--- title: Comparand Blog url: https://comparand.xyz/blog description: Plain-English guides to crypto charts, past cases, backtests and AI trading agents, from Comparand. updated: '2026-10-21' --- # Comparand Blog Plain-English guides to crypto charts, past cases and AI trading tools. Past cases, not a forecast. ## Basics - [Crypto base rates: how often does Bitcoin go up?](https://comparand.xyz/blog/past-crypto-returns-not-forecasts.md): Binance Bitcoin rose in 51.88% of 24-hour periods, and 10 big coins ranged from 48.39% to 53.84%. Why a base rate comes before any pattern claim. - [How does liquidation work in crypto? Margin, price, cascades](https://comparand.xyz/blog/how-liquidation-works-crypto.md): Liquidation closes a leveraged trade when margin runs out. How to work out your liquidation price, what maintenance margin is, and how cascades start. - [How to read a candlestick chart, with real Bitcoin candles](https://comparand.xyz/blog/how-to-read-candlestick-chart.md): How to read a candlestick: open, high, low, close, body, wicks and colors, using real Bitcoin candles. Green and red were close to even: 50.61% green. - [What is a Sharpe ratio? Explained with real crypto tests](https://comparand.xyz/blog/sharpe-ratio-explained-crypto.md): A Sharpe ratio is return per unit of swing. How to annualize it for a market that never closes, and real Bitcoin numbers: just holding BTC scored 0.78. - [What is a chart analog? A guide with real Bitcoin data](https://comparand.xyz/blog/what-is-a-crypto-chart-analog.md): A chart analog is a past chart shaped like today's. What bars, windows and shapes mean, how matches are picked, and how a Bitcoin analog rule did. - [What is a funding rate? A plain guide with real numbers](https://comparand.xyz/blog/crypto-funding-rate-explained.md): A funding rate is a payment between longs and shorts on a perp. Who pays whom, hourly vs 8-hour rates, and what a BTC long cost per day, week and month. - [What is a perpetual future? Perps explained simply](https://comparand.xyz/blog/what-is-a-perpetual-future.md): A perp is a futures contract with no end date. How funding keeps it near spot, what margin means, and how close Bitcoin's perp stayed to spot each year. - [What is crypto volatility and how do you measure it?](https://comparand.xyz/blog/what-is-crypto-volatility.md): Volatility is how much price moves. Six ways to measure it, how to work it out, and why Bitcoin's middle day (1.38%) and average day (2.10%) differ. - [What is max drawdown? Formula and real Bitcoin examples](https://comparand.xyz/blog/what-is-max-drawdown.md): Max drawdown is the biggest fall from a high to a later low. How to calculate it step by step, and Bitcoin's drawdown inside each year from 2020 to 2026. - [What is open interest in crypto? Explained with Hyperliquid data](https://comparand.xyz/blog/what-is-open-interest-crypto.md): Open interest is the total size of trades still open. How it differs from volume, the four price and open interest mixes, and how often each happened. - [What time does the crypto daily candle close? UTC and local](https://comparand.xyz/blog/what-time-does-crypto-daily-candle-close.md): Binance and Coinbase daily candles close at 00:00 UTC, 8 pm in New York in summer. The next hour moved Bitcoin 0.44% on average. Local times inside. - [Why is Bitcoin price different on exchanges?](https://comparand.xyz/blog/why-bitcoin-price-differs-between-exchanges.md): Coinbase Bitcoin was a median 0.03% above Binance over 62,342 hours since 2019. See the gap by year and why currency and contract type matter. ## How-to - [CCXT vs a hosted data API: which for crypto history?](https://comparand.xyz/blog/ccxt-vs-hosted-crypto-data-api.md): CCXT is a free library that asks each exchange for candles. A hosted API stores them for you. See when each fits, plus measured exchange differences. - [Crypto position sizing by volatility: the