---
title: TradingView similar pattern scripts vs past-case search
url: https://comparand.xyz/blog/tradingview-pattern-similarity-scripts
description: 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.
updated: '2026-10-21'
category: Using Comparand
---

# TradingView similar pattern scripts vs past-case search

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.

Yes, TradingView has similar-pattern indicators, through community scripts. The TradingView guides we read show no built-in tool that ranks past look-alike charts. One script is LuxAlgo's open-source Echo Forecast. It compares recent prices with similar or dissimilar past stretches using correlation, and its author describes the output as a forecast. Comparand's chart search lists each same-coin, same-exchange case with its own later moves, and gives no average. The script fits you if you already work inside TradingView. A past-case search fits if you want separate cases, or want them from code or an AI app.

Both kinds of tool look for old charts that resemble a recent one; they differ in what they show. We did not run Echo Forecast. This article compares what each tool's own pages say.

For TrendSpider's version, see [Comparand vs TrendSpider](https://comparand.xyz/blog/comparand-vs-trendspider).

## What is a pattern-similarity script?

A script is a small program that runs on a TradingView chart. Community members write and share them. Some are open source, which means anyone can read the code.

A similarity script asks one question: which past stretches of price look most like the latest one? It needs a way to score "looks like". Echo Forecast uses correlation, a measure of how closely two price paths move up and down together.

This is different from a named pattern. TradingView's built-in tools find shapes with names, such as flags, triangles and head-and-shoulders. A similarity search needs no name. It just ranks past stretches.

*Diagram: Show a recent chart on the right and three older stretches on the left, each scored by how closely its shape matches the recent chart.*
## What do we know about Echo Forecast?

The facts we could confirm are few. LuxAlgo's [Echo Forecast](https://www.tradingview.com/script/Y8v4W3mX-The-Echo-Forecast-LuxAlgo/) is open source. It compares recent prices with similar or dissimilar historical windows using correlation. Its author describes the output as a forecast.

We could not confirm which market's history it searches, how many matches it uses or whether it avoids overlapping matches. Check the script page for those. TradingView's own documents did not show a built-in ranked search of past charts. That does not prove none exists, since we only read guides and one script page.

## What does TrendSpider's version do?

TrendSpider sells a similar tool in its store, an add-on called [Analog Path Projection](https://trendspider.com/trading-tools-store/indicators/6a8bda-analog-path-projection/). Its listing says it searches a symbol's own history for similar price episodes, marks where they happened and builds a consensus cone and median path.

We could not confirm its price or trial terms, or whether a base plan includes it. TrendSpider's help pages say its platform covers 61,000+ assets, including 28,000 crypto pairs on more than 10 exchanges. Those are the vendor's claims. Our [Comparand vs TrendSpider](https://comparand.xyz/blog/comparand-vs-trendspider) page compares the two in more detail.

## Should a tool show an average path or separate cases?

This is the main design choice, and it changes what you can learn.

A median path is one line. It is easy to read. But it can hide that the cases pointed different ways. If five past cases rose and five fell, the median can look like a gentle drift and describe none of them.

A list of separate cases lets you see every one, including the ones that went the other way. Comparand returns the list, with no average or hit rate. Splits are common. In 300 searches on Binance Bitcoin hourly charts, the ten closest past charts split 4 to 6 up in 65% of cases a day later. Our guide on [how often similar charts disagree](https://comparand.xyz/blog/when-past-crypto-cases-disagree) has the numbers for 12 markets.

Neither design is a forecast. Which one helps depends on whether you want a summary to start from or the cases to study.

## How do the three compare?

| Question | Echo Forecast (as described) | Analog Path Projection (as described) | Comparand chart search |
| --- | --- | --- | --- |
| Where does it run? | On a TradingView chart | In TrendSpider | Website, REST API or AI app through MCP |
| Which history is searched? | Not confirmed by what we read | The symbol's own history | Only the same coin on the same exchange and timeframe |
| What do you see? | Output that its author calls a forecast | Marked episodes, a consensus cone and a median path | Up to 20 separate cases (10 by default), each with its path and later moves |
| Is there an average? | Not confirmed | Yes, a median path | No average or hit rate |
| Do matches overlap? | Not confirmed | Not confirmed | No: one per episode, never overlapping your chart |
| Price | TradingView plans: Basic is free; paid plans from $12.95 a month billed yearly | Add-on price not confirmed | $0.01 per search from credits |

Sources: the [Echo Forecast page](https://www.tradingview.com/script/Y8v4W3mX-The-Echo-Forecast-LuxAlgo/), the [TrendSpider store listing](https://trendspider.com/trading-tools-store/indicators/6a8bda-analog-path-projection/), [TradingView pricing](https://www.tradingview.com/pricing/) and the Comparand docs. "Not confirmed" means we did not find it on the pages we read.

