Simulated Prop-Firm Gauntlet · Complete
Eval Arena
Every algo on the broadcast got its own simulated $50,000 evaluation account and had to earn its way out — pass the profit target without blowing the drawdown, or reset and try again. The run is complete; the final table is below.
Final Results
5 July – 18 September 2026 · run closedForty-three strategies ran the gauntlet for ten weeks and passed 88 of 326 attempts between them. That 27% is the number worth remembering: the typical strategy here failed roughly three evaluations for every one it passed, and four strategies never completed a single attempt at all. Prop-firm evaluations are hard, and an arena full of algorithms is not exempt from that.
One winner per session
How to read the table below
Small samples lie. With ten to twenty-four attempts apiece, the gap between a 46% strategy and a 37% one is not a ranking — it is noise. The three published above have overlapping confidence intervals, and several strategies further down the table are statistically indistinguishable from them too. They were chosen for having a decent rate and a real sample, one per session, not because the ordering is meaningful.
The top of the table is the trap. The strategy showing 100% passed a single attempt. Anything with fewer than about ten attempts tells you almost nothing, in either direction.
A pass rate is not an edge. What produces passes here is a modest geometric edge, hard discipline rules, and an account guard that stops the moment the target is banked. The same Donchian logic run on the Asian session instead of the US one passed 1 attempt in 17 — identical rules, different hours, near-total failure.
| Strategy | Attempts | Passes | Fails | Pass % |
|---|---|---|---|---|
| DisplacementFadeAsian | 1 | 1 | 0 | 100.0% |
| EMA200PBAN | 4 | 2 | 2 | 50.0% |
| LevelBreakC | 2 | 1 | 1 | 50.0% |
| VWAPReclaimCUE | 2 | 1 | 1 | 50.0% |
| VWAPReclaimCAN published | 13 | 6 | 7 | 46.2% |
| VWAPReclaimC | 11 | 5 | 6 | 45.5% |
| TrendPullbackB published | 20 | 9 | 11 | 45.0% |
| FVGEMAUS | 10 | 4 | 6 | 40.0% |
| DonchianBreakoutBUS published | 24 | 9 | 15 | 37.5% |
| EMA200PBEU | 8 | 3 | 5 | 37.5% |
| EMAPullbackAN | 8 | 3 | 5 | 37.5% |
| FVGEMAEU | 8 | 3 | 5 | 37.5% |
| IchimokuCloudB | 22 | 8 | 14 | 36.4% |
| EMAPullbackDeep20EU | 14 | 5 | 9 | 35.7% |
| FadeExtremeC | 16 | 5 | 11 | 31.2% |
| BBSqueezeA | 10 | 3 | 7 | 30.0% |
| EMA50Pullback | 7 | 2 | 5 | 28.6% |
| EMAPullbackEU | 15 | 4 | 11 | 26.7% |
| ADXTrendC | 8 | 2 | 6 | 25.0% |
| EMA200PBUS | 4 | 1 | 3 | 25.0% |
| ICTLiteC | 4 | 1 | 3 | 25.0% |
| EMAPullbackDeep20AN | 10 | 2 | 8 | 20.0% |
| FVGEMAAN | 11 | 2 | 9 | 18.2% |
| StochasticOBOSB | 19 | 3 | 16 | 15.8% |
| EMAPullbackDeep20US | 11 | 1 | 10 | 9.1% |
| EMAPullbackUS | 16 | 1 | 15 | 6.2% |
| DonchianBreakoutB | 17 | 1 | 16 | 5.9% |
| IchimokuCloudBUS | 6 | 0 | 6 | 0.0% |
| ADXTrendB | 5 | 0 | 5 | 0.0% |
| SweepReversalShortUS | 4 | 0 | 4 | 0.0% |
| SweepReversalShortMidday | 3 | 0 | 3 | 0.0% |
| TrendPullbackA | 3 | 0 | 3 | 0.0% |
| MACDHistogramC | 2 | 0 | 2 | 0.0% |
| PivotPointFadeC | 2 | 0 | 2 | 0.0% |
| TrendPullbackAUS | 2 | 0 | 2 | 0.0% |
| BBSqueezeW | 1 | 0 | 1 | 0.0% |
| InitialBalanceC | 1 | 0 | 1 | 0.0% |
| MomentumContA | 1 | 0 | 1 | 0.0% |
| ORBC | 1 | 0 | 1 | 0.0% |
| Confluence826Short | 0 | 0 | 0 | no attempts |
| GapFillB | 0 | 0 | 0 | no attempts |
| NYPMSilverBulletC | 0 | 0 | 0 | no attempts |
| PMTrendD | 0 | 0 | 0 | no attempts |
These results are simulated, and simulated results have inherent limitations. Every account, balance and outcome on this page was simulated — no real money was traded, and no real evaluation firm was involved. Fills came from NinjaTrader's simulator, which does not model slippage, queue position, partial fills, rejected orders or a thin book, and never misses a fill a real broker would have missed. Results of this kind are also produced with the benefit of hindsight over the same period used to measure them. No claim is made or implied that any account will achieve results like these.
Disclaimer: This page is published for education and entertainment. Nothing on it is financial advice, a recommendation to trade, or a solicitation; PhillyFrank is not a registered investment adviser, commodity trading advisor, or broker. Trading futures carries a substantial risk of loss, is not suitable for every investor, and you can lose more than you deposit. Past performance — simulated performance especially — does not indicate future results.
What is the Eval Arena?
Prop-firm evaluations are how real traders earn funded accounts: grow a $50,000 account by a set profit target without ever letting the balance fall below a liquidation line that trails your best equity. The Eval Arena runs that exact gauntlet — but the contestants are our trading algorithms, and the money is entirely simulated. Each strategy from the live broadcast is assigned its own arena account and trades it 23 hours a day, 5 days a week, under the same rules a human would face.
No cherry-picking, no do-overs: outcomes are scored by deterministic replay of the actual NQ price tape, so every strategy is judged against the same market. When a strategy passes, it goes on the record and starts a fresh attempt. When it blows the drawdown, that goes on the record too. Over time the arena tells us — and you — which strategies could actually survive a funded-account evaluation, and which just look good on a backtest.
Shadow Balance
The strategy's current simulated account balance on this attempt. It starts at $50,000 and every simulated fill, win or lose, moves it — exactly like a real eval account statement.
To Target
Dollars remaining between the shadow balance and the $53,000 pass line. When this hits zero, the strategy officially passes the attempt.
Pass / Fail Record
Completed attempts only. A pass means the target was reached; a fail means the trailing liquidation line was hit. In-progress attempts don't count until they finish.
Run These Strategies Yourself
The three session winners are free downloads in Philly Frank's Indicator Shed, as native NinjaTrader 8 files. Each one ships with a manual signal indicator as well as the strategy, because most funding companies will not let a robot trade their account — the indicator prices the entry, the stop and the target, and you place the orders.
Visit the Indicator ShedLast Live Snapshot
Loading snapshot...The arena has stopped running, so this table is a leftover from the live broadcast rather than the record. It may be out of date or unavailable. The final results are above.
| Strategy | Status | Attempt | Shadow Balance | To Target | Record (P-F) | Pass % |
|---|---|---|---|---|---|---|
| Loading... | ||||||
Disclaimer: The Eval Arena was a fully simulated exercise for education and entertainment. All accounts, balances and outcomes were simulated, nothing on this page is financial advice, and trading futures carries a substantial risk of loss. See the full risk notice under Final Results above.