← Strategies

Settlement projection none

Trade the arithmetic, not the chart. In the last minute of a 15 minute Bitcoin market, part of the settlement average is already decided; this rule projects it and buys the side it favours when the projection sits far enough from the line. You set the window, the distance and the price cap. The lab's live rule in this family runs with settings that are held until its verdict; these defaults are neutral.

Markets: Bitcoin price up down · every 15 minutes KXBTC15M

Modify

The lab's record

Not a lab result: the lab's own live rule in this family uses held settings. The public measurement behind the idea, over 635 real settlements: with 10 seconds left the projected average missed the final value by about 47 cents, the spot price by about 4 dollars; in close calls ($5 to $20 from the line) the projection picked the right side 98 times in 98, spot 92 in 99. Whether the side it picks is worth what the market charges is the whole question.

Money: Paper
Period: measured Oct 2026

Read the write-up

Pass mark

written before the first run

On a member's copy, a strategy cannot run until its pass mark is written down: minimum trades, cents a contract after fee, both halves positive, a cap on one day's share of profit, a loss-rate cap, and a judgement date. The verdict is computed only from fills made after the mark, so it can never be fitted to a result that already exists. The lab's own bar is the default.

The lab tested this shape

public numbers, 2026-10-26

The lab tested this shape. paper: 1 graded ledger, 135 trades, +14.60¢ a contract weighted, 1 passed the bar.

Run	Market	Trades	¢ / contract	Thirds	Verdict
Last minute (projection, old fill rule)	BTC15M	135	+14.60	8.51 / 16.19 / 17.45	pass	read

Information, not a block. Rows marked real are from the public real-money table; the rest are paper ledgers from the lab's all-time grade. No live-lane settings or maker parameters are shipped.

Your paper result

Fills12

Win rate67%

¢ / contract-7.75

Touched, not filled0

Settled fills12 (8 won) · -93.0¢ net · 2026-10-31 to 2026-11-01
First half vs second half-19.83 on 6 · +4.33 on 6
Best single day's share of profit– (2 days with fills)

Scored only on settled fills, the way the lab's bars are. Export this strategy's orders (CSV)

Backtest

not run yet

This rule reads the live settlement index, which Kalshi's free history does not include, so it cannot be backtested here. Run it on paper and judge it by the fills.

Rule

Declarative: Ping never runs code from a strategy file.

Kind: projection
Series: KXBTC15M
Acts: 5 to 60 s before close
Price range: 50¢ to 95¢
Contracts: 1
Order life: 45 s
Max open: 1
min distance usd: 10
max price cents: 95

Strategy file (TOML)

id = "projection"
name = "Settlement projection"
series = [
    "KXBTC15M",
]
description = "Trade the arithmetic, not the chart. In the last minute of a 15 minute Bitcoin market, part of the settlement average is already decided; this rule projects it and buys the side it favours when the projection sits far enough from the line. You set the window, the distance and the price cap. The lab's live rule in this family runs with settings that are held until its verdict; these defaults are neutral."
contracts = 1
order_life_seconds = 45
max_open = 1
cooldown_seconds = 0
pack = "core"

[gate]
seconds_before_close_min = 5
seconds_before_close_max = 60
max_price_cents = 95
min_price_cents = 50
min_contracts_offered = 1

[rule]
kind = "projection"

[rule.params]
min_distance_usd = 10
max_price_cents = 95

[record]
status = "none"
summary = "Not a lab result: the lab's own live rule in this family uses held settings. The public measurement behind the idea, over 635 real settlements: with 10 seconds left the projected average missed the final value by about 47 cents, the spot price by about 4 dollars; in close calls ($5 to $20 from the line) the projection picked the right side 98 times in 98, spot 92 in 99. Whether the side it picks is worth what the market charges is the whole question."
real_money = false
period = "measured Oct 2026"
link = "https://pondletter.com/kalshi-bot-course-advanced-1-settlement-average/"

Paper orders

Placed (UTC)	Market	Side	Price	Status	Fill	Fee	Result	P&L
11-01 12:35	KXBTC15M-26NOV011236-81250	YES	94¢	open
11-01 12:30	KXBTC15M-26NOV011231-81000	NO	86¢	open
11-01 12:09	KXBTC15M-26NOV011210-81250	NO	86¢	filled	through	0¢	win	+14.0¢
11-01 00:24	KXBTC15M-26NOV010025-79500	YES	65¢	filled	at price	0¢	win	+35.0¢
10-31 21:24	KXBTC15M-26OCT312125-80000	NO	91¢	filled	through	0¢	win	+9.0¢
10-31 20:39	KXBTC15M-26OCT312040-80000	YES	95¢	filled	through	0¢	win	+5.0¢
10-31 19:24	KXBTC15M-26OCT311925-81500	NO	50¢	filled	at price	0¢	loss	-50.0¢
10-31 16:54	KXBTC15M-26OCT311655-78000	NO	87¢	filled	through	0¢	win	+13.0¢
10-31 16:24	KXBTC15M-26OCT311625-78000	NO	54¢	filled	through	0¢	loss	-54.0¢
10-31 11:24	KXBTC15M-26OCT311125-81750	NO	71¢	expired
10-31 09:54	KXBTC15M-26OCT310955-79250	NO	62¢	filled	through	0¢	loss	-62.0¢
10-31 08:24	KXBTC15M-26OCT310825-78750	YES	79¢	filled	through	0¢	win	+21.0¢
10-31 03:54	KXBTC15M-26OCT310355-80500	YES	50¢	filled	at price	0¢	loss	-50.0¢
10-31 03:09	KXBTC15M-26OCT310310-78000	YES	81¢	filled	at price	0¢	win	+19.0¢
10-31 00:54	KXBTC15M-26OCT310055-79500	NO	93¢	filled	at price	0¢	win	+7.0¢