NilsTrades

Testing 100 TradingView indicators for statistical edge. Each one converted into a trading signal. No cherry-picking. No tweaking results. Identical conditions — same market, same timeframe, same rules. XAUUSD 15m • Last 10 years Posting 1 result every day for 100 days. 100 indicators. One leaderboard. The rankings start now. Spoiler: Some of the most popular indicators won't hold up.

SuperTrend's big brother? SuperTrend is currently leading my ranking of public TradingView indicators after 9/100 tests — standardized conditions, several thousand trades per entry signal, same market and rules across all of them. If you want SuperTrendy in the ranking, say the word, and I'll run it through the same framework. 💪

Introducing SuperTrendy Indicator! ✅ Confirms flips before firing - no more whipsaws ✅ Adapts to volatility automatically ✅ Grades every signal A, B, or C ✅ Tells you when a trend is overextended ✅ Fully open source Drop it on any chart, any timeframe. tradingview.com/script/jsZD85P…

The 2nd most used indicator on TradingView. Raw Entry Signal Test #8 Indicator: Squeeze Momentum Entry Signal: Squeeze release 10-year backtest — XAUUSD 15m 36088 trades Best Profit Factor: 1.06 (No TP | 200 EMA) Current rank: #6 (8 of 100 tested) This is the originally intended entry method. Testing a different approach tomorrow. Takeaway: Built-in direction. Built-in momentum. Still breakeven without the 200 EMA. Popularity ≠ edge.

60 trades doesn't really create an overfitting problem — it creates a sample size problem. At that count, the margin of error on your win rate alone is around ±13%. You can't tell if results are from the signal or randomness, let alone optimize anything. Getting to ~400+ trades first is more useful than any overfitting fix.

@MatiasScalbi What's the best way to prevent overfitting when optimizing a trading strategy? I'm using TradingView Pine Script with a 15-minute GOLD chart and only have 60 trades in my backtest.

GUARDEN ESTO Ahora con Claude Code se pueden crear estrategias de inversión en TradingView 🔥 Vas a poder crear lo que quieras con 0 conocimiento en programación Se conecta y hace toda la edición en PineScript solo Aca lo tienen directo para implementar github.com/tradesdontlie/…

@Fairy_Trades Agreed — and the hard part is most traders can't tell what edge they have until they've tested it properly. Most never do — so they're trading an assumed edge thinking it's a real one.

A lot your trading problems will disappear when you have an edge that you consistently see play out. Not an edge you assume you have Not an edge you’re assuming it works Not an edge that you rarely see Not an edge that is popularly said to work An edge you consistently find, and see it play out.

@Team2Trading Fair — but most traders haven't tracked enough trades to know what their A setup actually looks like in data. The classification exists in their head, not in their results.

@nilstrades Everyone knows when they are taking a B or C grade setup.

Your A+ setups make you profitable. Your B setups bring you back even. Your C setups turn you red. So technically you are already a profitable trader but you are just doing too much. Cut out all those shitty little B & C setups this month!

@SJosephBurns Discipline is only as good as what you're following. Most traders call it a system. Few have the data to prove it has edge.

A trader without the discipline to follow a system with an edge is just a gambler.

@jtrader Most traders who say they don't backtest are already doing it — just badly.

Do you backtest?

@zaruww "Study losses, copy wins" only works if you've defined what a win actually looks like before the trade. Without that, you're not copying a pattern — you're copying what felt right in hindsight.

1 system. 100 trades. Study losses. Copy wins. Repeat. You will Thank yourself in 5 years.

Leaderboard update — 7 of 100 tested. XAUUSD 15m — 10-year backtest — identical conditions. Raw indicator signal. No trade management. 1. SuperTrend — 1.22 (No TP | 200 EMA) — avg 1.05 2. EMA 20/50 — 1.19 (No TP | 200 EMA) — avg 1.03 3. QQE MOD — 1.16 (No TP | 200 EMA) — avg 1.04 4. SSL Hybrid — 1.12 (No TP | 200 EMA) — avg 1.05 5. MACD — 1.07 (No TP | 200 EMA) — avg 1.02 6. RSI Divergence (Regular) — 1.04 (No TP | 200 EMA) — avg 1.00 7. RSI Divergence (Hidden) — 1.03 (No TP | 200 EMA) — avg 0.99 Ranked by Profit Factor of strongest variant. Drop your indicator below. Public TradingView indicators only.