math, real numbers](https://comparand.xyz/blog/position-sizing-by-volatility.md): Position size is your risk divided by your stop distance. The math, and why a stop set by each coin's own hourly range was hit about as often on 10 coins. - [Find similar chart patterns in Python, or use an API?](https://comparand.xyz/blog/python-find-similar-chart-patterns.md): Build a similar-chart search in Python with pandas, or call a hosted one? The code steps, five pitfalls from look-ahead to overlap, and what each costs. - [Freqtrade vs vectorbt vs Backtrader for crypto backtests](https://comparand.xyz/blog/freqtrade-vectorbt-backtrader-compared.md): Freqtrade, vectorbt, Backtrader and QuantConnect LEAN compared from their own docs: crypto data, live trading, licences, limits, and a look-ahead trap. - [How many trades does a backtest need to be significant?](https://comparand.xyz/blog/how-many-trades-backtest-sample-size.md): Luck moves a 50% hit rate about 10 points at 100 trades and 3 at 1,000. Why your trades may not be separate, and what 3,600 real tests could not show. - [How to backtest a crypto chart pattern: 5 ways compared](https://comparand.xyz/blog/ways-to-backtest-a-crypto-chart-pattern.md): Compare five ways to test crypto chart patterns, with real Bitcoin results showing why later dates and trading costs change the answer. - [How to download Hyperliquid historical data and check it](https://comparand.xyz/blog/how-to-download-hyperliquid-historical-data.md): Get Hyperliquid candles, funding and liquidation data as whole-day files, check them with a hash, and spot gaps first. Real coverage numbers. - [How wide should a crypto stop loss be? What the data shows](https://comparand.xyz/blog/stop-loss-distance-crypto-data.md): On Binance Bitcoin, a price 2% below was touched within a day 39.29% of the time, 1% below 63.54%. See 10 coins, ATR distances and what wicks do. - [What is a markout in trading? How to check your fills](https://comparand.xyz/blog/trading-markouts-explained.md): A markout is the price move after your fill. Learn to work it out, what it says about being picked off, how fees count, and what a normal Bitcoin hour is. ## Coins - [Avalanche (AVAX) price history: drops, recoveries and big days](https://comparand.xyz/blog/avalanche-avax-price-history-numbers.md): AVAX fell 90.47% within 2022 on Binance, from $114.11 to $10.87, and fell 10% in a week on 20.40% of days. Yearly drops, recoveries, big days. - [Bitcoin Ethereum correlation by year, plus 8 more coins](https://comparand.xyz/blog/bitcoin-ethereum-correlation-by-year.md): BTC and ETH daily moves had a 0.82 correlation on Binance since 2019, from 0.77 in 2021 to 0.91 in 2026. See 8 more coins and what this means. - [Bitcoin halving price history: what came after each one](https://comparand.xyz/blog/bitcoin-halving-price-history.md): Bitcoin was up 281.5%, 562.0% and 31.1% a year after the 2016, 2020 and 2024 halvings. See the peaks, the falls after them, and why 3 cases are few. - [Cardano (ADA) price history: drops and recoveries measured](https://comparand.xyz/blog/cardano-ada-price-history-numbers.md): Cardano fell 46.54% to 84.77% within each year since 2020 on Binance, and four of seven yearly highs were not regained. Daily moves and weekly falls. - [Chainlink (LINK) price history: drops, recoveries and big days](https://comparand.xyz/blog/chainlink-link-price-history-numbers.md): Chainlink's yearly drops ran 40.25% to 80.43% on Binance, and six of seven yearly highs were regained. Daily moves, weekly falls and 24-hour results. - [Do altcoins follow Bitcoin? How far each coin moves](https://comparand.xyz/blog/do-altcoins-follow-bitcoin.md): On Binance, Dogecoin moved 1.33% per 1% Bitcoin move and TRX only 0.72%. Beta for 9 altcoins, by year, and on Bitcoin's big down and up days. - [Does Ethereum fall more than Bitcoin in crashes?](https://comparand.xyz/blog/ethereum-vs-bitcoin-crash-size.md): On 