GET /v1/insights/strategies or acting on a match.
Listing rules
The first rule reads: across both venues, Culture markets whose Yes was priced 80¢ to 90¢ one day before close resolved No 37% of the time. No cost 15¢ on average, so $100 spread across all 119 came back as $243, and the 95% range on that rate puts the return between $86 and $200.
The id
{venue}~{category}~{horizon}d~b{band}~{side}, for example kalshi~tech-and-science~7d~b1~yes.
Ids are stable, so a rule can be bookmarked and re-read.
How returns are computed
A rule is one calibration band with a side attached. With the band’s average pricep and the side’s win rate w:
- Price. For Yes,
pis the band’savg_price_usd; for No, it is1 − avg_price_usd. History prices markets at the daily midpoint, not at an ask. - Return. $100 spread evenly buys `100 / p` contracts; winners pay $1 each.
return_per_100_usd = 100 × (w / p − 1), before fees. - Range.
wis bounded at 95% byw ± 1.96 × √(w (1 − w) / n), withnthe settled markets, then clamped to 0 and 1.return_low_usdandreturn_high_usdare the return at the two ends. Large samples give a narrow range; small samples give a wide one. - Horizons. A band has an edge on a horizon when its Yes rate differs from its average price by at least 2¢ in the side’s direction.
horizonscounts the horizons where it does.
return_low_usd, ties by markets, so a large sample with a steady edge outranks a lucky small one.
Markets that fit now
- is on the rule’s venue (
allmatches both) and whose event is in the rule’s category; - has a Yes ask at or above
from_usdand belowto_usd, whatever the side; - closes inside the horizon’s window, measured from now: within 2 days for the 1-day horizon, 4 to 11 days for 7, 20 to 45 days for 30;
- has not started, for games;
- has a No ask, for a No rule.
price_usd, the ask for the rule’s side, and closes_at. Matches come tradeable first, then most traded, then soonest to close, and the liquidity floors apply unless include_thin=true. limit runs from 1 to 100, default 30. An id that is well-formed but has no rule behind it returns 404 with strategy_not_found.
Related
- Calibration: the bands the rules come from.
- Liquidity: the floors applied to matches.
- Signals: the ticker mixes the strongest rule’s matches with moves and fills.
- Strategy matches: watching a rule’s matches over time.