Raw Indicator Signal Test #7 Indicator: RSI Divergence Indicator Entry Signal: Hidden Divergence XAUUSD 15m Test period: Last 10 years (ending previous month) 16919 trades tested Best Profit Factor: 1.03 Takeaway: After the surprisingly bad result of Regular Divergence, I wanted to test Hidden Divergence. Only to find that it performs even worse. This might be due to the indicator identifying divergence 5(!) bars later than where it gets drawn. Current rank: #7 (7 of 100 tested)

@dangadofx Partial profits are a solid way to manage downside — especially in volatile markets. The break-even move is where it gets nuanced though. Whether it helps or hurts expectancy depends on the signal. Some entries need room to breathe past 1R.

My model is simple. I’ve used it consistently for years, targeting 3-4RR. No greed. I always secure partial profits because the market can reverse at any time. Once price hits 1R, I move to break-even to protect my capital. You don’t need a Lambo before December. That mindset will only make you watch profits turn red. Discipline over hype. Fix that💪

@tradertheory The stop loss moved because the exit rule wasn't defined before the trade. When exits are pre-defined and backtested, "tactical" stops being an option.

You moved the stop loss. You told yourself it was a tactical decision. You used the word *tactical* out loud, to no one. The trade is still open. The stop loss is still moving. Where did it all go wrong?

@bl_qee Nice. If the approach is working consistently, worth running it through a longer backtest to make sure it holds across different market conditions — not just the current phase. Short timeframes can look very consistent until they don't.

I did this TradingView premium not for backtest but so I could see charts in 30s, I’ve been so consistent with 1m and 3m decided to try the 30s.😂

@PatternProfits Win rate matters, but only relative to RR. A 70% win rate with a 0.3R average winner and 1R average loser still loses money. Profit Factor captures both in one number — cleaner primary metric for this reason.

With all due respect... if there's an inability to turn a 70% or better probability in the markets into a profitable trading strategy, one doesn't need to look deeper into the data; one needs to look deeper into the mirror. In the markets, it doesn't get much better. ✌️

Solid principle. The only thing I'd add: 50 trades is usually too small to draw conclusions. At that sample size you can't separate real edge from a lucky streak. I run 10-year backtests on XAUUSD 15m — that's typically 1000+ trades per signal. Results look very different if you compare chart history (a few hundred trades) with deep backtest (~10x more data and multiple market regimes).

Backtest any strategy 50 times on TradingView before you risk real money. If it doesn't work on historical charts it won't work live.

Raw Indicator Signal Test #6 Indicator: QQE MOD Entry Signal: Histogram turns blue/red XAUUSD 15m Test period: Last 10 years (ending previous month) 66300 trades tested. Best Profit Factor: 1.16 (No TP | 200 EMA) Takeaway: Decent performance with tons of entry signals. That means -> a lot of room for adding filters to improve the entry signal, and reduce drawdown. Current rank: #3 (6 of 100 tested)

Early observation after 5 tests: No TP + 200 EMA is the strongest variant in 5 out of 5 indicators. This means: - Letting trades run (no fixed target) adds more value than locking in profit early - 200 EMA filter consistently improves results (this does not mean that 200 EMA is the best trend filter option). One indicator showed no edge at all: RSI Divergence. Avg. PF 1.00 across all 6 variants — one of the most popular indicators on TradingView. Popularity ≠ edge. Insights based on XAUUSD, 15m — 2016/04–2026/03

No TP behavior No TP variant trades stay open until: - Stop Loss is hit, or - An opposite Entry Signal appears If the opposite Entry Signal is not suitable for exits, a Trailing ATR is added as an additional exit trigger — closing the trade when price closes more than 2× ATR below the previous candle close (for buys), or above it (for sells). The Trailing ATR exit level only moves in the trade's favor, never against it. An opposite Entry Signal also closes the trade — whichever comes first.

Indicator settings All indicators use their default settings. If an indicator does not produce an Entry Signal by itself, a 20-period EMA is added to generate a standardized cross Entry Signal.

Testing 100 TradingView indicators for statistical edge. Each one converted into a trading signal. No cherry-picking. No tweaking results. Identical conditions — same market, same timeframe, same rules. XAUUSD 15m • Last 10 years Posting 1 result every day for 100 days. 100 indicators. One leaderboard. The rankings start now. Spoiler: Some of the most popular indicators won't hold up.

Updated my multi-inside bar indicator with some nice features. tradingview.com/script/sZlYn2d…