polymarket-tui/blog/trade-tape

Reading the Polymarket Trade Tape

An order book shows you what people are willing to do. The trade tape shows you what they actually did. Both matter, but they answer different questions - and on Polymarket, where every print is public, the tape is the closest thing to watching other traders think in real time.

01An offer versus a fact

Why the tape and the book disagree on purpose

A resting order in the book is a standing offer: "I'll buy at 51c" costs the trader nothing until someone accepts it, and it can be pulled any time. A print on the tape is the opposite - it already happened. Two named counterparties agreed on a price and size, capital changed hands, and nothing about it can be withdrawn. The book tells you what liquidity is available right now; the tape tells you what liquidity actually got used, by whom, and in which direction. A market can have a deep, tight book and still print nothing for an hour - depth with no conviction behind it - or a thin book that nonetheless prints steadily as size trades through it one lot at a time.

02Five fields, one story

What a print actually contains

Every row on the tape carries the same five fields: a timestamp, a side (BUY or SELL), a size, a price, and the trader who took it. Polymarket's public trades endpoint (data-api.polymarket.com/trades?market=<conditionId>, the same one polymarket-tui's live tape rail polls) returns exactly this, plus a wallet address and display name so you can tell whether a run of prints came from one trader or many. Nothing here is inferred - it is the matched-trade record, not a model's guess at intent.

03Who crossed the spread

Side tells you who was in a hurry

"BUY" and "SELL" on a print name the aggressor - whoever crossed the spread to trade immediately, rather than the counterparty who was already resting in the book. A BUY print means someone paid the ask because they wanted in now; a SELL print means someone hit the bid to get out now. Reading a run of prints as a sequence of aggressor decisions is different from reading the book, which only shows patient orders waiting to be taken. A market that prints mostly BUY at the ask is being bought into by people unwilling to wait for a better price - a stronger signal than the same price sitting unchanged in a quiet book.

04A live example

Book and tape, side by side

Here is a genuinely two-sided market as of 2026-08-05 09:17 UTC: Polymarket's "no change" outcome on the Fed's September 2026 rate decision, last trading at 51c on $370,583 in 24-hour volume and $366,328 in liquidity, fetched from Gamma and the CLOB order book at that moment. The book shows a 1c spread with real size resting on both sides:

TOP OF BOOK, "NO CHANGE" - SEP FED DECISION, 09:17 UTC
bid  51c x 3,093     ask  52c x 9,184
bid  50c x 37,023    ask  53c x 30,611
bid  49c x 79,589    ask  54c x 48,794

The tape from the same window shows how that quote actually got made - a mix of small BUYs lifting the 52c ask and small SELLs hitting the 51c bid, from several different traders, none large enough alone to move the quote:

TAPE, "NO CHANGE" - SEP FED DECISION, 08:53-09:09 UTC
09:09:12  SELL  20.0  @ 51c   Impish-Disparity
09:07:59  SELL   3.8  @ 51c   Unripe-Suite-Nightlight
09:04:15  BUY    3.8  @ 52c   Unripe-Suite-Nightlight
09:03:09  BUY   19.2  @ 52c   Rotten-General-Mouth
09:01:21  BUY    5.0  @ 52c   General-Stonework
08:58:41  SELL   2.6  @ 51c   (unnamed)
08:58:18  BUY    1.9  @ 52c   Energetic-Worm

Read together: the quote itself (51c bid, 52c ask) hasn't moved across this whole window, but the tape shows genuine two-way flow crossing it in both directions at similar size - nobody is leaning on the market hard enough to walk the price, and depth three levels deep on both sides backs that up. That is a market where 51c reflects real disagreement, not one side quietly waiting for a chance to sell into strength.

05Size and clustering

One whale or many small trades?

The same net volume can come from very different flow, and only the tape distinguishes them. A single 50,000-share print is one trader expressing a view all at once - often the more informative signal, since someone was willing to move a large amount of capital in one decision. Fifty prints of 1,000 shares each, arriving steadily over an hour, usually means either an algorithm working an order in slices to avoid moving the price, or many independent traders happening to agree. The example above leans toward the second case: names repeat (Unripe-Suite-Nightlight buys then sells within minutes, a market-maker's two-sided pattern) but no single print dominates the window. A tape dominated by one name in one direction is the pattern worth stopping to look at.

06Reading them together

The book is the present, the tape is how it got there

Neither view alone is enough. The book without the tape can't tell you whether today's quote is fresh conviction or last week's orders nobody has bothered to cancel. The tape without the book can't tell you whether the next trade fills at the same price or walks five cents up an empty ladder. Watching both is how you catch the moment a quiet market starts printing one-sided - the tape usually moves first, and the book's mid catches up a few seconds later once the resting side gets taken out.

07In the terminal

The tape as a rail, not a separate screen

In polymarket-tui, the trade tape sits as a live rail beside the order book on every market page rather than behind a separate tab - each row colored by side, newest on top, polling the same public trades endpoint used above. Cursor onto any row to open that trader's public portfolio and see what else they hold. No account is needed to watch it.

$ uv tool install polymarket-tui
$ polymarket-tui

The install page has Homebrew and one-liner options.

Prices and trades are a snapshot from 2026-08-05 09:17 UTC and will have moved since. More posts on the blog index, or subscribe via RSS. Not affiliated with Polymarket; nothing here is financial advice.