103 Binance days when Bitcoin fell 5% or more, Ethereum fell more on 70.87%, but on big up days it rose more only 38.17% of the time. - [Dogecoin volatility in numbers: how big DOGE's days get](https://comparand.xyz/blog/dogecoin-volatility-in-numbers.md): A typical Dogecoin day moved 2.33% on Binance vs 1.38% for Bitcoin, and 1 day in 5 moved over 5%. Big days, weekly falls and recoveries. - [Ethereum price history in numbers: moves, drops, recoveries](https://comparand.xyz/blog/ethereum-price-history-in-numbers.md): Binance ETH moved a typical 1.95% a day, fell 10% in a week on 10.83% of days, and needed 628 days after its 2022 low to recover. Compared with BTC. - [HYPE funding rate: what holding a HYPE perp cost, measured](https://comparand.xyz/blog/hype-funding-cost-hyperliquid.md): A $1,000 HYPE long on Hyperliquid paid a median $9.65 in funding per 30 days, the most of 24 coins we measured. Daily, weekly and monthly numbers. - [How much does Bitcoin move in a day? 11 years measured](https://comparand.xyz/blog/how-much-does-bitcoin-move-in-a-day.md): A typical Bitcoin day moved 1.38% on Binance and 1.43% on Coinbase, yet 9.24% of Binance days moved over 5%. See each year, big days and 9 more coins. - [Is BNB less volatile? BNB price history in numbers](https://comparand.xyz/blog/bnb-price-history-numbers.md): On Binance, BNB's median day moved 1.57% vs 1.38% for Bitcoin, yet its worst yearly fall was deeper in 6 of 7 years. See weekly falls and 24-hour rises. - [Is Uptober real? Bitcoin's returns by month since 2015](https://comparand.xyz/blog/best-month-for-bitcoin-uptober.md): Coinbase Bitcoin rose in 9 of 11 Octobers since 2015, median +14.44%. See every month, other coins, and why 11 years is a tiny sample. - [Most volatile crypto: 10 big coins ranked by daily moves](https://comparand.xyz/blog/most-volatile-big-crypto-coins.md): SOL had the biggest typical day of 10 big Binance coins (2.91%), but AVAX led 4 of 6 years and DOGE had the biggest days. See the order by year. - [Solana price history in numbers: how wild is SOL?](https://comparand.xyz/blog/solana-price-history-in-numbers.md): Binance SOL moved a typical 2.91% a day, fell 94.63% in 2022, and ended higher over 24 hours 49.66% of the time. Compared with BTC, year by year. - [TRON (TRX) price history in numbers: the quiet big coin](https://comparand.xyz/blog/tron-trx-price-history-numbers.md): TRX had the smallest median daily move of 10 big coins on Binance, 1.28%, yet one day rose 95.82%. See weekly falls, yearly drops and 24-hour rises. - [What happened to Bitcoin after similar charts?](https://comparand.xyz/blog/bitcoin-after-closest-past-charts.md): In 300 Binance tests, the 10 closest past Bitcoin charts split 4 to 6 up in 65.00% of cases. See Coinbase too, fees, and Bitcoin's normal range. - [What happens after Bitcoin hits an all-time high? 13 cases](https://comparand.xyz/blog/bitcoin-after-all-time-high.md): After 13 new Bitcoin highs on Coinbase, price was up 30 days later in 53.85% of cases, with a middle move of +0.40%. Every case is listed with its date. - [XRP price history in numbers: big days and long waits](https://comparand.xyz/blog/xrp-price-history-in-numbers.md): Binance XRP had a typical 1.96% day but a 73.23% one-day jump, and its 2021 drop took 1,229 days to recover. Moves, drops and 24-hour odds vs BTC. ## Patterns - [Best timeframe for crypto? 1-hour vs 4-hour vs daily data](https://comparand.xyz/blog/timeframe-1h-vs-4h-vs-daily-chart.md): On Binance Bitcoin the typical candle ranged 0.64% on 1-hour, 1.35% on 4-hour and 3.74% on daily charts. Wicks, green share and noise compared. - [Bitcoin 200-day moving average: what came next, above vs below](https://comparand.xyz/blog/bitcoin-200-day-moving-average.md): Coinbase Bitcoin was higher 90 days later 66.79% of the time after a close above its 200-day average, 48.64% after one below. See 10 more coins. - [Bitcoin Sunday pump, Monday dump: does the pattern hold?](https://comparand.xyz/blog/sunday-pump-monday-dump.md): After an up Sunday, Coinbase Bitcoin fell on Monday 54.03% of the time vs 48.55% for any Monday. See other coins, each year and the weekend move. - [Do breakouts work in crypto? 