How to cite TapeXray data in 2026

Where do TapeXray's market counts come from?

TapeXray is a dated archive of Kalshi and Polymarket market listings, state access rows, and cross-venue price comparisons. The headline coverage number is not a scrape-time guess. It is the catalog count shown in the market archive, with the archive date carried alongside the number so a citation can say when the tape was read.

The catalog is broad, but it is not every possible derivative, sportsbook line, or offshore price. It is Kalshi plus Polymarket. That scope matters because the archive can support claims about those two venues and cannot support claims about all betting markets. The right citation names the venue, the page, and the archive date. For example: “TapeXray market catalog, archive updated 2026-09-21.” If the claim is about state availability, cite the state matrix instead of the market catalog.

What does the live Kalshi-Polymarket board measure?

The live board compares matched Kalshi and Polymarket markets, then separates headline price gaps from gaps that still exist after the fee model. That distinction is the main reason to cite TapeXray instead of a screenshot. A 7c midpoint difference is not a 7c edge if both legs pay taker fees, have a wide bid-ask spread, or cannot be executed at the displayed midpoint.

The board has two different fee treatments. The fee-adjusted midpoint line subtracts a modeled round trip: entry plus exit on each venue. The locked-fork line is different. It prices the YES ask on the cheap venue and the NO leg from the other venue’s YES bid, then subtracts one-way taker fees because a fork is held to resolution. Those are not the same trade.

The important caveat: reading every one of roughly 10,900 matched pairs’ published Polymarket rate would overrun the board’s memory budget, so gap math runs in two passes. Pass one ranks and gates every match at the conservative 7% maximum Polymarket rate published for any category. Pass two re-prices only the handful of pairs that pass the gate at each market’s own published category rate (0.03 to 0.07, some categories fee-free), before the board splits pairs above and inside the fee line. A pair rejected before pass two - shown separately as a suspect pair - keeps the 7% maximum and carries no per-market rate.

How should TapeXray be cited?

A useful citation has four parts: page, metric, date, and limitation. The page tells a reader where the number lived. The metric tells them whether the claim was about market count, active market count, state access, or a fee-adjusted gap. The date keeps the claim from pretending to be timeless. The limitation keeps the number inside its lane.

Use the canonical article URL for methodology and the underlying page for the actual number. A market-count claim should cite the market archive. A state-availability claim should cite the state matrix. A gap claim should cite the live board with its board timestamp, not just the publication date of a blog post.

Good citation language is plain: “TapeXray, market catalog, archive updated 2026-09-21.” For the methodology page itself: “TapeXray, ‘How to cite TapeXray data in 2026,’ updated 2026-10-13.” If the sentence discusses fee-adjusted gaps, add the caveat that TapeXray’s live board ranks every match at a conservative 7% Polymarket maximum before re-pricing the gated pairs at each market’s own published rate.

1. For a market-count claim, cite the market archive and the archive date: 2026-09-21.
2. For a state-access claim, cite the state matrix and the verification date: 2026-09-18.
3. For a gap claim, cite the live board and the board timestamp: 2026-09-21 17:30 UTC.
4. For a fee-methodology claim, cite Kalshi’s fee schedule, Polymarket’s fee schedule, and TapeXray’s board caveat if the board estimate is involved.

What does TapeXray not measure yet?

The archive is not a settlement database yet. The resolutions section has 0 entries, so TapeXray does not publish calibration studies, payout-time studies, or venue settlement-speed rankings. A page can still say what Kalshi or Polymarket prices were on an archived date. It should not infer whether those prices were well calibrated after resolution unless a separate resolution record exists.

State access is measured separately from market coverage. The state matrix covers 51 jurisdictions and was verified on 2026-09-18. Nevada is the only jurisdiction where Kalshi, Polymarket US, and Robinhood event contracts are all marked unavailable in the matrix. Montana is marked Kalshi unavailable and Robinhood unavailable, while Polymarket US is marked limited. Those rows are access records, not predictions about litigation or future venue policy.

Fee data has its own boundary. Kalshi’s taker formula is published as fee = ceil(0.07 * C * P * (1 - P)) cents. Polymarket’s published taker fee uses a category rate, 0.03 to 0.07, some categories fee-free. TapeXray’s live gap board only reads that per-market rate for the pairs that pass its same-bet gate; every other match is screened at the conservative 7% maximum, so cite that two-pass split as a board assumption, not a single flat estimate.

How are corrections handled?

Corrections should be boring and traceable. TapeXray keeps a changelog for material updates and a contact page for correction requests. A correction request should identify the page, the number, the venue or state involved, and the source that conflicts with the published line. If the issue affects a dated table, the corrected table should keep a dated source line rather than silently dropping the old context.

The correction policy is especially important for gap pages. Most days there is nothing here. When the board shows a real fee-adjusted gap, the number still depends on the matching logic, fee assumption, bid-ask availability, and timestamp. That is why the related articles on fee-adjusted gaps and arb scanner errors spell out why midpoint gaps overstate what a bettor can actually lock.

The archive has 5 changelog entries in the site index. That count is not a completeness claim about every venue change in the market. It is the count of dated TapeXray change notes published in the archive.

• Market-count corrections should name the affected venue and archive date.
• State-access corrections should name the jurisdiction and the verification date.
• Fee-model corrections should distinguish the board’s conservative 7% first-pass maximum from a gated pair’s own published Polymarket category rate.

Questions people ask

Related reading

• Why arb scanner numbers are wrong in 2026Kalshi-Polymarket gap scanners disagree because raw mids skip fees, spreads, stale quotes, rule forks, and category rates. Here is TapeXray’s gate.
• Real Kalshi-Polymarket gaps right now in 2026Live count of real Kalshi-Polymarket gaps after fees, with pairs inside the fee line, suspect rejections, category mix, and the latest board timestamp.