## What about the Hyperliquid search?

Comparand has a second search, hl_analogs. It works only on Hyperliquid. Instead of matching just the shape of the chart, it finds the past moments nearest to one coin's state at a time you choose. That search also checks buying, selling, open interest and liquidations. It returns the outcome over the next 1 hour or 4 hours, 20 cases by default, for about $0.01 a search.

It also compares a coin only with its own past.

## What do both approaches leave out?

- **Neither shows that matches beat chance.** In our own test, following past look-alikes did not beat always guessing up. See [Comparand vs TradingView](https://comparand.xyz/blog/comparand-vs-tradingview) for the wider comparison.
- **Settings change the answer.** A longer or shorter chart changes which past cases come back. Choose your chart length before you look at outcomes. See [why chart length matters](https://comparand.xyz/blog/chart-window-length-matters).
- **A list is not a backtest.** A backtest applies entry and exit rules with costs. A list of cases only shows what happened.

## How can you use both?

Keep the market the same in each tool. TradingView lists Binance BTCUSDT.P as a perpetual contract, a futures contract that never expires. A Coinbase Bitcoin spot chart is a different market, and [the exchange matters](https://comparand.xyz/blog/why-same-exchange-matters-chart-search) for which cases come back.

1. In TradingView, note the exchange, symbol, bar size and the time of the last completed bar.
2. Run the same market in Comparand, ending at the same time.
3. Compare the lists. Count the cases that rose and the ones that fell.

## How we measured this

This is a comparison of what each tool's own pages say, not a performance test. We read TradingView's guides and pricing, the Echo Forecast script page, TrendSpider's pricing and store listing, and Comparand's docs. We did not run any script, compare outputs or test accuracy. Where a page did not say something, we write "not confirmed", which does not prove the feature is missing.

## Try this in Comparand

With Comparand connected to your AI agent, ask:

- "Find the 10 closest past charts to the latest 48 completed hourly bars of Binance Ethereum, and show what price did after each." shape_analogs returns each case's own path and its later moves after a quarter, half and full window (12, 24 and 48 hours here), in basis points. One basis point is one hundredth of one percent. It returns no average or hit rate.
- "What past moments on Hyperliquid were closest to SOL's state at this time, and what followed over 4 hours?" hl_analogs returns the nearest past moments and their outcomes.

Each search costs about $0.01 from credits. The [methods page](https://comparand.xyz/methods) explains how matching works, and you can [try 3 coins free](https://comparand.xyz/try). Results are past cases, not a forecast.

## The short version

- Community scripts such as Echo Forecast run inside a TradingView chart; its author calls the output a forecast.
- TrendSpider's add-on builds a consensus cone and median path; Comparand lists separate cases with no average.
- Comparand compares a coin only with its own past on the same exchange and timeframe.
- Neither approach is shown here to beat chance; past cases are not a forecast.
- To use both, keep the exchange, market type, bar size and end time the same.

## Questions people ask

**Does TradingView have an indicator that finds similar past patterns?**

Yes, through community scripts. LuxAlgo's open-source Echo Forecast compares recent prices with similar or dissimilar past stretches using correlation, and its author describes the result as a forecast. TradingView's own drawing, replay and named-pattern tools do different jobs.

**What is the difference between a similarity script and Comparand?**

A script runs inside a TradingView chart. Comparand runs a search from its website, API or an AI app, and lists up to 20 separate past cases from the same coin on the same exchange. Each case shows its own later moves. It gives no average or hit rate.

**Does TrendSpider have something similar?**

Its store lists an add-on called Analog Path Projection. It searches a symbol's own history for similar price episodes, marks where they happened, and builds a consensus cone and median path. We could not confirm its price or whether a base plan includes it.

**Is a similar-pattern search a forecast?**

No. It shows what happened after similar charts in the past. The cases can point different ways, and in our tests the ten closest past charts often split between up and down. Past cases are not a forecast.