Martial Systems LLC
v1 stopped · historical record
BTC Research · KXBTC15M-style

Kalshi BTC 15 min research, v1 closed

v1 stopped on 2026-10-14. Simulated paper record, unverified. Decision sample: n = 6,441 unique windows, 2026-07-27 to 2026-10-14. This page keeps the historical aggregates and the failure figures. v2 coming soon.

Skip to the historical shadow note

v1 stopped (2026-10-14)

Simulated paper figures. Unverified. The paper watcher is stopped. Decision sample: n = 6,441 unique windows, 2026-07-27 to 2026-10-14, public export 2026-10-14 05:51 UTC. On that sample the simulated path is +$27.85 on 2,138 win70 entries (about 1.3 cents a contract). The comparison below is that published sample. This stopped page does not load the export chart file.

A simulated positive share of 58.3% (n = 4,272) was graded on the last spot print, often on a partial window, with the scored row chosen after later rows. On the 2026-09-18 calibration (n = 4,661) the model Brier score was 0.169 and the market's own price was 0.166.

v2 coming soon. The next measurement starts from the Level-2 tape (2,086,954 snapshots) and from the exchange settlement.

v1 close note (PDF)

Why the sample failed

Decision sample, 2026-10-14 05:51 UTC. One comparison. The close note has the rest of the tables.

Sim path rises, $22.37 to $27.85. Settlement cooldown falls, $189.12 to $111.41. The stricter gate is higher on one ledger and lower on the other. Simulated paper, unverified.

Method note & pre-correction baseline (reset 2026-07-26)

Post-reset research sample (unique markets, last write wins; path $ from equity series only). Pre-reset archive PDF remains for method notes.

Download pre-correction research baseline (PDF) - estimator error, correction, and pre-correction aggregates (no trade tapes)

Sample size
Unique windows6,441
Date range2026-07-27 to 2026-10-14
Export2026-10-14 05:51 UTC
Statusstopped, unverified

Simulated paper figures. Unverified. v1 stopped on 2026-10-14. Decision sample: n = 6,441 unique windows, 2026-07-27 to 2026-10-14.

Calibration snapshot (2026-09-18, historical)

Mid-window forecast measurement on the published calibration sample. Simulated paper, unverified.

Sample n = 4,661
Model BS 0.169
Market BS 0.166
Calibration 2026-09-18 11:39 UTC

Published pair only. Simulated paper, unverified. v1 stopped.

Status
v1 stopped
Paper watcher stopped 2026-10-14
Decision sample
6,441 windows
Simulated paper, unverified
Date range
2026-07-27 to 2026-10-14
Published sample
Record
unverified
Simulated paper. See close note
Calibration
n = 4,661
2026-09-18 11:39 UTC
Export
2026-10-14 05:51 UTC
Stopped. File not loaded

Simulated paper measurements. Unverified. v1 stopped on 2026-10-14. The cards above are the published sample, not a live feed and not a trade list. Decision sample: n = 6,441 unique windows, 2026-07-27 to 2026-10-14. Model research-post-reset.

Historical: cumulative simulated path (1 contract)

Unique markets, last write wins. Simulated cooldown path. Unverified. v1 stopped on 2026-10-14. This page does not load the export chart file, and it does not list trades.

Historical: rolling path +rate (active windows)

Share of recent non-flat windows that were positive on the stopped sample. Separate from the shadow positive share. The series is not loaded here.

Historical: entry threshold (min predicted win probability)

Mean simulated dollars per counted row when the threshold moves (cooldown, 1 contract). Counts only. No individual rows. The bars are not loaded here.

Historical: shadow exit comparison

Same first_hot markets under alternate exit rules (1 contract, estimated fees). Simulated paper, unverified. v1 stopped. This page does not publish the rows or a chart.

Historical: cheap-open scalp

Isolated paper path. Unverified. The model does not choose the side. Separate from the first_hot shadow rows and from calibration. This page does not publish the path or the rows.

Instrument and method

Venue (context only). Kalshi is a CFTC-regulated U.S. exchange for event contracts. KXBTC15M-style markets are short-dated binaries on BTC vs a strike over a fixed interval. This page does not republish live books, quotes, or market-by-market rows.

Research process. Internally we estimate win probability and apply thresholds and risk rails. This page publishes the stopped sample summary only. It does not list trades, quotes, or rows that could be copied into an order.

Publication scope. Personal non-commercial research summary. Speculative risk remains; capital can be lost. Not investment advice.

Who we are

Martial Systems LLC - quantitative research on short-dated event contracts.

Terms

Event contract

Exchange-listed contract that pays on a defined outcome at a fixed time (here: BTC vs strike after 15 minutes).

YES / NO

Contract sides on binary event contracts. Public stats show aggregate side counts only.

Predicted win probability (p)

Model estimate that the chosen side wins by settlement. Thresholds such as p ≥ 70% leave out lower estimates.

Cooldown

Minimum wait after an entry before another entry is counted on the baseline path.

Shadow positive share (entry)

Share of unique first_hot markets with a positive simulated current exit (1 contract). Separate from the path positive share, which includes no-trade windows. Unverified.

Simulated path ($)

Equity-series sum of the 1-contract cooldown sim path on unique last-write windows. Excludes trade_outcomes closes. Unverified.

Path vs closed rows

The path series is a simulated paper measurement. Closed outcome rows are separate aggregates. No per-trade tape is shown. Unverified.

Aggregate-only public page

No tickers, strikes, settle spots, entry prices, per-window row lists, or CSV downloads on this site.