20-day highs tested](https://comparand.xyz/blog/do-breakouts-work-crypto.md): On 10 Binance perps, 39.83% of daily 20-day-high breakouts fell back in 3 days. Ten days on, the average gain was 6.26%, but the middle case was 1.53%. - [Does Bitcoin always fill CME gaps? A spot-price test](https://comparand.xyz/blog/does-bitcoin-fill-cme-gaps.md): Coinbase Bitcoin returned to its Friday price within 3 days after 71.81% of weekends, but a same-distance level was touched about as often. - [Does RSI below 30 work in crypto? 10 coins tested](https://comparand.xyz/blog/rsi-oversold-crypto-backtest.md): After RSI first fell below 30, 10 Binance coins rose a day later 54.27% of the time, but the average move matched normal hours. See fees and years. - [Does volume confirm a crypto breakout? 1,411 tested](https://comparand.xyz/blog/does-volume-confirm-breakouts.md): On 10 Binance perps, high-volume daily breakouts rose 9.15% on average over 10 days vs 5.26% on normal volume. The medians were far smaller. - [Golden cross and death cross: every Bitcoin case measured](https://comparand.xyz/blog/golden-cross-death-cross-bitcoin-results.md): Coinbase Bitcoin was higher 90 days after 7 of 8 golden crosses, but only 8 cases exist. See each cross, 10 more coins, and a rule after fees. - [How many candles for pattern matching? Chart length matters](https://comparand.xyz/blog/chart-window-length-matters.md): We changed chart length from 24 to 48 to 96 hourly candles for Bitcoin and Ethereum. Most past matches changed, and no length won everywhere. - [How much do trading fees hurt a crypto backtest? We tested](https://comparand.xyz/blog/crypto-trading-fees-backtest-impact.md): A Bitcoin rule that trades about 210 times a year went from 27.52% to 3.41% a year as fees rose. The daily version barely changed. See the table. - [How to tell if crypto is trending or ranging: 3 measures](https://comparand.xyz/blog/market-regime-trending-or-ranging.md): Three simple regime measures on Binance Bitcoin by year: variance ratio, days above the 200-day average, and daily range. What each shows and misses. - [Look-ahead bias backtest example: one timing bug](https://comparand.xyz/blog/look-ahead-bias-backtest-example.md): One timing bug turned a losing hourly Bitcoin rule into a fake winner: Sharpe −8.00 became 59.41. See how it happens and how to check your test. - [Overfitting in trading backtests: we tested 1,000 random rules](https://comparand.xyz/blog/overfitting-1000-random-trading-rules.md): We picked the best of 1,000 random Bitcoin rules: +155.22% after fees in training, then −51.72% on later dates. Here is what overfitting looks like. - [Similar crypto charts, different outcomes: how often?](https://comparand.xyz/blog/when-past-crypto-cases-disagree.md): In 3,600 tests on 12 crypto markets, the 10 closest past charts split 4 to 6 up in 62.72% of cases a day later. What a split does and does not show. - [Walk-forward analysis explained with a real crypto test](https://comparand.xyz/blog/walk-forward-testing-explained.md): Walk-forward testing checks a method one step at a time, using only past data. Rolling vs anchored windows, and a Bitcoin test scored year by year. - [Why do crypto candles have long wicks? 62,357 candles measured](https://comparand.xyz/blog/crypto-long-wicks-explained-data.md): In 61.82% of Binance Bitcoin's hourly candles, the wicks together were longer than the body. See wick sizes for 12 markets and busy vs quiet years. ## Data - [Best day of the week to buy Bitcoin? 