Reading the tape
Gamma describes the terrain. The tape describes what is being done on it right now. Every column in the Scanner is an attempt to answer one question — is this trade unusual, and unusual in a way that carries information? This chapter explains what each measure sees, and just as importantly what it cannot see.
Volume against open interest
Volume counts contracts traded today. Open interest counts contracts that exist. The ratio between them is the first and most durable filter for unusual activity, because it separates ordinary churn from something new arriving.
A V/OI above 1 is the clearest structural evidence available that something new is happening at that strike. It cannot be produced by existing holders trading among themselves in normal size. This is why the Scanner exposes V/OI as a first-class column and why the "Cut noise" preset filters on it.
The limitation is timing. Open interest is published once daily, overnight, so today's V/OI compares live volume against yesterday's position count. On a contract that saw heavy activity yesterday, today's ratio understates what is happening — the denominator has already absorbed it.
Sizzle
V/OI asks whether today is large relative to existing positions. Sizzle asks a different question: is today large relative to this contract's own normal. It divides today's volume by the contract's median daily volume over the last twenty completed sessions.
Three deliberate choices sit behind that number, and each exists to stop it lying to you.
- Median, not mean
- A single enormous day in the baseline would inflate an average and permanently suppress the sizzle reading afterwards — the contract would look calm precisely because it had one violent day. The median ignores that outlier.
- Gated at eight sessions
- A contract with three days of history has no meaningful "normal". Rather than print a spurious 40× on a thin baseline, FlowRadar shows a dash. An absent number is more honest than a confident wrong one.
- Capped at 50×
- Beyond a point the exact multiple stops carrying information — 80× and 300× both mean "far outside anything this contract normally does". The cap keeps sorting sane and stops one freak reading dominating the board.
Use sizzle and V/OI together. High sizzle with low V/OI is a busy contract that already had large open interest — real activity, but on established ground. High on both is the configuration worth looking at: unusual for this contract and large relative to what was already there.
Sweeps and blocks
These two tags describe how an order was executed, and they mean opposite things about the trader behind it.
- Sweep — urgency
- The trader wanted the position immediately and was willing to pay worse prices at successive venues to get filled now. Sweeps are read as conviction about timing: someone who thinks they have hours rather than days does not work an order patiently. FlowRadar requires prints across at least two exchanges inside a 50-millisecond window, aggregating to 100 contracts or more.
- Block — size
- A single large trade, typically arranged rather than swept up off the screen. Blocks say something about conviction in magnitude, not urgency, and they are far more likely to be institutional. FlowRadar requires 250 contracts and $250,000 of premium in a standalone print.
A row can carry both tags. That is not a contradiction — it means the contract saw a sweep and, separately, a large standalone print on the same tape. Two different trades, both notable.
Side tagging: the arrows
A sweep tells you someone was in a hurry. It does not, by itself, tell you which direction they were leaning. For that we compare each print to the prevailing national best bid and offer at the moment it occurred.
The side then combines with the contract type to produce the arrow: calls bought and puts sold read bullish; calls sold and puts bought read bearish. Where the evidence does not lean clearly either way, no arrow is shown.
Arrows are also deliberately suppressed on deep in-the-money contracts. Those have wide spreads and sparse quotes, so the midpoint rule is guessing more than measuring. Showing no direction is more useful than showing a direction the data cannot support.
The unavoidable limitation, and the one most often ignored elsewhere: this measures aggression, not intent. Someone opening a new bullish position and someone closing an old bearish one both buy at the ask, and the tape cannot tell them apart. A hedge against stock, one leg of a spread, and an outright directional bet all print identically. Treat the arrow as "who crossed the spread", never as "what someone believes".
Delta, premium and location
Three supporting columns give a trade its shape.
- Premium is dollars committed — contracts times price times 100. It is the honest measure of how much someone actually risked, and it separates a genuine institutional position from a large number of cheap lottery tickets.
- Delta is directional exposure per contract. A 0.15-delta call is a low-probability punt; a 0.70-delta call is close to owning the stock. Same ticker, same direction, entirely different statements about conviction.
- Location shows where the strike sits relative to spot. Far out-of-the-money activity is cheap and speculative; near-the-money activity costs more and carries more gamma, which feeds straight back into the dealer hedging of Chapter 1.
Premium and delta together are the fastest filter against noise. Large premium at meaningful delta is someone taking real risk. Enormous contract counts at 0.02 delta for pennies are, more often than not, exactly what they look like.
The score
The score is a convenience for sorting, not a verdict. It combines the measures above into one number so the board can rank hundreds of contracts, adding weight for unusual sizzle, for sweep and block execution, and for the ratio of volume to open interest.
It is a ranking aid and should be treated as one. A high score means several unusual things coincided on the same contract, which is a reason to look — not a reason to trade. Two rows with identical scores can tell completely different stories, and the columns behind the number are where the actual information lives.
No composite score has predictive power on its own. If it did, it would stop working the moment enough people used it. What it does reliably is direct attention, which is the real problem when hundreds of thousands of contracts trade every day.