets a position-size cap. Overtrading gets an attempt limit. Rule errors get a checklist. Fear-based skips get opportunity-capture tracking.
Do not respond to a low pass rate with vague “more discipline.”
Statistics are valuable when they create specific process changes.
Akash's research lens: I care less about the percentage of failed accounts than the percentage of failures that came from each controllable cause.
Book insight: Black Box Thinking by Matthew Syed is useful because systems improve when failure is turned into precise feedback rather than shame. Page: varies by edition.
A dashboard keeps the comparison honest and easy to update.
Count all attempts under the same definition.
Keep date range and product type.
This is the main first-stage denominator.
Calculate pass and failure rate.
Separate abandonment.
Record median days where useful.
This should generally equal Phase 1 passes that actually activated the next stage.
If some accounts never begin Phase 2, record that separately.
The transition itself can be a drop-off point.
Calculate conditional pass rate from Phase 2 starts.
Keep the denominator visible next to the percentage.
Never show the rate alone.
Divide by original Phase 1 starts for full-funnel completion.
This is a different statistic from conditional Phase 2 success.
Show both.
Use consistent categories.
Compare which problems rise or fall after Phase 1.
This is the most actionable part of the dashboard.
Track both because extreme accounts can distort averages.
Compare in R and percentage where useful.
Drawdown path often explains failures better than final result.
Count off-plan trades divided by total trades.
Compare stages.
This shows behavioral drift.
Valid setups taken divided by valid setups available after account filters.
Low capture can identify fear.
High trade count with low valid-opportunity count can identify overtrading.
Count operational mistakes separately.
These should become less frequent with experience.
Use every error to improve the transition checklist.
Akash's research lens: My dashboard puts the denominator beside every percentage. A statistic without its population is not allowed on the page.
Book insight: Measure What Matters by John Doerr is useful because visible metrics are useful only when they are tied to clear definitions and objectives. Page: varies by edition.
The final framework gives traders a simple way to read any Phase 1 vs. Phase 2 statistic without being misled.
Is it a provider, platform dataset, review site, survey or anecdote?
Source quality affects interpretation.
Do not treat all percentages equally.
Accounts, attempts, people, Phase 2 entrants or funded accounts?
Write it beside the number.
If unknown, confidence drops sharply.
Two-step only or mixed account types?
Targets and drawdown structures matter.
Do not compare unlike evaluations.
Rules and markets change.
Use current data where possible.
Label older benchmarks clearly.
One trader can generate many accounts.
Know whether the dataset counts each separately.
This can change the headline rate substantially.
Drawdown, rules, inactivity, abandonment or all?
Cause classification matters.
Completion failure is broader than trading-loss failure.
Phase 2 pass among entrants is conditional.
Full completion among original starts is a funnel rate.
Do not swap them.
Target, drawdown, minimum days and time limits can explain some stage differences.
Statistics should be interpreted with account geometry.
Numbers without rules are incomplete.
Phase 2 entrants already passed Phase 1.
This makes the populations different.
A higher conditional rate can be expected without proving the stage is easier.
Benchmark your expectations, but do not use industry percentages to size trades.
Your account risk needs your own strategy data.
Population success does not control the next trade.
Track attempts, passes, causes, drawdown and behavior by stage.
Over time, personal data becomes more actionable.
Keep uncertainty visible with small samples.
The most useful statistic is often not which phase fails more. It is which controllable behavior ends your accounts most often.
Fix that behavior.
Then the personal failure rate can change regardless of the industry average.
Akash's research lens: The answer to “which phase fails more?” matters less than the method used to reach it. Good statistics should improve decisions, not create fear.
Book insight: The Art of Statistics by David Spiegelhalter is useful because responsible statistics communicate uncertainty, definitions and context instead of presenting numbers as magic truth. Page: varies by edition.
There is no independently audited universal industry-wide answer. Phase 1 often filters the full starting population, while Phase 2 contains only Phase 1 survivors, so conditional rates can differ substantially.
Phase 2 entrants are a selected group that already passed Phase 1. The second target can also be smaller in some models. This does not automatically mean Phase 2 is easy.
Only after checking the source, denominator, product, date range, repeat-attempt treatment and definition of failure.
It is the percentage of accounts that pass Phase 2 out of accounts that actually entered Phase 2, not out of all original challenge starts.
It is the percentage of original Phase 1 starts that eventually complete both stages under the chosen account-level definition.
One trader can make many attempts. Account-level data counts each attempt, while unique-trader data can count eventual success once.
No. The population is broader and includes more unprepared, inactive and rule-learning attempts. Structural difficulty is only one factor.
No. Population statistics should never determine position size. Risk must come from your account drawdown and strategy variance.
Phase 1 starts/passes, Phase 2 starts/passes, full completions, failure causes, drawdown, rule errors, setup drift and opportunity capture.
Always define the denominator and population before comparing percentages. Then focus on the controllable reasons your own accounts succeed or fail.
Final takeaway: The internet loves one number. Trading reality rarely gives one. Phase 1 and Phase 2 failure rates depend on who enters, which product they trade, how the rules work, how repeat attempts are counted and what the dataset calls failure. Phase 1 can show a harsher raw pass rate because everyone starts there. Phase 2 can show a higher conditional pass rate because only survivors enter. Neither statistic tells you what the next trade will do. Use data to understand the funnel, then use your own risk and process data to change the part you can control.
Prop Firm Bridge’s Evaluation Mastery Center focuses on transparent evaluation logic, clear definitions and process improvement instead of recycled failure-rate myths.
There is no independently audited universal industry-wide answer. Phase 1 usually contains the full starting population, while Phase 2 contains only accounts that already passed Phase 1.
Phase 2 entrants are a selected population and some models use a smaller second target. A higher conditional pass rate does not automatically mean the stage is easy.
Only after checking the source, denominator, product, date range, repeat-attempt treatment and exact definition of failure.
It is Phase 2 passes divided by accounts that actually entered Phase 2, not by all original challenge starts.
It is the percentage of original Phase 1 starts that complete both evaluation stages under a consistent attempt-level definition.
One person can make many account attempts. Account-level data counts each attempt while unique-trader data can count eventual success once.
No. Phase 1 also contains more unprepared, inactive and rule-learning attempts, so population composition affects the raw rate.
No. Population statistics should never determine position size. Risk must come from current drawdown capacity and strategy variance.
Track Phase 1 starts and passes, Phase 2 starts and passes, full completions, failure causes, drawdown, rule errors, setup drift and opportunity capture.
Always define the denominator and population first, then focus on the controllable reasons your own accounts succeed or fail.
