What the Order Book Actually Shows You
An order book is a list of resting limit orders at various price levels. The bid side shows buyers willing to purchase at specific prices, and the ask side shows sellers offering at specific prices. The gap between the best bid and best ask is the spread. For liquid crypto pairs, the visible order book on a single exchange represents maybe 20-40% of actual available liquidity, since a large share of trading now happens through hidden orders, iceberg orders, and orders routed through aggregators.
Even with those limitations, the order book contains useful information. The key is knowing what to look for and, just as importantly, what is likely to be fake.
Order Book Imbalance as a Directional Signal
The simplest useful metric from order book data is the bid-ask imbalance. Take the total bid volume within some distance of the mid-price (say, 0.5%) and compare it to total ask volume within the same distance. If bid volume is 3x the ask volume, there is more resting buy interest than sell interest near the current price. Empirically, this imbalance has modest but consistent predictive power for short-term price direction, on the order of seconds to minutes.
The intuition is straightforward. If there are many more resting bids than asks near the current price, an incoming market sell order has to chew through more liquidity to push price down than a market buy order needs to push price up. Price follows the path of least resistance, and order book imbalance tells you which direction offers less resistance.
The catch is that this signal decays quickly. Order book imbalance measured right now predicts price movement over the next 30 seconds reasonably well, but its predictive power drops off sharply beyond a few minutes. Resting orders get cancelled and replaced constantly, and what the book looks like at 10:00:00 may bear little resemblance to what it looks like at 10:05:00.
Spoofing and How to Spot It
A significant portion of visible order book depth is not genuine trading interest. Spoofing involves placing large visible orders with the intent to cancel them before they execute. A trader might place a 500 BTC bid wall at a price level to create the impression of strong support, encouraging other traders to buy, and then cancel that order once price has moved in the desired direction.
There are a few signatures that help distinguish genuine orders from spoofing. Genuine large orders tend to be placed and left in the book for extended periods. They often appear at psychologically significant price levels (round numbers, previous support/resistance). Spoofed orders tend to appear suddenly, move price through the influence they exert on other traders, and then vanish within seconds or minutes.
You can track this by monitoring order book snapshots over time and looking for large orders that appear and disappear without trading. If a 200 BTC bid appears at $64,000, sits for 45 seconds while price moves up $50, and then vanishes, that was almost certainly a spoof. Aggregating the frequency and size of these phantom orders gives you a sense of which direction the spoofers are trying to push the market, which, ironically, is itself a useful signal.
Absorption and Stacking Patterns
When a large resting order absorbs repeated market orders without the price level breaking, that tells you something meaningful. If there is a 300 BTC bid at $63,500 and you watch 150 BTC of market sells get filled against it without the price dropping below that level, the bid is genuine, well-capitalized, and the entity behind it wants to accumulate at that price. This is called absorption, and it is one of the more reliable order flow signals.
The opposite pattern is stacking, where a trader places a cluster of large orders at consecutive levels to create a wall. If you see 100 BTC bids at each of $63,400, $63,450, $63,500, $63,550, and $63,600, that is a deliberate attempt to establish a floor. These stacked orders are more likely to be genuine when they appear at technically significant levels and when the entity behind them has demonstrated willingness to get filled (through absorption at the first levels).
Practical Application
For most traders, the actionable use of order book data is not to predict specific price movements but to improve execution. If you want to buy and you can see a large resting ask that has been absorbing bids at $64,200, you know that level is a ceiling in the short term. You might place your buy below that level and wait for a dip, rather than paying the ask and getting filled at a worse price.
Similarly, if you see a sudden appearance of heavy bids right below the current price, you might wait before selling, since those bids (whether genuine or spoofed) are likely to support the price for at least a few minutes.
The tools for this have gotten more accessible. Platforms like Bookmap, Coinalyze, and even TradingView with exchange data feeds let you visualize order book depth and flow in real time. The information advantage is not in having access to this data, since everyone does, but in interpreting it correctly and acting on it faster than the median participant.