NBA · Moneyline · Alexander Morgan “The Quant”
Toronto Raptors moneyline
Published odds
-104
kalshi · 1.96 dec
Model prob.
59.1%
Market (no-vig)
50.0%
Probability edge
9.1 points
Estimated return / stake
+10.0% (capped)
After uncertainty: +10.0% (capped)
Stake
1.5u
confidence 75%
A large model edge is a reason to be cautious, not a promise.
We cap displayed expected return at +10% and probability edge at 10 percentage points. Higher estimates are marked “Estimate caution.” This is a display safeguard, not proof the true edge is 10%. Original estimates remain visible in each pick’s analysis. Missing injuries, lineups or stale odds can inflate an estimate; any pick can lose.
Original estimates at publication (uncapped)
Expected return: +15.9% · Probability edge: 9.1 percentage points · Uncertainty-adjusted return: +12.1%. These are model estimates, not realized profits or guaranteed advantages.
Analysis
Executive summary
Selection: Toronto Raptors to win at Washington Wizards, priced at 1.9615 decimal on kalshi.
The calibrated ensemble gives Toronto a 59.1% win probability, versus 50.0% in the no-vig market. That supports a 1.5-unit position: combine the signals, calibrate to the market, and require a margin of safety—not certainty.
Statistical rationale
- Toronto leads in Elo: 1564.53 vs. 1321.98.
- Across the supplied 60-game season samples, Toronto scored 115.57 and allowed 111.48 points per game; Washington scored 113.25 and allowed 123.42.
- The scoring projection favors Toronto, 120.53–113.48.
- The raw ensemble probability of 72.7% is reduced to 59.1% after calibration, with a 60.0% market weight.
Recent performance
Over the supplied 10-game recent samples:
- Toronto: 117.4 points scored and 111.4 allowed per game.
- Washington: 113.6 scored and 121.0 allowed.
Rest favors Washington: 2 days, compared with Toronto’s 1 day.
Projected probabilities
- Toronto win: 59.1%
- Washington win: 40.9%
- Push: 0.0%
The supplied model confidence is 74.9%; this is distinct from Toronto’s win probability.
Market price comparison
The 1.9615 offered price implies a 51.0% break-even probability. The supplied no-vig market probability is 50.0%, leaving Toronto a 9.1-percentage-point model edge against that benchmark.
Expected return is 15.9% per unit staked, with supplied adjusted EV of 12.1%.
Reasons supporting the selection
- Elo, season scoring, recent scoring, and the projected score favor Toronto.
- Market calibration substantially tempers the raw model estimate.
- The offered odds exceed the model’s 1.6926 fair price, providing a margin of safety.
Factors that could invalidate it
- Washington’s rest advantage could matter more than estimated.
- Specialist outputs are mixed: matchup 27.3%, pattern 30.6%, momentum 35.2%, and recovery 54.2%. Their probability-side definitions are not supplied, limiting interpretation.
- Unavailable personnel information could materially change the assessment.
- A shorter market price could eliminate the value.
Uncertainty and data limitations
Injuries, lineups, starting goalies/pitchers, weather, and advanced metrics (xG, EPA, DVOA, FIP) were not considered because they were listed as unavailable. Referee and active-official information were also unavailable and not considered.
The supplied season and recent samples are not dated. Specialist disagreement and missing personnel inputs warrant caution; the selection is not a guarantee.
Price threshold
Break-even decimal odds = 1 / win probability = 1 / 0.5908196999136142 = 1.6926.
Odds must exceed 1.6926 for positive model-implied expected value. The offered 1.9615 clears that threshold; break-even itself provides no margin of safety.