Edge Finder
5 Cities • 3-Model Ensemble • Auto-refresh
Current Conditions
Loading...
Humidity
--%
Dewpoint
--°
Visibility
--
Pressure
--
Wind
--
mph from --
🔬 Enhanced Prediction Factors
Pressure Trend
--
--
Temp Rate
--
--
Marine Effect
--
--
Precip Impact
--
--
Frontal
--
--
Mixing
--
--
Dewpoint Spread
--
--
Bias Correction
--
--
NWS Tracking
--
--
🌡️ Diurnal & Seasonal Context
Time
--
--
Season
--
--
Sky
--
--
Diurnal
--
--
🌬️ Air Mass Advection
--Temp Δ
--
°F
Press Δ
--
mb
Rate
--
°F/hr
ETA
--
min
🎯 Enhanced Predictions (0.5° precision)
🔬 Model Inputs
▼Hourly Outlook
📅 Extended Forecast
🌡️ Tomorrow.io Analysis
Loading...📡 Ensemble Weights
Weights adjust based on historical accuracy
🧮 How It Works
3-Model Ensemble: Combines Tomorrow.io, Open-Meteo, and NWS forecasts. Only shows bets with ≥3% model probability to filter out extreme longshots.
YES Bets: When model probability > market probability, bet YES on that bucket.
NO Bets: When model probability < market probability, the bucket is overpriced — bet NO against it.
Edge = Model Prob - Market Prob (for YES) or Market Prob - Model Prob (for NO). 🔥 Strong = Edge > 30%
⚠️ Not financial advice. Past performance ≠ future results.
🔒 LOCK: End-band bucket ("X or higher" / "X or lower") where all models agree = near-certain win
📊 Bucket Agreement Overview
🧮 Safe Bets Philosophy
Same Bucket Required: All 3 models (Tomorrow.io, Open-Meteo, NWS) must predict within the SAME Polymarket bucket (typically 2°F range like "44-45°F").
Worst-Case Analysis: We calculate edge using each model individually. Only bets where EVERY model shows ≥10% edge are included.
🔒 LOCK Bets: When all models agree on an end-band bucket ("X or higher" or "X or lower") with ≥10% edge, this is flagged as a LOCK — the highest confidence opportunity.
Why End-Bands are Special: End-band buckets have infinite range on one side, so if all models agree the temp will be there, it's almost certain to hit.
⚠️ Not financial advice. Even "lock" bets can lose due to forecast errors.
Select a market above
📊 Bucket Analysis
| Bucket | Model % | Mkt % | YES Edge | YES EV | NO Edge | NO EV | Best Bet |
|---|---|---|---|---|---|---|---|
| Select a market | |||||||
📡 Data Sources (3-Model Ensemble)
📈 Source Accuracy & Weights
Weights auto-adjust based on historical accuracy
📈 Prediction Record
No predictions recorded yet
🧮 Methodology
3-Model Weighted Ensemble: Blends Tomorrow.io, Open-Meteo, and NWS forecasts with dynamic weights based on historical accuracy. Sources that perform better get higher weight.
YES Edge = Model Prob - Market Prob. Positive = underpriced bucket, bet YES.
NO Edge = Market Prob - Model Prob. Positive = overpriced bucket, bet NO against it.
EV% = Expected Value as percentage of stake. Both YES and NO EV are calculated independently.
⚠️ This is not financial advice. Past performance does not guarantee future results.
Accuracy by Timeframe
--
NWS MAE
--
Enhanced MAE
0
Verified
0
Enh Wins
--
MAE by Timeframe
Error Distribution
0-1° Excellent
0
0%
1-2° Good
0
0%
2°+ Poor
0
0%
Prediction History