How to Read the Data
Understanding the signals, probabilities, and metrics displayed throughout Arctic Odds.
Temperature Models
Arctic Odds runs two temperature prediction models that analyze different aspects of daily weather:
Predicts daily high temperatures using NWS forecasts, historical patterns, and real-time atmospheric data.
Predicts daily low temperatures by analyzing overnight cooling patterns, cloud cover, and radiative cooling conditions.
City Summary Cards
Current High
The running high temperature observed today so far, pulled from live weather data.
Current Low
The running low temperature observed today so far. This updates as overnight lows are recorded.
Model Forecasts
Our model's predicted high and low temperatures for the day. These are the values we're betting on.
Signal Count
Combined number of active trading signals for this city today across both high and low temperature brackets.
Best Edge
The highest edge among all active signals (high or low) for this city.
Signals Table
Type (HIGH/LOW)
Indicates which temperature model generated the signal:
HIGH = Daily high temperature prediction
LOW = Daily low temperature prediction
Bracket
The temperature threshold for the market contract. For example, "≥45°F" means the contract pays out if the temperature reaches or exceeds 45°F.
Signal (YES/NO)
YES = Our model thinks the temperature WILL hit this bracket, and the market is underpricing it.
NO = Our model thinks the temperature will NOT hit this bracket, and the market is overpricing it.
Model vs Market
Model = Our calculated probability that the temperature will hit the bracket.
Market = The implied probability based on current Kalshi contract prices.
When these diverge significantly, there's potential edge.
Edge
The percentage difference between our model's probability and the market price. Higher edge = bigger potential opportunity.
Example: If our model says 70% chance and the market is priced at 62%, the edge is +8%.
Understanding Edge
Edge is the foundation of profitable trading. It represents the difference between what we believe the true probability is and what the market is pricing.
Positive edge means our model thinks the market is underpricing the outcome. We'd want to buy YES contracts.
Negative edge means our model thinks the market is overpricing the outcome. We'd want to buy NO contracts.
We typically only generate signals when the absolute edge exceeds our threshold (usually 5%+), ensuring we're only acting on meaningful discrepancies.
Dashboard Stats
Total Signals
All signals generated by our model, including historical ones.
Active Signals
Signals for markets that haven't resolved yet - these are current opportunities.
Win Rate
Percentage of resolved signals where our prediction was correct.
Avg Edge
The average edge across all active signals - higher is better.
Ready to explore the signals?
View Weather Dashboard