What Is Bitcoin Tick Data and Where to Find Free Samples

If you are building a trading bot or backtesting a strategy, you have probably encountered two types of historical data: OHLCV candles and tick data. Candles are convenient but they discard most of what actually happened in the market. Tick data keeps everything.

This article explains what Bitcoin tick data is, what each field means, why it matters for serious backtesting, and where you can download free samples today.

What Is Tick Data?

Tick data is a record of every individual trade executed on a market. Each row represents one transaction: a buyer and a seller agreed on a price and exchanged a quantity of an asset at a specific moment in time.

For Bitcoin spot markets, a typical tick dataset contains the following fields:

trade_id       — unique identifier for the trade
price          — execution price in USDT
qty            — quantity of BTC traded
quote_qty      — notional value (price × qty) in USDT
time           — Unix timestamp in milliseconds
is_buyer_maker — True if the buyer was the market maker

The is_buyer_maker field is particularly valuable. When it is False, a buyer hit the ask — a buy-side aggressor. When it is True, a seller hit the bid — a sell-side aggressor. This is the foundation of order flow analysis.

Tick Data vs OHLCV Candles

A 1-minute OHLCV candle compresses every trade in that minute into four price points (open, high, low, close) plus total volume. That compression destroys information:

• You cannot tell how many individual trades occurred
• You cannot distinguish buy-initiated from sell-initiated volume
• You cannot detect the sequence of aggressive orders
• You cannot reconstruct the exact price path within the candle

For many strategies — momentum, market making, order flow imbalance — that destroyed information is precisely what drives the signal. Tick data preserves it.

How Much Data Are We Talking About?

Bitcoin has been one of the most actively traded assets in the world since 2020. The Last Year Bitcoin tick dataset (rolling 12 months) contains approximately 1.42 billion individual trades. Stored in Apache Parquet with ZSTD compression, that compresses to around 3.6 GB on disk.

Loading 6.2 billion rows into a pandas DataFrame on a typical laptop is not feasible. You need columnar storage (Parquet), lazy evaluation, and a query engine capable of predicate pushdown — tools like DuckDB, Polars, or PySpark.

Reading Bitcoin Tick Data in Python

Here is how to load a 1-week sample with Polars and compute basic aggressor-side statistics:

import polars as pl

df = pl.read_parquet("BTC_trades_sample.parquet")

print(f"Rows: {len(df):,}")
print(f"Columns: {df.columns}")
print(f"Date range: {df['time'].min()} → {df['time'].max()}")

# Aggressor side breakdown
aggressor = (
    df.group_by("is_buyer_maker")
    .agg(
        pl.len().alias("trades"),
        pl.col("quote_qty").sum().alias("notional_usdt")
    )
    .sort("is_buyer_maker")
)
print(aggressor)

On a 1-week sample of approximately 11 million rows, this runs in under two seconds on a standard laptop.

Why Audited Data Matters

Not all tick datasets are equal. Public archives occasionally contain timestamp anomalies — for example, a silent unit change from milliseconds to microseconds mid-dataset — that corrupt any time-based analysis without raising an obvious error. A strategy backtested on unaudited data may be optimized around an artifact, not a real market signal.

Audited datasets document every known anomaly, gap, and correction applied during consolidation. That transparency is not a weakness — it is the only honest way to distribute historical data.

Where to Download Free Bitcoin Tick Data Samples

The Glitch List publishes free 1-week samples of audited Bitcoin tick data in Apache Parquet format. The sample covers a full trading week with millisecond precision, zero gaps, and a documented schema.

You can download it from:

• Kaggle — bitcoin-tick-data-free-sample
• HuggingFace — bitcoin-tick-data-free-sample
• Direct download — free sample

Summary

Bitcoin tick data is the complete record of every trade executed on a spot market: price, quantity, timestamp, and aggressor side. It is the highest-resolution market data available for backtesting and microstructure research. OHLCV candles are derived from it — but once compressed, the underlying information cannot be recovered.

If your strategy depends on order flow, trade frequency, or volume imbalance, tick data is not optional. It is the only data that contains your signal.