Every result on this site answers one question about one strategy. Put all 36 tests together and a few patterns show up that matter more than any single verdict.
The scoreboard
Out of 36 strategy tests:
- 16 came out positive. They passed a challenge or finished clearly up.
- 12 were mixed. Usually a pass in one stretch and a loss in another, or a profit that depended on one good year.
- 7 failed.
- 1 was inconclusive.
That sounds better than you might expect. But how a strategy was tested changes the picture a lot.
Replays pass, code is harsher
Most of my early tests were manual replays: stepping through the chart candle by candle, usually for 20 to 100 trades. 14 of the 27 replay tests came out positive, about 52%.
The coded tests run the rules over years of data, often 5 to 10 years and hundreds or thousands of trades. Only 1 of the 7 coded tests came out positive: Justin Chuk's ORB. Three were mixed and three failed.
It's not that coding a strategy breaks it. A short replay can only show a good stretch or a bad one, and the strategies people make videos about are usually showing a good one. When the same rules run for years, the good stretch gets diluted by everything else.
The clearest example is Joovier Gems' London breakout:
- A 20-trade replay passed a practice challenge, up 16% with a 60% win rate.
- A 57-trade retest in 2026 won 35% and didn't pass.
- The 10.5-year coded test lost money in both versions. Only about 20% of simulated challenges passed.
Same idea, three very different answers. The longest test is the one to trust.
Claimed win rates were roughly cut in half
When a video gives a win rate, I check it. The gap is usually big:
| Strategy | Claimed | What I measured |
|---|---|---|
| Nico Trades ORB reversal | 80–85% | 42% |
| Simplified Trader 9:30 model | ~71% | 34% |
| Simplified Trader gold OTE | 70% | 33% |
A lower win rate isn't automatically bad. A 34% win rate can make money at 1:2 or better. But a strategy sold as a 70% winner that actually wins a third of the time will feel very different to trade, and most people quit it during the first losing streak.
Sometimes the creator's numbers are real, but cherry-picked
Some creators' numbers held up exactly, on the dates they chose:
- [Casper SMC's Short Session Method](/results/J2FEm3B8g4c): on his dates I got a 72.2% win rate and +$45,410, close to his own numbers. But over 5 years it made about $49,000 in total, and the four years before his window added only about $4,000.
- [Tim Hutter's EMA VWAP](/results/wTtYTJwUi9M): I matched his backtest to the dollar at $29,985. But most of the money came from trades held overnight, which the firm in question doesn't allow, and on his exact settings 2016 to 2022 lost about $31,000.
Neither of them made anything up. They showed the best window. That's why the question to ask is always "what about the years you didn't show?"
Warning signs I look for now
- A small sample. Under about 50 trades, a win rate can be 10 or more points off its true value.
- One recent window. If the results only cover the last 6 to 12 months, ask what happened before.
- A win rate above 70% with a normal target. Possible, but it's the claim that failed most often in my tests.
- Rules that need judgment. "Wait for a clean setup" can't be tested, and it lets every losing trade be explained away.
- No prop firm math. A strategy can be profitable and still fail most challenges if the drawdown is too big for the account. See prop firm drawdown rules.
So, do they work?
Some do, a little. The best ones I've found have a modest edge, a win rate a few points above their break-even line, and drawdowns you can survive at a sensible risk size. None of them are the "70% win rate, $10,000 a month" strategies the titles promise.
If you're about to trade a strategy from YouTube, look it up on the results page first. If it's not there, request it, or test it yourself using the way I backtest.