Order flow field guide
How to Read a Crypto Footprint Chart Without Mistaking Aggression for Direction
A practical guide to bid/ask ladders, delta, imbalances, POC and absorption on crypto footprint charts—with the data limits and failed signals that matter.
Basis Research · 2026-10-03 · 9 min read
A candle tells you where a market opened and closed. A footprint asks a harder question: what traded at each price inside that candle, and which side crossed the spread? The distinction matters most when a large amount of aggressive buying produces almost no upward progress—or aggressive selling fails to push price lower.
This guide uses executed trades, not a prediction of the next candle. It explains what the numbers can establish, what they cannot, and a repeatable way to inspect a level in Basis without treating every green cell as a long signal.
01 / 07
Start with the instrument, not the colored cells
The same BTC move can produce different footprint readings on Binance spot, Binance perpetuals and another venue. Each feed represents its own executed trades. Before making a comparison, record the exchange, spot or perpetual market, symbol and timeframe. Do not interpret a single-venue ladder as the complete global crypto market.
Open a liquid pair and a short enough interval that individual bars are readable. The footprint is built from trade data; older bars and thin markets can have less exact coverage. Wait for the tape coverage indicator before assuming a sparse historical ladder means no trading occurred.
02 / 07
What one row actually says
At each price level the footprint groups traded volume by aggressor side. “At the ask” means buyers lifted offers; “at the bid” means sellers hit bids. These are completed trades, unlike the DOM, where visible orders can still be moved or canceled. A row with 420 bid and 980 ask volume therefore records more aggressive buying at that price—it does not tell you how much passive selling remains.
Row size is an analytical choice. A one-tick ladder reveals more detail but may be noisy or visually expensive; grouping multiple ticks creates fewer, larger rows. If you change the grouping, you change the units being compared. Keep it stable across the bars in your review.
03 / 07
Delta measures urgency, not control
Bar delta is ask-traded volume minus bid-traded volume. For example, 1,800 units at the ask and 1,200 at the bid produce +600 delta. The arithmetic is simple; the inference is not. Buyers can be aggressive into a seller who absorbs them, leaving a positive delta bar that closes below its high.
Begin with the outcome: did price advance, stall or reverse? Then compare that outcome with the delta. Strong buying plus little upward progress at a pre-marked resistance level is a possible absorption clue. Strong buying with clean price progress through the level is a different auction. Neither condition guarantees what happens next.
04 / 07
Use imbalance and stacked imbalance as context
An imbalance highlights a large difference between opposing sides at nearby price rows. Basis lets you adjust the ratio and minimum volume. A 3:1 ratio on tiny absolute volume can be meaningless, which is why the volume floor matters. Several adjacent imbalances in one direction form a stacked zone: a stretch where aggression was concentrated rather than a single isolated print.
Mark the zone, then watch how price behaves when it returns. Was the area accepted, rejected, or crossed without reaction? An imbalance by itself is evidence of how a past move traded. It is not an instruction to enter when price revisits the same number.
05 / 07
POC is a location, not a magnet
The bar point of control (POC) identifies the most-traded price row within that bar. It tells you where business was concentrated. Compare it with the bar’s range: a POC near the lower end of a rising bar suggests a different path than a POC near the high of a bar that later rejects.
Do not confuse a bar POC with a session volume-profile POC. The first belongs to one candle; the second aggregates a much larger window. Both can be useful, but they answer different questions and should be labeled separately on a chart or in a review log.
06 / 07
A disciplined absorption test
Suppose BTC trades into a resistance area you marked before the move. The next bar shows strong ask-side volume, positive delta and a small upper wick, but no durable close above the area. That is a candidate for passive sellers absorbing aggressive buyers—not proof. A large player may be hedging, the level may still break, or the tape may be incomplete.
Write down the candidate before seeing the next bar. Require an observable follow-through rule, such as a close back below the area or a failed retest, and define the price at which the idea is wrong. If price accepts above the area, cancel the absorption thesis rather than relabeling the same bar after the fact.
- Pre-mark the price level from structure, not from the loudest footprint cell.
- Observe aggression and whether price made proportional progress.
- Wait for acceptance or rejection, then record invalidation and execution costs.
- Log failed candidates as well as attractive examples.
07 / 07
What the footprint cannot see
A footprint cannot reveal all hidden liquidity, all trading on other exchanges, or why a participant traded. Large prints may be hedges rather than directional bets. Historical exact-trade coverage may be narrower than OHLCV history, and live bars can change until complete. These limits are why a footprint should improve a price plan, not replace one.
A useful review compares the same setup with and without the footprint condition. Track candidates rejected by the extra filter, missed winners, timing delays, spreads and fees. A sharper-looking chart is not evidence of a better trading outcome until those costs and exclusions are counted.