Tape fit answers a question delta cannot: where the short strike sits relative to the range the stock has actually traded. It blends three components into one 0-100 subscore. Placement, half the weight, measures how far outside the recent trading box the short strike sits as a fraction of spot. Patterns, 30 percent, checks whether a triggered chart pattern opposes the trade, and zeroes the component when a measured-move target breaches the short strike. Regime, 20 percent, favors a ranging tape for premium selling and a trending one for debit trades. A negative dealer-gamma regime caps the regime component for credit trades.
How these pages were written: from the engine source as deployed in September 2026, with thresholds quoted as the code holds them. Thresholds change with releases and the changelog records each change. This is a description of arithmetic, not advice about what to trade.
Source module: strategy/tape_gate.py, function score_tape_fit, built on the Tech Charts range and pattern machinery.
The other subscores in the agent's gate are all options-chain and portfolio economics: liquidity, volatility regime, structure fit, event risk, portfolio fit. Nothing among them knows where the short strike sits relative to the observed trading range, whether a chart pattern is live against the trade, or whether the tape is trending against the spread's bias. Short delta is the only distance proxy, and delta is an implied-volatility guess: a short put at a range floor the tape tests weekly and one below the whole prior washout can carry the same delta while having very different real-world odds of being touched. Tape fit turns the chart machinery into that missing input.
Placement carries half the weight. For a credit vertical it measures how far outside the recent-tape box the short strike sits, as a fraction of spot; an iron condor scores both wings. The box is the recent-tape tightened range rather than a long-lookback high and low, so a stale ceiling from months ago cannot make a bad strike look safe. Distance is also expressed against an expected-move ramp scaled from implied volatility and days to expiry, so a strike two expected moves away scores higher than one inside the first.
Debit verticals invert the frame, because their edge is the move: placement scores the reachability of the breakeven, highest when the tape has already demonstrated the level, with the breakeven inside the recent box or already achieved, and lowest when the required move exceeds both the box and the expected-move ramp.
Patterns carry 30 percent. Only patterns with status triggered and a score of 60 or more count; forming patterns are conditional, and whipsaw or completed patterns are not fresh signals, so they are ignored. A triggered pattern opposing the spread's bias whose measured-move target breaches the short strike zeroes the component. Opposing without a breach scores 30. A triggered pattern aligned with the trade scores 85. A pattern invalidated by a close beyond its own stop never enters the summary at all, in either direction. A failed break contributes its inverted bias as a mild signal, since busted patterns tend to move opposite their original signal, but never the breach zero, because a busted pattern has no reliable measured-move target.
Regime carries 20 percent. For a credit trade a ranging or borderline tape scores 80, a trend against the spread's bias scores 20 (the classic slow bleed), a trend aligned with it scores 60, and a condor in any trend scores 30. Debit trades invert: a trend aligned with the thesis scores 85, a trend against it 15, and a range 35, because a mean-reverting tape resists the move a debit needs.
A dealer-gamma overlay then applies when gamma-exposure data is present. A negative-gamma regime, where dealer hedging amplifies every move, caps the regime component at 30 for credit trades. A positive-gamma regime, where dealers damp every move, caps it at 30 for debit trades. Negative gamma gives debits no bonus, because it amplifies moves in both directions and the journal holds no evidence yet that buyers win there. The former nudges for the largest gamma wall were made informational only after a pre-registered, control-matched test found the top wall statistically indistinguishable from the next four strikes; the wall is still journaled for display, but it no longer moves the score.
Everything here fails open: when the chart machinery cannot produce a range, a pattern summary or a regime, the component is absent and the blend renormalizes over what it has, and the card says which inputs were missing.
Where the short strike sits relative to the range the stock has actually traded, whether a triggered chart pattern opposes the trade, and whether the tape is trending against the spread's bias. Two strikes with the same delta can sit at very different distances from the levels the tape has been testing.
Placement against the recent trading box carries 50 percent, triggered patterns 30 percent and the trend or range regime 20 percent. A measured-move target that breaches the short strike zeroes the pattern component; a negative dealer-gamma regime caps the regime component at 30 for credit trades.
The component is absent, the blend renormalizes over what it has, and the card names the missing input. Nothing here fails closed or fills in a value.
More in Methodology: how the engine computes: How Exit Rules Are Applied to Open Positions · How the Conviction Score Is Computed · How a Structure Is Chosen: Signal, Volatility, VIX