Multi-Strategy Backtesting Engine for Crypto
Simpleton v0.01 CURSOR is a TradingView strategy() script with 12 strategies (11 individual + 1 consensus mode). It generates Buy/Sell signals with strength levels 1–5, has toggleable TP/SL, and is non-repainting by default.
| File | Type | Purpose |
|---|---|---|
Simpletonv0.01_CURSOR.pine | Pine Script v6 Strategy | Main TradingView tool — add to chart to backtest |
backtest_cursor.py | Python Script | Batch backtesting across symbols + timeframes |
simpleton-cursor-quickstart.html | HTML Documentation | This guide |
Simpletonv0.01_CURSOR.pine into the editor| Level | Name | Conditions | Position Sizing |
|---|---|---|---|
| L1 | Weak | Strategy fires alone — no confirmations | Paper trade only |
| L2 | Moderate | + HTF 4H alignment OR volume > 1.5x avg | Small position (25%) |
| L3 | Strong | + HTF aligned AND volume confirmed | Standard position (50%) |
| L4 | Very Strong | + 2 or more consensus votes | Increased position (75%) |
| L5 | Maximum | + 4 consensus + volume surge + regime fit | Full conviction (100%) |
| # | Strategy | Type | Best For | WR% | Status |
|---|---|---|---|---|---|
| 1 | Connors RSI-2 | Mean Rev | SPY, BTC | 75.7% | PROVEN |
| 2 | Z-Score MR | Mean Rev | BTC, ETH | 62–77% | Research |
| 3 | EMA + RSI | Trend | ETH, BNB | ~60% | Research |
| 4 | MACD + RSI | Trend | 4H Crypto | 73% | Research |
| 5 | Bollinger Squeeze | Volatility | SOL, Ranging | 55–60% | Research |
| 6 | VWAP Reversion | Mean Rev | Intraday | 62–68% | Research |
| 7 | Supertrend | Trend | LINK, Trending | 55–60% | Research |
| 8 | Ichimoku Cloud | Trend | Forex, 4H+ | ~62% | Research |
| 9 | HMA Trend | Trend | SPY, QQQ | 59.1% | Research |
| 10 | Swing Failure (SFP) | Reversal | 1H-4H Crypto | 58–65% | Research |
| 11 | Liquidity Sweep | Reversal | 1H-4H Crypto | 73% | Research |
| 12 | Consensus | Multi | All pairs | Varies | Composite |
| Pair | Best TF | Strategy | WR% | Risk | Cap Tier |
|---|---|---|---|---|---|
| BTCUSD | 4H / 1D | RSI-2 / Z-MR | 62.5% | Med | Large |
| ETHUSD | 4H / 1H | EMA+RSI / MACD | 64.2% | Med | Large |
| SOLUSD | 1H / 4H | BB Squeeze | 58.3% | High | Mid |
| BNBUSD | 4H | EMA+RSI | 65.5% | Med | Large |
| XRPUSD | 4H / 1D | RSI-2 / Trend | 60.1% | Med | Large |
| AVAXUSD | 1H | MACD+RSI | 55.8% | High | Mid |
| DOGEUSD | 15m / 1H | Consensus | 52.8% | V.High | Meme |
| ADAUSD | 4H / 1D | Z-MR / VWAP | 59.4% | Med | Large |
| LINKUSD | 4H | Supertrend | 57.2% | High | Mid |
| MATICUSD | 1H / 4H | Ichimoku | 56.5% | High | Mid |
When Non-Repainting Mode = ON (default), signals only appear after the bar closes. This means:
Use backtest_cursor.py to test all strategies against crypto pairs programmatically:
# Test all strategies on all pairs
python backtest_cursor.py
# Single pair
python backtest_cursor.py --symbol BTCUSDT
# Single strategy
python backtest_cursor.py --strategy rsi2
# Single timeframe
python backtest_cursor.py --timeframe 4h
# Save best strategy per symbol
python backtest_cursor.py --save-best
# Download data if not cached
python backtest_cursor.py --fetchResults are saved to backtest_results/cursor_results.json and cursor_best.json.
Connors RSI-2 — it's the only strategy with statistically significant backtest proof (75.7% WR on SPY, p=6e-06, 992 trades). Auto-Detect defaults to this.
Past backtest results do not guarantee future performance. Market conditions change. Paper trade for 3–6 months before risking real capital.
Yes, but the crypto pair recommendations table is crypto-focused. Strategies like Connors RSI-2, EMA+RSI, and MACD work across asset classes.
Signals only appear after the bar closes (barstate.isconfirmed). They will never disappear or change retroactively. This gives honest backtest results.
Select "Multi-Strategy Consensus" from the dropdown. Set "Min Strategies Agreeing" (default 3). The engine counts how many of the 11 strategies agree on direction, with correlation caps to prevent redundant signals.
Based on the Pine Script code for Kimi Claw Pro v6.5, the phrase "strong bullish consensus across timeframes" (or "strong bearish") is a specific tooltip message that appears on the chart when certain conditions are met. It's not a standalone strategy but a summary insight from the Multi-Timeframe (MTF) Consensus Engine. Here's exactly what it means, how it's calculated, and what drives it:
request.security (non-repainting, meaning it uses confirmed/closed bars to avoid false signals).consensus_score):
consensus_score = (0.5 * higher_trend) + (0.3 * intermediate_trend) + (0.2 * primary_trend)This system is inspired by ensemble methods (combining multiple models for better accuracy) and statistical agreement (like in machine learning voting classifiers).
