Over 600,000 people watched someone test a guru's A+ setup, which he says wins 70% of the time. We rebuilt it on Bitcoin and Ethereum 1-hour candles from 2020 to October 2026 with fees. It won 25% of trades on Bitcoin and 23% on Ethereum, and lost 82% and 92%. Verdict: fade it.
The viral claim
Top G, a gold day trader with daily livestreams and hundreds of thousands of followers, says his A+ setup "goes up to 70%, sometimes even 75". A video testing it passed 630,000 views. The rules sound simple enough to test, so we did.
The rules, and how we rebuilt them in Trigr
His rules as shown in his videos:
- Price breaks a recent high.
- Price comes back to the A+ zone, near the last low.
- A green candle closes there: buy.
- Stop under the candle, target the last high.
He skips some numbers, so we filled the gaps and say exactly how. We told the rules to Claude, connected to Trigr, and it built a fixed strategy.
| Setting | Our version |
|---|---|
| Markets | Bitcoin and Ethereum |
| Timeframe | 1-hour candles |
| A+ zone | Around the lowest low of the last 24 hours |
| Entry | A green candle that closes back above the zone, within 24 hours of the break |
| Stop | About 1% below entry on Bitcoin, 1.3% on Ethereum |
| Target | The highest high of the last 48 hours |
| Direction | Long and the mirror short |
| Test window | January 2020 to October 2026 |
| Costs | Fees included |
The results
| BTC and ETH, 1H, Jan 2020 to Oct 2026 | A+ setup | Just holding |
|---|---|---|
| Bitcoin return | -81.9% | +1,030% |
| Ethereum return | -91.8% | +1,726% |
| Win rate, Bitcoin | 25% | n/a |
| Win rate, Ethereum | 23% | n/a |
| Worst drop | -84% |
The win rate was not 70%. It was 25% on Bitcoin and 23% on Ethereum: a one-in-four setup. With a target at the last high and a tight stop, that is not enough to cover the losers and the fees.
Can we save it?
We tried five ideas on both coins.
| Idea | Bitcoin result |
|---|---|
| His setup, our reading | -81.9% |
| Same run before fees | -66.5% |
| Any green candle after a break, no A+ zone | -100% |
| Best fix: a fixed 3-to-1 target | -79.4% |
| Just holding Bitcoin | +1,030% |
Here is the surprise: the A+ zone is not useless. Buying any green candle after a break went to zero, so his zone is not the worst idea. It is just the second worst. Even before fees it lost about 67% on Bitcoin, and the best fix we found, a fixed 3-to-1 target, still lost 79%.
The honest part
The A+ zone is our reading of his video, not his exact chart. Zones drawn by eye cannot be tested exactly, so a trader who draws them differently may get a different result. Funding costs did not load for this test either, so the real number for perps could be different.
One point for him: in the last 25% of the data, from February 2025 to October 2026, the setup lost less than holding on Bitcoin (-9.6% vs -15.6%). On Ethereum it lost more (-17.5% vs -8.9%). That is not enough to change the verdict. If you still want to try it, paper trade it first. Read why backtests and live results differ before you put money on any setup like this.
Why high win-rate claims are hard to trust
A 70% win rate is a strong claim, and it is also easy to make. Win rates depend on how you count. If you move your stop after the fact, skip the trades you did not like, or only show the sessions that worked, the number goes up. A backtest cannot do any of that. It takes every signal the rules give, over years, and counts every loss.
The other problem is that win rate alone says little. A setup can win 70% of the time and still lose money if the losers are bigger than the winners, or if fees eat the small wins. On 1-hour crypto candles, fees and slippage add up fast, because the setup trades often. That is why we always report the return, the worst drop and the trade count next to the win rate.
When you see a high win-rate claim, ask three questions: over how many trades, with what fees, and compared with just holding the same market.
What we would test next
If you want to give the idea a fairer shot, three changes are worth testing: a wider stop based on recent volatility instead of a fixed percent, only taking longs when the daily trend is up, and trading it on gold, the market he actually trades, once enough clean intraday data is available. Each of those is a one-sentence change in Trigr, and each should be compared with just holding over the same window.
Verdict: trade it or fade it?
Fade it. It lost on both coins, even before fees, and the real win rate was about a third of the claim. A high win rate is easy to say on a livestream and hard to keep over thousands of trades.
Try it yourself
You can rebuild this setup, change the zone, stop or target, and backtest it with fees on Trigr. Start from the docs, and check your result against our overfitting warning signs.
Sources: his A+ claim on YouTube (1 May 2026 livestream) and his 8 Oct 2026 livestream.
Backtest, not live. BTC and ETH 1-hour candles, January 2020 to October 2026, fees included, funding not included. Past results are not a promise of future results, and this is not financial advice.