11 years of data](https://comparand.xyz/blog/best-day-of-week-to-buy-bitcoin.md): No weekday stood out in 11 years of Coinbase Bitcoin data. Monday led on average (+0.42%), but its median was +0.09% and the top day kept changing. - [Bitcoin after a 10% weekly drop: buy-the-dip data](https://comparand.xyz/blog/bitcoin-after-10-percent-weekly-drop.md): After 77 weekly drops of 10% or more on Coinbase since 2015, Bitcoin was higher 30 days later 57.14% of the time, about the same as any day (58.04%). - [CoinGecko API alternatives for historical OHLCV data](https://comparand.xyz/blog/coingecko-api-alternatives-ohlcv.md): CoinGecko's coin-chart history is not one exchange's candles. See what its docs say and compare exchange APIs, CCXT, CoinAPI, CoinDesk Data and Comparand. - [CoinGlass alternatives for liquidation and funding data](https://comparand.xyz/blog/coinglass-alternatives-liquidation-data.md): Compare CoinGlass, Coinalyze, Velo, Laevitas, exchange APIs and Comparand for liquidation and funding data: history, prices and limits. - [Coinalyze vs Velo vs Laevitas: derivatives data compared](https://comparand.xyz/blog/coinalyze-velo-laevitas-compared.md): Coinalyze vs Velo vs Laevitas for funding, open interest and liquidations: published prices, history depth, and which fits charting, options or an API. - [Crypto drawdown recovery time: 10 big coins after 2022](https://comparand.xyz/blog/crypto-drawdowns-and-recovery-times.md): Measured yearly drops for 10 big coins on Binance, how many days each took to get back to its high, and which have not recovered yet. - [DCA vs lump sum Bitcoin: every start date tested](https://comparand.xyz/blog/dca-vs-lump-sum-bitcoin.md): Buying Bitcoin all at once beat 52 weekly buys on 68.14% of 3,581 start days since 2016. See the start-year split, Ether, 4 years and the limits. - [Do chart patterns predict crypto prices? We tested look-alikes](https://comparand.xyz/blog/do-chart-lookalikes-predict-crypto.md): In 3,600 tests on 12 crypto markets, chart look-alikes picked the next-day direction 50.22% of the time, below the 50.92% from always guessing up. - [Do longs or shorts get liquidated more? Hyperliquid data](https://comparand.xyz/blog/longs-vs-shorts-liquidated-more.md): Forced longs were 58.95% of Hyperliquid liquidated dollars over 114 days, but fell to 49.64% by September. See the months and what followed. - [Does Bitcoin dump at the US market open? 1,788 days measured](https://comparand.xyz/blog/does-bitcoin-dump-at-us-market-open.md): Binance Bitcoin's 9:30 a.m. New York hour rose on 48.10% of 1,788 days, averaging +0.7 bp. No drop pattern, but moves were 2.02x a weekend hour. - [Does high funding mean price will drop? Hyperliquid data](https://comparand.xyz/blog/does-high-funding-predict-price-drop.md): On Hyperliquid, price fell the next day after 53.37% of high-funding days, versus 52.72% of all days. See low funding, 3 days and fees. - [Glassnode vs CryptoQuant: on-chain data compared](https://comparand.xyz/blog/glassnode-vs-cryptoquant.md): Glassnode and CryptoQuant compared by published metrics, history, plans and API limits, plus how on-chain data differs from exchange price history. - [How do new Hyperliquid listings trade in month one?](https://comparand.xyz/blog/hyperliquid-new-listings-first-month.md): For 19 Hyperliquid coins, the typical daily move was 5.86% in days 1-30 and 3.86% in days 31-120. 