The overall strategy is a "multi-tool" trading indicator that combines many ideas to spot buy/sell opportunities. It looks at trends, momentum, volume, risk, and more to give scores and signals. I'll explain each major part like I'm talking to a high-school student—think of it as tools in a toolbox, each doing one job, and the indicator uses them together to decide "buy now" or "sell now."
What it does: These are like "smart lines" on the chart that follow price but smooth out noise. KAMA adjusts based on how wild the market is (fast in trends, slow in choppy times). HMA is quick to spot turns.
How it works in the strategy: If price is above these lines, it's bullish (buy signal boost). Below = bearish. It's like checking if a river is flowing uphill or downhill.
Why simple? Imagine averaging your test scores but weighting recent ones more—helps see if you're improving.
What it does: Figures out if the market is "trending" (steady up/down), "ranging" (bouncing around), or "normal" (mixed).
How it works in the strategy: Uses math like Hurst (checks if prices are random or patterned) and ADX (measures trend strength). Adjusts other tools—e.g., in trends, it favors following the flow; in ranges, it looks for bounces.
Why simple? Like weather: Sunny (trending) = go outside and run with it; rainy (ranging) = stay inside and wait for breaks.
RSI-2: Measures if something is "overbought" (too expensive, sell) or "oversold" (cheap, buy) over just 2 periods. Quick for short-term flips.
RVOL-T: Checks if trading volume is unusually high for the time of day (like a sudden crowd at school—something's happening).
Choppiness: Scores how "jumpy" prices are (high = messy, low = smooth trend).
SFP (Swing Failure Pattern): Spots when price pokes beyond a recent high/low but snaps back (like a fake-out in sports).
How they work in the strategy: Add points to buy/sell scores. E.g., extreme RSI-2 + low choppiness = strong mean-reversion buy.
Why simple? RSI-2 is like checking if a toy is too popular (sell) or ignored (buy). SFP is spotting a trick play that fails.
Ichimoku Cloud: A "cloud" on the chart—price above = buy, below = sell. Like a weather cloud forecasting rain or sun.
Supertrend: A line that flips colors for buy/sell. Combines trends with volatility.
Liquidation/Flash Crash/Sweep: Detects big price drops with volume spikes (buy the dip) or fake-outs.
How they work in the strategy: Each "votes" on buy/sell. Combined into "Elton Net" score (e.g., +20 = bullish bias).
Why simple? Think of a team vote: If most players say "pass the ball," you do it. Each strategy is one player.
Z-Score MR: Checks if price is too far from average (buy low, sell high).
Absorption/Climax: Spots when big trades are "absorbed" without moving price much (hidden buying/selling).
CVD (Volume Delta): Estimates if buyers or sellers are stronger by looking at candle shapes.
TTM Squeeze: Detects "squeezed" low volatility before big moves (like a spring uncoiling).
Cointegration: For BTC/ETH, checks if their prices are out of sync (buy the cheap one).
Kyle's Lambda: Measures how easily price moves (high = low liquidity, risky).
How they work in the strategy: Short-term signals for quick trades. Combined into "Scalp Net" score.
Why simple? Z-Score is like grading on a curve—if your score is way below average, it's a comeback opportunity. Squeeze is like pressure building in a balloon before it pops.
What it does: Checks trend agreement across 3 timeframes (short, medium, long).
How it works in the strategy: Averages EMA, MACD, and price signals per timeframe, then weights them (longer TFs count more). Score >80% bullish = "strong consensus" message.
Why simple? Like asking friends in different grades if a movie is good—if most say yes, it's probably worth watching.
What it does: Scores volume strength (spikes, trends) to confirm signals.
How it works in the strategy: High volume boosts buy/sell scores. Required for signals if enabled.
Why simple? Volume is like crowd noise at a game—loud means excitement, quiet means boring.
What it does: Measures speed and extremes (e.g., RSI >70 = overbought).
How it works in the strategy: Adds to confluence scores. E.g., bullish MACD = +10 points.
Why simple? Momentum is like a ball's speed—if rolling fast uphill, it might keep going.
What it does: Combines RSI, volatility, momentum, and volume into a "mood" score (0=extreme fear, 100=greed).
How it works in the strategy: Extreme fear boosts buy scores (contrarian).
Why simple? Like gauging if people are panicking (buy cheap) or too excited (sell high).
What it does: Calculates overall danger (0-100) from volatility, liquidity, size, etc.
How it works in the strategy: High score tightens TP/SL to reduce risk.
Why simple? Like a "danger meter" in a video game—high danger means play safer.
What it does: Combines all above into buy/sell signals if scores > threshold (e.g., 75).
How it works in the strategy: Checks trend/momentum/volume/MTF, adds bonuses (e.g., from Elton/Scalp). Uses Kelly for sizing, ATR for stops.
Why simple? Like a recipe: Mix ingredients (scores), bake if hot enough (threshold), serve with safety (stops).
What it does: Spots when multiple groups "ignite" (agree) for high-confidence alerts. ALT-ELT gives a final "buy/sell/wait" recommendation.
How it works in the strategy: Counts agreeing groups (e.g., trend + momentum = ignite). Boosts scores.
Why simple? Ignite is like multiple friends shouting "go!" at once. ALT-ELT is the final vote.
It's like a voting committee of 9+ tools. Each tool votes buy/sell based on its specialty (trends, bounces, volume). In "Dynamic" mode (recommended), votes are weighted by market type (trend vs range). If enough agree (e.g., 4+ votes) and filters pass (volume, trend bias), it signals. Risk tools (stops, sizing) keep you safe.
For High-School Analogy: Imagine planning a party. Tools check weather (regime), crowd excitement (volume/momentum), budget (risk), and friend opinions (votes). If most say "yes" and it's sunny, party on—but with a backup plan (stops) if it rains.