Track record / methodology

/ ensemble_meta

Typical horizon: INTRADAY • Season 1 — all time

What it does

Mirrors fresh signals from strategies the calibration table proves UNDERCONFIDENT (realized win rate beats predicted probability on 30+ samples) — turning calibration measurement into capital allocation.

Key parameters

behavior-shaping settings only
count 1

Record (fee-inclusive, this season)

W–L
0–7
Open
0
Positions
7
Realized
$-1.48
Marked
$-1.48
Fees
$0.03

Calibration

predicted vs realized win rate, by confidence bucket
bucket 4 3 signals p̂ 0.37 → 0.33
bucket 5 3 signals p̂ 0.45 → 0.0
bucket 7 1 signals p̂ 0.62 → 0.0
Caveats: All results shown are simulated (paper) trading performance, not actual trading, unless explicitly labeled live. Fees and slippage are modeled (Kalshi's taker fee formula plus 1¢ slippage), which cannot fully reflect real market impact or liquidity constraints. Past performance, simulated or actual, is not indicative of future results. Parameters shown are the current configuration and change over time as the strategy is tuned. Nothing on this page is investment advice — see the full disclaimer.