17 of 19 calmed down. See which two did not. - [How long do Bitcoin rallies last? 51 rallies measured](https://comparand.xyz/blog/how-long-do-crypto-rallies-last.md): Coinbase Bitcoin rallies of 20% or more took a median 26 days and gained 48.23%. Winning streaks lasted a median 2 days. Ethereum too, and the limits. - [How much does Bitcoin move on CPI and Fed days?](https://comparand.xyz/blog/bitcoin-cpi-fomc-day-volatility.md): Binance Bitcoin's Fed decision hour averaged a 1.10% move vs 0.42% on normal days. CPI hours averaged 1.05%, but the typical one was about normal. - [How often does Bitcoin drop 10% in a week?](https://comparand.xyz/blog/how-often-bitcoin-drops-10-percent-week.md): Bitcoin closed 10% or more below its close a week earlier on 6.79% of Binance days and 8.41% of Coinbase days. Each year, 20% and 30% drops, other coins. - [How often would 10x or 20x leverage get liquidated?](https://comparand.xyz/blog/how-much-leverage-liquidation-odds.md): On Binance Bitcoin, a 10% move against you came within a day 1.94% of the time and within a week 16.21%. Odds for 5x to 100x, with the limits. - [Hyperliquid funding rate history: how often is it negative?](https://comparand.xyz/blog/hyperliquid-funding-rates-data.md): Measured hourly Hyperliquid funding for 24 coins: how often it was negative, typical rates, yearly averages, and how much it changed by month. - [Hyperliquid wallet trackers compared: what each shows](https://comparand.xyz/blog/hyperliquid-wallet-trackers-compared.md): Hyperdash, Arkham, Nansen, Hyperliquid's public API and Comparand compared on what a wallet view shows, live or history, and the price each publishes. - [Is crypto more volatile on weekends? Our measured answer](https://comparand.xyz/blog/crypto-weekend-volatility.md): Binance Bitcoin moved 0.29% per hour on weekends vs 0.42% on weekdays, and all 10 coins we measured were calmer. Saturday vs Sunday, and the limits. - [Kaiko and Amberdata alternatives for crypto market data](https://comparand.xyz/blog/kaiko-amberdata-alternatives.md): Kaiko, Amberdata, CoinAPI, CoinDesk Data, Tardis.dev and Comparand compared by published coverage, history, delivery and price, and who each suits. - [Most volatile crypto hours of the day, in UTC and local time](https://comparand.xyz/blog/most-volatile-hours-crypto-utc.md): We measured average hourly moves for 12 crypto markets on Binance and Coinbase. See the busiest hours in UTC, New York, London and Tokyo time. - [Polymarket and Kalshi historical data API: where to get it](https://comparand.xyz/blog/polymarket-kalshi-historical-data.md): Where to get Polymarket and Kalshi price history: what each API returns, its time steps, its limits and cost, plus one tool for both. - [Rising open interest and falling price: what came next?](https://comparand.xyz/blog/rising-open-interest-falling-price.md): On 169 Hyperliquid coins, rising open interest with falling price was followed by a higher next day 51.44% of the time, vs 50.43% for all days. - [Tardis.dev alternatives for crypto tick and trade data](https://comparand.xyz/blog/tardis-dev-alternatives-tick-data.md): Tardis.dev, CoinAPI, Kaiko, CoinDesk Data and exchange archives compared for tick, trade and order-book data: what each sells and who each suits. - [Volatility clustering in crypto: does a big day mean another?](https://comparand.xyz/blog/volatility-clustering-crypto.md): After a big Binance Bitcoin day, the next day was also big 15.84% of the time vs 9.05% normally. See 10 coins, each year, and what this cannot show. - [What happened after crypto liquidation cascades? 300 cases](https://comparand.xyz/blog/after-crypto-liquidation-cascades.md): We measured Hyperliquid's 300 largest liquidation episodes: price 15 minutes to a day later, longs vs shorts, before fees, and the limits. - [What happens after Bitcoin drops suddenly in one hour?](https://comparand.xyz/blog/bitcoin-after-big-one-hour-drops.md): After big 1-hour drops, Binance Bitcoin was up a day later 56.16% of the time vs 52.04% for all hours, and 53.77% with drops spaced 48 hours apart. - [What happens after Bitcoin has 5 red days in a row?](https://comparand.xyz/blog/red-days-in-a-row-bitcoin.md): After Binance Bitcoin's fifth red day in a row, the next day was up in 63.33% of 30 cases, but 3 days later only 46.67%. See all lengths and 12 markets. - [What happens after Bitcoin's biggest 1-hour pumps?](https://comparand.xyz/blog/what-happens-after-bitcoin-pumps.md): After big 1-hour