Signal methodology

Category Methodology

Scoring weights and Bayesian priors vary by market category. Here is the full breakdown.

← Back to Methodology
01

Component Weight Matrix

Weights reflect where genuine skill concentrates in each category. Politics scores Consistency highly because political edge is durable; Sports scores Edge highly because market inefficiency is event-specific.

CategoryEdgeSluggingCapacityConsistency
Politics30201535
Elections40302010
Iran45351010
Geopolitics35301520
Sports45152515
Esports35251525
Crypto45252010
Tech35301520
Finance45153010
Economy35302015
Weather25203520
Culture35301520
Mentions40153015

All values are percentage weights. Each row sums to 100.


02

Per-Category Bayesian Priors

The Bayesian prior represents the population of skilled (not lucky) wallets in each category. Weather's 75% skilled baseline is unusually high because public ECMWF forecast models already hit ~70% — wallets must clear that bar to be considered skilled.

CategorySkilled WRBreakevenNotes
Politics60%53%Tight band; >65% over <40 N is variance
Elections57%52%Even 538-tier hits ~57% on contested races
Iran60%54%Tiny N — requires posterior CI to fully clear 54%
Geopolitics58%53%Wider tail than Politics — genuine edge rare
Sports55%49%After vig-equivalent
Esports58%50%Public dumber, books softer than sports
Crypto60%52%Reflexive markets — concentrated edge bursts
Tech64%52%Soft-info edges available
Finance60%52%Efficient vs CME FedWatch — alpha narrow
Economy57%52%Fatter tails than Finance
Weather75%70%Public ECMWF gets ~70% — this is the unusual one
Culture58%52%Real narrative noise; insiders barely above chance
Mentions62%52%Quant-tractable but no canonical public model

Skilled WR is the prior mean for wallets that have genuine edge in this category. Breakeven is the vig-adjusted minimum win rate to generate positive EV. The posterior 90% credible interval must fully clear the breakeven rate before a wallet qualifies.


This platform is for educational purposes only. Not investment advice. Past signal accuracy does not guarantee future results. All prediction market participation carries risk.

Category Methodology | ARCANIQ | Arcaniq