pumps, Binance Bitcoin was up a day later 52.28% of the time vs 52.04% for all hours. Compare drops, years, Ethereum and Solana. - [What time of day do crypto liquidations happen?](https://comparand.xyz/blog/when-do-crypto-liquidations-happen.md): On Hyperliquid, liquidations peaked in the hour starting 15:00 UTC (11:00 in New York in summer) and were rarer on weekends. Hours, days, limits. ## AI agents - [AI trading agent risks: 6 guardrails before you trust one](https://comparand.xyz/blog/ai-trading-agent-guardrails.md): Six checks before you trust an AI trading agent: stale data, sources, order access, spending caps, tests on unseen dates and a human read. - [Best crypto MCP server? CoinGecko, CCXT, CoinAPI compared](https://comparand.xyz/blog/crypto-mcp-servers-compared.md): CoinGecko, CCXT, CoinAPI, Laevitas and Comparand MCP servers compared by data, history, keys, published prices and whether they can place orders. - [Best crypto historical data API: 6 options compared](https://comparand.xyz/blog/crypto-market-data-apis-compared.md): CCXT, CoinGecko, CoinGlass, Kaiko, Amberdata and Comparand compared by data type, published prices, history limits and who each one suits. - [Can ChatGPT or Claude backtest a trading strategy?](https://comparand.xyz/blog/can-ai-backtest-a-trading-strategy.md): Yes, if it runs code on real bars. In our Bitcoin test, leaving out fees showed 464.36% where the rule made 167.73%. Four mistakes and a checklist. - [ChatGPT vs Claude for crypto trading research: what each does](https://comparand.xyz/blog/chatgpt-vs-claude-crypto-research.md): What ChatGPT and Claude each document for web search, files and MCP tools, what we could not confirm, and why both need a data tool for real numbers. - [Connect Claude or ChatGPT to crypto data with MCP](https://comparand.xyz/blog/connect-claude-chatgpt-to-crypto-data-mcp.md): Give Claude or ChatGPT real crypto price history through MCP: the settings, a first question to ask, and a measured Bitcoin example. - [How to use ChatGPT or Claude to analyze a crypto chart honestly](https://comparand.xyz/blog/how-to-use-ai-to-analyze-crypto-chart.md): How to ask ChatGPT or Claude about a crypto chart so it counts past cases instead of guessing: what to give it, prompts that work and checks to run. ## Using Comparand - [Comparand vs TradingView: which fits chart pattern search?](https://comparand.xyz/blog/comparand-vs-tradingview.md): TradingView is for charting, Bar Replay and strategy tests. Comparand ranks a coin's own past look-alike charts and shows what came next. - [Comparand vs TrendSpider: two ways to find past look-alikes](https://comparand.xyz/blog/comparand-vs-trendspider.md): TrendSpider's Analog Path Projection draws one consensus path from past look-alikes. Comparand lists each same-coin past case by API or AI app. - [How to check crypto data quality and where it came from](https://comparand.xyz/blog/check-crypto-data-provenance.md): Four checks for crypto price data: missing bars, unfinished candles, exchange gaps and file hashes, with real Bitcoin numbers for each. - [How to find past crypto charts that look like today's](https://comparand.xyz/blog/find-past-charts-like-today.md): How to find past crypto charts like today's in five steps: coin, exchange, candle size, window length and match count, then how to read each case. - [TradingView similar pattern scripts vs past-case search](https://comparand.xyz/blog/tradingview-pattern-similarity-scripts.md): How community scripts like LuxAlgo's Echo Forecast compare with a past-case search that lists each same-coin, same-exchange case and its own outcome. - [Why chart pattern matching should stay on one exchange](https://comparand.xyz/blog/why-same-exchange-matters-chart-search.md): The same Bitcoin chart search scored 50.67% on Binance and 47.33% on Coinbase. Why Comparand matches a coin only with its own past on one exchange.